Decimal point tempos also might be necessary for calculating/programming metric modulations, (polyriddim, lol) but honestly, you could round up or down and nobody would notice. Watch this video ad-free on Nebula! nebula.tv/videos/adam-neely-this-tempo-is-trash-and-should-never-be-used
From a different perspective you could say it's similar to the increasing complexity of Cooking, Baking, Gastronomy... or wood floor, cement floor and linoleum... there's always good reasons for increased complexity, but yes... if it doesn't create a noticeable change then you have to ask Why you're doing it.
All systems need 3 base components 2:1. Water H2O 2:1 electricity current resistance voltage. All this knowledge yet you still can't see how it relates to you?
If Sungazer were playing on a space ship, an interstellar Titanic, who instead of crashing into an iceberg collided with a black hole: as time itself slowed to a stop, would you steadily increase your tempo to preserve the audience's experience or maintain tempo and allow the song to slow to a stop, drawing attention to the strange environmental physics?
I think perhaps the ultimate example of western notation just not being able to represent something is SWING! Of course, you can try triplets, dotted eighth notes or other tuplets, but the best solution is just to notate the music as if it were straight, and just tell the performer “swing!”. In other words, if you micromanage the swing, the swing dies.
My way of notating that is to denote the entire piece as 12/8, which is convenient because sheet music software will play it correctly (sometimes the recipe isn't for a Human)
I sing with choirs professionally and you get the same thing when more traditional white choirs try and do gospel or spirituals. They read it instead of feeling it and just doesn’t grove right. Sometimes they even screw up otherwise really easy syncopated sections because they are reading the notes, but they should just be feeling the rhythm.
It's even better to write it straight and tell the musicians to swing than trying to accurately describe it in notation. That way they can use their ears to find the ration that better suits the piece and even vary it across its duration. Also, I'm not from a black community or even American, but I'd consider the second option to be disrespectful and ignorant about the nature of swing itself, it's a result of the embodiment of rhythm and movement derived from African traditions and not suitable for a fundamentally European form of notation.
The amount of swing is always improvised depending on the player/group. Same way tempo is a suggestion and the conductor will follow their own beat note to note.
@@cy-bernet-ix Maybe. Or you could just ignore it because "cool whatever". On the very off chance that I actually can feel that decimal point (almost never) then I'll probably prefer how I felt it to whatever changes I make. And in the reality that I probably didn't feel it, I won't be bothered to change it in my DAW because I won't notice a difference either way, so why do the extra step of manually changing the tempo? I already got it in like, two seconds, I'm not gonna add two more to erase then type in more numbers. At least, that's my thought process if I'm recording something random and do the same thing. I'll often round up or down, but I imagine OP's thought process is the same as mine on the days where I don't even do that. Nobody will notice the difference anyway unless they're trying to record a cover at the exact same tempo or something. I could be wrong, but I vaguely remember hearing that Andrew Huang does something similar instead of deciding "this must be a whole number because Reasons". In all other circumstances I definitely avoid the decimal point BPMs, though I had no thought process for that before this video beyond a vague sense of wrongness or whatever.
@@RobertKnutzen so this is something i was actually thinking about in general during the vid. one reason you might make micro corrections is either youre "repitching" samples and trying to pitch / tempo match, or because the performer could not physically play the riff / melody etc at that tempo, IE shredder guitarists doing extremely fast sweeps etc. as a composer i totally agree with just rounding up or down to a whole digit, but tons of people definitely just hit the tap tempo and move on.
@@Allen-rv9jz I think this is in large part due to the structure of the videos themselves. Adam scripts these out, so he can read large chunks at a time, whereas Charles speak more off the cuff (especially in his reaction videos of course) so they have to edit it down to make it a usable video
Weird fact: in rhythm heaven fever the music for the minigame air rally is played at 162 bpm, but when it returned in rhythm heaven megamix, it was sped up to 162.01 bpm
I DJ electronic music and I always told people and beginners, that you can easily change the tempo of a track by about 3 BPM without the crowd noticing and even 3-5 is ok. I never had any sientific base for that, it was just feel. But it makes total sense
You do need to be careful with sudden shifts, though. Especially with older tempo changing methods that would adjust the pitch as well. A semitone is a 6% change in note frequency (aka pitch), and so a sudden change of 1% is noticable, as that 1% is about twice the Just Noticable Difference in pitch (generally quoted as being betwen 6 and 10 cents).
I absolutely feel 5bpm - that's the number I go up when I'm practicing technique. 5bpm is a big jump when you're at the top of your level. But even without trying to play it, you can easily hear a 5bpm increase just listening to the metronome.
@@jada90 for musicians for sure I agree. Even 2-3 is noticeable if it's something you've spent time playing. But for the general audience that can't clap on 2 and 4, I'd argue you could push it abt 6 bpm lmao
changing tempo in dj sets can actually be pretty cool if done right. it feels weird if you do a small amount (like 3 bpm) slowly, because the groove is lost. but if you cut the previous track entirely and then play the next track +/- 3 bpm, it sounds super dope. also it can be a good way to transition into slower/faster stuff when you dont have enough room on the pitch slider to beatmatch it with what youre currently playing
I loved this. As a linguist, I wanted to say "yes, in language, too!" to literally every section of this essay. Phonetic transcription is similarly lossy and sometimes overly detailed without being able to tell you the difference between whether it sounds like Steve Urkel or Darth Vader. The term for overspecificity changing the meaning of something is called "Gricean Maxim violation" in linguistics/pragmatics, and it might be interesting to look into more with respect to music.
Yeah, that's true. The IPA is basically like music notation. Sure, it's possible to annotate how a vowel is lengthened or fronted or aspirated or nasalized, but that'd just be like saying you went to the party at 7:42. There's a reason for the lossiness; that's required of any system to model something.
@@vigilantcosmicpenguin8721 the IPA vowel notation already includes information about height, frontness and roundness though, without any additional diacritis and markings used, /i/ for instance is different from /o/ on the basis of these features, so they are not redundant and it's far from being similar to the 7:42 situation that Adam talks about I'd agree with nasality and aspiration, but only as far as English goes, since there are languages that have minimal pairs where only nasality or aspiration changes (ancient greek for the latter, for example)
"If it *sounds* right, it *is* right" - groundbreaking British recording engineer and producer Joe Meek. I wonder what the *precise* tempos (and how unsteady they are) of many of the most highly regarded songs are? That would be an interesting analysis.
A long while back I did a remix of "Single Ladies" alongside "Liquidator" by the Harry J All Stars, and... that took a lot of tweaking. And there's nothing wrong with either track, but the one that was clearly recorded to a steady clicktrack was a lot easier to work with. This is where decimal tempos come into their own though - when you want to sync two tracks that aren't the same tempo, you need to find their precise bpm and then find a happy medium between them.
@@gabrielv.4358 I made a spreadsheet for myself, of the 1100 most popular tracks in the past 20 years, all with precise BPMs specified. I won't share it cause... hey that's a lot of work xD But if you want I can give you averages or stuff, anything you wanna know?
Absolutely agree with you regarding notation, but within the DAW I’m an ardent decimal points tempo person since I use tap tempo to “feel” where I’d like my recording to be. I’d also point out that if you’re only measuring the just noticeable difference of quarter notes, you overlook that we can perceive differences in durations of longer notes within that tempo :)
To me it seems kind of like comparing 29.97fps to 30fps. Your eye can't tell the difference between the two and it generally doesn't matter which frame rate a video is in, but it becomes extremely important when combining and syncing audiovisual content. It is the same for music, an accurate bpm is sometimes very useful. If the source audio is at 100.25bpm and you add 100bpm audio to it, it gets out of sync and sounds absolutely awful. This is something that I occasionally come across when creating mashups and it is terrible to see people mark down the bpm incorrectly without the decimal point.
@@budswayzeniger3473 You really can't tell that difference by eye. The reason that decimal exists is actually because of the color TV and the way it was implemented on top of black-and-white television.
@@budswayzeniger3473 Both 29.97fps and 30fps are standards and commonly used. Mixing them up can lead to issues, hence it is important to know which one a video clip is using in some technical contexts.
2:58 A slight detail here: we can't actually say that the just noticeable threshold is at 30 bpm based on Shawn's experiment. His experiment asked participants to correctly indicate the pulses were changed. This is finer information than just that the pulse was changed in some (perhaps unidentifiable) way, which is what the just noticeable threshold measures.
I haven’t watched Shawn’s video, so I can state this without bias: That’s technically true if there is no control. But the distinction isn’t meaningful. If people can’t tell 500 ms didn’t shift to 499 ms (the smallest “earlier” at this resolution) when it actually shifted to 530 ms, then the “actual” just noticeable threshold is 31ms.
You are describing something other than the just noticeable threshold, though. We want the pulse frequency where, when altered, 75% of participants can indicate that it has been changed. Shawn's experiment had three responses: late, early, or on time. By misidentifying those who noticed a change but misidentified the type of change, you are introducing a systematic statistical bias and you will misidentify the just noticeable threshold. Now for Shawn's purposes, this is fine because he wasn't looking for the just noticeable threshold, as explained in his video. But this does make the experiment unsuitable for Adam's purposes.
@@tracyh5751 It's easy to imagine that you're hearing a change if you're actively listening for it. I think a fourth response option would be better ("could hear change but not identify late or early"), than to rephrase the question to be just "Did you hear a change or not?", as that would introduce much more uncertainty into the study with only two responses "yes" or "no". Also, by adding a response option, you could measure at what point (average ms) did the fourth response option ("could hear change but not identify late or early") start to get used more and more frequently.
There is also the issue of consistency: How often do we actually play a piece in EXACTLY the same tempo from start to finish? I find that I often have to change the tempo on the playback from say 80 to 79 to 82 in another passage, but I see no point in writing it down that exactly. The thing is that the "right" tempo mgiht also depend on the performers and how they percieve the piece, accentuations, dynamics etc.
I don't think I even play a piece in the exact written tempo. I tend to approximate using the pulses of pieces I do know well that I also know the tempo for, and there are some large jumps in that list.
@@papercraftcynder5430 BPM is also a strange way of measuring, since the difference between each step becomes smaller the faster tempo we are dealing with. In a tempo of 190, a change to 191 is barely noticable if at all, but in a tempo of 22, a change to 23 is huge - it becomes a completely different piece.
@@NidusFormicarum This is actually true for all things sound related. We measure it in frequencies, which tell us how many times it cycles per *second*. So a 20 Hz tone cycles 20 times per second. A 20,000 Hz tone cycles 20,000 times per second. That's our audible range (at birth and during childhood), and humans can be pretty good at hearing a difference between 20 Hz and 21 Hz. They just absolutely cannot tell any notable difference between 10,000 and 10,001 Hz though. So we devolved scales that compress these into formats that match our perception better. The exact same thing is done with the amplitude (or "volume") of a sound. The dB scale gets wild, because 10dB is barely perceptible...but 100dB is quieter than a car horn...and is has 10 BILLION times more energy behind it. That isn't a typo. So the BPM measurement is...fine.
I wouldn't even know what markings are on a real metronome, I've never seen one in person. I didn't know they 'skipped' numbers til this video. When I make music I just start in a ballpark BPM number and maybe adjust a few up or down from there over the course of making the music. Rigidly adhering to the specific numbers on a centuries old device seems far more pretentious to me.
@@robinlydian4452 Personally, I could understand decimal points drawing remarks or possibly odd numbers such as 91, but being called out for something like 90 to me as a failure to accept and adapt to modern sensibility. Electronic music is often written in multiples of 5 (not always obviously) and thus I would argue is common practice, just not traditional practice
Sheet music never lies. I remember getting a sheet from another student (Rachmaninov Prelude), and there was a note made with a pencil: 'You can't play this' Guess what... they were right
@@kurumi394 I've seen other scores that are not possible for people with less than six fingers on a hand, or look like they are reductions of a duet that were rendered sloppily as to what is actually possible.
I think your recipe analogy is right on the money. I'm part of the chef team of a restaurant and we have discussions all the time about how much detail to put into our recipes. It seems intuitively obvious that the more detail you put in, the more consistent and accurate the final dish will be, but if you make the recipe too detailed it gets in the way of people using their experience, judgement, and ultimately taste. The feel of a decimal tempo is kinda the same as the feel of an exact number of grams of salt; I appreciate the effort but I'm mostly going to use this as a rough guide and adjust to make it the best I possibly can.
I hope to some day own a restaurant and am studying at a cooking school atm, I agree that too many details get in the way of a personal touch and can even make it a bit bland
Yes exactly! I see the decimal notation things as goofy recipes that are intentionally outrageous. Like Allegra chicken. No one wants to make that, especially not via the recipe. But it's really entertaining to make jokes about it and write unusable recipes for it.
@@amora1246 This is going to sound really glib and shallow, but honestly just find a good chef and work for them. School is useful for teaching the general principles but nothing can really show you how those principles apply in the real world besides doing it in the real world under a chef that wants to see you grow. When it's time to move on, find another chef to learn from and work for them. And the whole time make sure you're making it worth their while to teach you. Best of luck!
I recently listened to a lecture given by Brian Eno, where he spoke of his collaboration with Italian tenor, Pavoratti. The first time that Pavoratti entered a studio with Brian Eno, Eno and the musicians present started to attempt to throw something together, to write something new using the studio techniques and equipment they had available. Each musician probably arrived with a few ideas that they had developed on their own, and the group started to try and see how they could get these ideas to gel together. Eno said he looked at Pavoratti, who had a look of surprise and incredulity on his face. He said something to Eno like “You are making this music right now?” He had never in his life experienced anything like that. A recording session to him was where you got together with a symphony orchestra, broke out some old warhorse, e.g. Bizet’s Carmen, and started to rehearse and record. He just never realized that it was possible to make music on the fly. And he was amazed, and Eno and Pavoratti began to have deep discussions on the topic.
@@psykeeq … Pavo is turkey in Spanish, and Pavorotti was a turkey whenever he attempted to lipsync a concert instead of rendering a live performance like the people were paying for…
one piece of “transcription notation” that i love is just writing “laid back” or “rushed” instead of accurate notation as it allows for you to worry less about accuracy in notation and more about accuracy in performance
Same thing applies as when Adam speaks about BPM. Including the exact rhythm of rushed or pulled back notates is useless because it’s a pain in the ass to read and it’s not how the performer was hearing it. Whilst notating the exact rhythm is more ‘accurate’ - its useless and helps nobody
That's just "rit..." or "accel..." (if you mean gradual change) or "meno mosso" or "piu mosso" (if you mean sudden change), but using English because you want to look cool.
In fighting games it's common to have combos that require one or multiple 1 frame links, meaning you have a 1/60th of a second time window to hit the combo. Always found the rhythms of combos interesting especially when played on an arcade stick
They can make an interesting rhythm, definitely. You can literally hear if somebody's combo will come out, based on the noise their button presses and stick make.
a few years ago when I first getting into music production, I made this melody that had a super wack temp like 135.27 or something like that. it happened after I accidentally messed with something and I was to inexperienced to know the difference. A year later I revist the project because I like the melody. Noticing the wack tempo, I go to fix it, and for some reason the synth sound I was using on the melody sounded super different and not nearly as good when I fixed the tempo. So somehow I have this melody that only sounds good at a wack tempo lol
It either 1. Is a mistake somehow or 2. You just got so used to it that you don’t like how it sounds. Once you listen to a song enough it can be impossible to be objective about it
@@melody3741Yes, exactly. I listen to my own stuff in the car, cuz I often get ideas for revisions or new songs there, and take it from me, if you want to listen to your own stuff regularly, you HAVE to be good at fighting this feeling. It's possible, but not easy sometimes lol
@8:35 "Man, this food analogy is going too far" - No, thank you for it. By making that analogy, you helped me link aspects of music (which I'm bad at) to cooking (which I'm good at). The examples you gave demonstrated playing by ear vs by practice vs by sight as ways we think about many things in life, and I appreciate the depth you put into the analogy :).
@13:49 "Because it's inherently a lossy format" - Now you're speaking my kind of nerd-speak! Musical notation is like JPEG - our minds fill in the missing bits :)
@@Tawnos_ I'm surprised your mind jumped to JPEG instead of MP3 (or any other lossy audio compression format), but I had the same basic reaction. I was like "AHA! Now I understand!". 😆😆
Yes, in fact he kind of didn't go too far enough - there's a salient point about good cooking and good music that the recipe analogy highlights - one major advantage of the lack of precision is that it allows for interpretation - such as adjusting ingredient ratios or cooking temperature to get a unique taste/texture, or micro-variations in tempo or volume to highlight the feel of the music. Precision recipes/sheet music are great for mass production, but stifle the creative performance aspect.
Adam's channel, at this point, is an extended series convincing people that music is generally overanalyzed in all the wrong ways and underanalyzed in all the actual interesting ways.
Strange example of decimal BPM: Hurricane by Kanye West is 160 BPM on the dot, but its earlier leaked versions are all 157.61 BPM, since they all use the older Yandhi beat. Since Moon was initially created as an outro to those earlier versions, it, too, is 157.61 BPM, even though Hurricane isn't anymore.
The idea is that, although it can be played in decimal BPMs, when composing it's almost never that specifc, as that's pointless. Supposedly the only exception is for composing for films when audio/video synchonization is needed, but I personally doubt that justifies using decimal BPM either, as it's also possible to just mark down a few key notes that are required to be synced with the video, and everything else can shift a bit without getting noticed.
@@FTZPLTC Film scores are often recorded live I think, aka performing in real time along with the movie being played and the conductor can control the pace to meet those key points (or at least that's how I would imagine things to go). Alternatively it's also possible to stretch the audio in post to meet those key points. And for live mixing, yes, exact BPM is important. But just relying on the BPM reading is not enough, you got to have a device to send the metronome across all devices, and then turn on BPM syncing for it to work, otherwise everything would still drift away from each other. So in summary, although decimal BPM technically exists in recorded or played back tracks, they should be perceived as whole number BPM with a margin of error as well.
@@FlameRat_YehLon You're right that scores are sometimes recorded with the movie playing. I don't think it's super-common for music to be played to sync up precisely with the action though. Sometimes it's done, but if anything I'd say the metre of the music is more likely to be used to set the pace of the edit. It would be unusual for a film editor to just happen to making edits that formed a coherent pulse for the music, after all. I'm rambling though, I find this subject really interesting. And, again, watch Cats to see this done fantastically badly and arse-about-face.
@@FTZPLTC Thus I said key points. Movies can be designed to have a few key points that are synced to specific notes on a piece of music, and every other notes can be played in a more flexible way, just to fill in the duration between key points.
I've heard it said of scientific models that all models are wrong, but some models are useful. Maybe something similar applies here. All notation is wrong. Some notation is useful. And we can even take the parallel even further. Each scientific model has a certain regime under which it useful and outside of which it's not. Some notation is useful in one context, but not in others.
"I've heard it said of scientific models that all models are wrong, but some models are useful." That's a quote often attributed to statistician George Box (though it predates him as a concept)! The more you know 🌈
@@danparish1344could apply to modeling in general though because all models are constructed based on summerising as many variables as can be accounted for in order to abstract, reflect or predict the shape or pattern of something. For intents and purpose, if we consider enough variables, we have a functionally usable model. But in order to stay consistent, it must be open to unforseen variables and adapt to new information. No model can be absolute; therefore all models are still "wrong", no matter how functional and useful they are at the time.
Regarding wrong models, that's probably true. Take for example all physics calculations: they all ignore extreme variables, like the inner flexibility of an object. There's a video about how long does it take for one end of a metallic bar to move after you push the other end. In plain physics it would be 0.
Talking about tempo and perception, it might be fun to look into the music genre "Extratone", where the BPM is so fast it becomes an, you guessed it, extra tone
The issue here is known in statistics as "overprecision", or more fittingly, "drama digits". It is the misguiding use of precision that simply makes no sense in a specific context.
I haven't heard "drama digits" before, but I may have to find an excuse to use it now. (I do bioinformatics things, so tools giving incredibly fine grained results when you put in a handful of vague data points is basically standard.)
@@dnebdal haha, indeed, I like that word, too. I can well imagine that in your field you will have amble opportunity to tell people to drop the “drama digits”. Maybe at least sometimes opt for calling it something more scientific-sounding, like “spurious accuracy“, because … well, just because. ;-)
The food recipe analogy was so perfect, I absolutely love it. I think it’s so true that the notation is merely a recipe for us musicians to make the musical story come to life, and it is incomplete as well. And I completely agree that there are times when notation is not at all helpful, and it would be used as analysis or even something more trivial. However, I find myself so grateful for music notation and desiring to keep using it all the time, I love it so much. I don’t feel the need to rely on notation, but just the sight of a five-line staff, clefs, key signatures, time signatures, and notes on that staff bring me such a joy that I cannot describe. I love music notation, I just love it.
another time to use decimal points in bpms is when you have a tempo change that's fractional to the base tempo! like in a song i'm writing where quintuplets in 4/4 become 16th notes in 5/4. it goes from 94.4 bpm to a clean 118
That's technically true, but most of the time it's better to just notate what change happened instead of putting the exact tempos. Like, start the song in 95 bpm and at the transition put something that says quintuplet = 16th note. That way, it's obvious what you actually intend to happen instead of having to check the new tempo on an actual metronome.
The beginning could just be written as 20/16 time signature w 118 bpm no? The main pulse felt would change from every 5 16ths to every 4, but AAL do exact same thing in their song The Brain Dance. A lot easier to see the correlation between the 2 that way imo. I understand recording with a click would be trickier tho, could also use some midi instrument to make your own click emphasizing every 5 16ths.
a similar issue i had was trying to make a quintuplet swing groove 'last' for two more notes, (essentially as though it were 5/16 with occasional bars of 7/16, but i'd already established that it was quintuplets so didn't want to change the whole system). I found the fractional bpm thing way too technical for how easy it was to feel.
@@goatmagic5220 You could do that, but it would be a bit harder to read than quintuplets in 4/4. I think just writing it as metric modulation is nice, but either way works as long as you don't try to make someone read 94.4.
I disagree. What you're describing is a recreation. They create perfect 1:1 reproductions of terrain for the purpose of forensics, or military training. They're less common but sometimes you really need to re-create a sense of scale.
@@nathanjasper512 No you've missed the point as another user has pointed out. The point of sheet music is not to recreate, only to provide a useful abtraction of the music - much like how a map is an abtraction of an area; in the reduction of details comes the utility. Sure, you can have a 1:1 reconstruction but that's is not what we are discussing today. You're disagreeing with the paticulars of Three Coloured Squares' analogy but not his point
The point of sheet music is to be able to hear your ideas from the playback function of the notation software before you go on to orally explain them to other musically illiterate musicians
3:43 Adam is explaining the linguistic theory of Relevance. It's sometimes known as Grice's Maxim of Quantity. Always pleased when Adam sneaks some linguistics into music theory videos
As an admirer of the existential philosophy, I would say, this thought he is sharing is actually saving music from mechanical objectivity and justifying the randomness of human actions, like a slightly delayed note that brings all the beauty.
Yes! I'm actually now more inclined to produce music with decimal BPMs because (a) it's not noticable and (b) it will absolutely be felt. Taping a BPM in a DAW always made me dread the decimals but now I see it in a new light, I'm keeping the decimals in.
Having worked in music scoring for film and TV for 15 years, I can say you don’t always need fractional tempo to match musical beats to picture, but sometimes you do. Especially for cartoons with a lot of action or comedic beats, or to hit picture cut, etc. And in those instances decimals are sometimes required. Without them the piece is simply incorrect and does not contain enough information to achieve its core purpose which is to match picture. But if you want to adapt that music for public performance, by all means strip off the decimal points and round it off because of course no one will perceive the difference.
I have to call you out for this. Especially with cartoons, they're originally close to 11 frames per second, so I can't see why you would need to have hundredth of a second precision with sound. Even when you have 1 minute long scene with one score using 120bpm vs 120.5bpm you have roughly .2 second drift. In one minute you have skipped literally one frame. Half a BPM I can somehow understand in very long scenes with very long scores, but going any more precise than that is just not very useful. Also, many composers just play watching the scene organically, like it was used to do before computers. People didn't use fractions of a BPM then and I hardly believe many does so today.
Also let's not forget physics. 30ms is the time that sound waves in air need to travel roughly ten meters. 5ms is the delay of a sound played 1.7m away. Have you ever stood 1.7m away from your fellow musician (or your amplifier box) and felt the delay, or that you couldn't get the beat right? I haven't.
I freaking love your videos. The preparation, the info, the delivery, the editing, the everything…it’s all so enjoyable and interesting every single time.
As a musician, I was glad, that you revealed, that the fifth note was slightly late. I "felt" it. It is like the ear trainings for pitch. I can tell if the next note played is 0,7 hz higher or lower. But I don't hear it anymore. It is just a feeling. But in the end, I am right. I love these challenges!
I think it's always connected to what you're doing and how often you do it. I didn't notice the change in tempo... but by zeus I'll see small dips in framerate where my buddies say "you're imagining things, looks fine to me"
I've been playing rhythm games for nearly 15 years now. When the difference between a perfect score and a "choke" can be as little as 10 milliseconds, the late note was heard loud and clear.
This is the sort of stuff I addressed in my Grad Dip thesis, "The Communication of Type-Setting". I came up with: "The purpose of typeset music is to let the reader recreate the composer’s intent under the worst performance conditions. Printed music is a set of instructions for the performer, not the composer. The printed document can be used in two distinctly different ways: either for slow time-relaxed consumption (such as for practice, memorisation or analysis), or for time-critical events, such as sight-reading or live performance. The music has to be able to meet both uses." But. Maybe there's another purpose for written music - to record aural objects. But I think a sound recording would be better. Thanks for the thought-provoking discussion. -------- I was at a rehearsal this week and the drummer wondered why ♩=69 rather than ♩=70. We all agreed that while 69 was "nice", 70 would probably be simpler to comprehend. I thought it was either to do with gears in mechanical metronomes (numbers of teeth on cogs, etc) or to make simple ratios for changes between 2-based time and 3-based times, such as 3/4 to 6/8, or hemiolas. Now I know it's the minimum perceptual distance from the previous marking. I presume the datum is ♩=60, and the rest were derived - probably empirically - from there. -------- I sometimes add stuff to amuse future-me. In my thesis, in the section on timbre: "Timbre also seems to miss out in the typesetting world. This is paradoxically because there’s too little to say, or too much to say. It would be extremely difficult to get a clarinet quartet where: · One clarinet should sound like the second flautist of the Berlin Radio Orchestra in 1963. · One clarinet should sound as though it’s using a split number 1½ reed where the player is used to playing on a number 4. · One clarinet should sound like it was invented on Neptune, at Easter. · One clarinet should sound like it’s being played by a trombone player, who had to quickly borrow any instrument that was available, because his own instrument was stolen by a gang of fake portrait artists on a tram in Rio di Janeiro. It would be impossible to get a second quartet to play the same music. Although rather extreme, the example shows that timbre cannot be made too specific. Rather, “four clarinets” has to suffice."
One would have to approximate with terms like "sweet," "cronchy," and so forth. There are even some standard Italian musical terms for this. If you really want to capture a timbre, then find a way to record a sample of it.
@@SreenikethanI Depends on your definition of the word "hear". They could both perceive vibrations well enough to distinguish variations in both pitch and rhythm. Does that _have_ to be done using the ears to count as hearing?
@@EmpiricalPragmatist ah gotcha, I had the same thought in mind, just that I wasn't sure to consider that as "hearing"... maybe "perceive" is a better word here?
I love this video and I agree wholeheartedly with it... apart from one little caveat: When creating a video animation that moves to the beat, I compose a piece of music (or time-stretch an existing one) to match a beat to a whole number of frames. That means, at 25fps, if I want a beat every 12 frames, I need 125bpm, or for every 15 frames it's 100bpm. But if I want a beat every 14 frames, it's 107.142857bpm to stay in time. 16 is 93.75, 18 is 83.333333 (or 166.666666). 🎶
I feel that was mentioned (as one of the quoted reasons that wasn't further engaged with). For such purposes though, wouldn't it be more informative to write 25/14*60 bpm (or even 25/14 bps?), rather than 107.143?
I also would be that specific if I needed someone to record their part separately. However, I'd also be including a guide track, and the tempo is just in case they want to DAW it up themselves before they submit it.
The first thing that came to my mind when I saw the decimal points was synchronisation of different tracks. I quickly tested a 120 bpm with a 120.5 bpm and after about 5 seconds (or 10 beats) the asychronicity became noticeable to me, a few seconds later it was quite obvious. This kind of corresponds to visual perception, where fine differences in structures might be unnoticeable, but overlaying these structures makes the differences become easily visible (think of a Vernier scale or Moiré pattern). An interesting and relaxed video as always. :)
I thought of such music structures which go in and out of sync periodically. However it's probably more convenient to use fractions rather than decimals so you could tell the pattern by looking at them (even if you're a mere human). Sadly DAWs don't usually support fraction tempo and we have to stick to decimals. And decimal values cannot be precisely represented using binary values. Game over :-D
I guess that's where the just noticeable difference comes in. If you put the two against each other, sure, you can notice it, but seems like no one would ever really play anything like that.
@@vigilantcosmicpenguin8721 A friend who was DJ for electronic music for a few years explained to me that the problem can arise during a transition between two tracks. With a fading transition the two beats inevitably overlap for a few seconds and thus the bpm must match perfectly (at least good enough so the asynchronicity would become noticeable just after the transition).
So there is something off... On from minute 4:06 to 4:19, Adams says that he arrived at the Party at 6:42, but would say that he arrived at 6:45 to avoid seeming suspicious. But from minute 15:01 to 15:09, he refers to arriving at the party at 7:42 or 7:45... And this sounds suspcious, like a massive deception in need for an alibi!
Another justification for a decimal BPM for a performer could be when that performer isn't human. Currently a machine can't feel a tune so precise notation is required for the "performer" to create the music as intended. Otherwise it is a BPM of how things ended up lining up because a machine was running at such a speed
The way you described the inherent impossibility of perfectly complete and unbiased sheet music reminded me of something I've heard once about maps. It is impossible to fit all the detail of a perfect aerial view onto a map, and if there was a marker for every detail the map would literally be nothing but markers. So, it's up to the artist to decide what is important, what can be simplified, what details can be left out, and what should be marked. Thus, there are an infinite number of ways to make a map of the exact same plot of land, no matter how big or small, and each map is inextricably tied to the mind of the person that made it.
This kind of reminds me of this little "amazing facts" book I had in my childhood. It listed the height of Mt Everest down to the cm well before we had the technology to do stuff like that. Eventually I figured out they had simply taken the height in whole feet from the original edition and done a conversion that resulted in that false precision.
If I remember it right, the original measured height of Mt. Everest was exactly 29,000 feet (to the nearest foot), but it was then declared 29,002 feet to prevent people from thinking "29,000 feet was just a rounded estimate".
Decimal point tempos is actually a really big problem for people who do mashups because for a little while it seems normal and then it becomes a mess…… #JusticeForMashupMakers Also i love the music that comes before you talk about a topic, i want a full version please!
Was once told a story by a percussionist who was playing a piece written mainly on a computer. The percussion part was all over the place, no identifiable pulse, etc. So he asked the composer how it should go. Composer couldn't tell him. He'd essentially just clicked his mouse randomly into the software. So the next question was "would you notice if I got it wrong?" and the answer was "no". precise notation doesn't always help convey what you want.
I'm a classical composer, which means I almost always deal with live players and conductors and almost never with computers. I always remember a thing one performer said to a classmate during a composition masterclass: people aren't metronomes. Unless you have a clicktrack in their ears with the precise 133 BPM you want, no performer is gonna perfectly adhere to that. In fact, since in most professional settings performers have very little time prepping for a piece, metronome markings will be practically pointless unless you specifically bring in a metronome. Because of that, I always use words to convey tempo (adagio, presto etc.) and even if I add a tempo marking, it's always in brackets, always marked as approximate, and always rounded to the nearest multiple of 10 or 4.
@@gyroninjamodder first of all, in chamber music it's rare to have a conductor second, even if you have a conductor, they don't go on stage with a metronome or a click track, and so an exact metronome mark wouldn't be of use once again. I sometimes put in an approximate metronome marking in case there's a soloist or a conductor who studies the piece in advance and can practice it with a metronome, but even then there's really no use for a mark like "131", as their actual performance will always be somewhere around that and not the exact value. so I write "~130" or "~84" etc, because the exact value really doesn't matter all that much.
Not at the level of precision of 132 vs. 130 bpm, but some conductors, such as Toscanini or more recently Benjamin Zander, made powerful arguments that Beethoven meant for performers to take his metronome notations seriously, and that the fashion for playing the symphonies at much slower tempi made them lugubrious and robbed them of the energy Beethoven intended for them to have. The point isn't to hit a specific tempo as though it were a click track, but that the metronome notation conveys very important information about how the piece should feel when it's played. th-cam.com/video/L_bd1J98jNQ/w-d-xo.html
@@jonathangilligan6019 hard agree. I actually believe there's also pieces that didn't have a metronome marking initially that are misplayed, e.g. Moonlight Sonata mv. 1 (there's no way that when he wrote "adagio sostenuto" he meant the almost "grave" that people play nowadays)
04:30 you might want to visit Switzerland. Here every minute is important when telling the time. So if you ask someone what time is it hey mostly reply with the exact minute. It also depends wich part in Switzerland. People living in the mountains usually round up time too because it is easier and they don’t need to do everything by the minute.
@@Anonymous-df8it because time is money. If the train is a few minuted late then chances are high you won’t catch the next one. Here trains are very punctual and that’s why we have only small window to go from one track to another and it works very well, like a Clockwork.
@@Nanomaroni 😂 Okay. No. Switzerland is nowhere near Japan in it's punctuality and nowhere on the same planet as America in it's economic significance. You don't NEED to tell the exact minute. It's far more likely a quirk of the language.
Polyrhythms get the clicks - best description of “the algorithm” ever. Great video (from a DJ who started without software and decimal tempo displays).
When it comes to rhythm games, typically the size of the timing window to hit each note at the highest accuracy of judgement (e.g. hitting 'Perfects') is around 30 ms. The timing window for Dance Dance Revolution, however, is (crazily enough) half that at 15 ms! What is crazier still is that there are players out there who strive to clear songs hitting every note precisely to within that 15 ms window: the coveted "Marvelous Full Combo". In fact, top DDR player iamchris4life in particular is well-known within the rhythm dance game community for his quest to get Marvelous Full Combos on as many songs as he can. This video does bring this premise of hitting notes super-precisely into another perspective for me, since having predetermined rhythmic challenges for players to attempt to exactly recreate while getting scored on their accuracy to such an insane degree feels almost robotic and absolute. Even swung notes are precisely laid out for the player to try to hit exactly, laid out in a way not too dissimilar to the precise transcriptions from these music social media posts.
I love the inherent contradiction in that. Can you groove/swing *exactly*. Alright I suppose that’s also grooving/swinging *like someone else*, which is both what makes sounding exactly like other musicians hard, but is all the more impressive when someone can do that for multiple performers.
recommend replacing "however, is crazily enough" with "crazily enough, is". i understand "however" is a transition word which is used to express contrast, but "crazily enough" also serves more or less the same purpose. as it is now, it'd be drowning in commas as, like, "...Revolution, however, is, crazily enough, half that at 15 ms!". might be salvageable with parenthesis instead like "...Revolution, however, is (crazily enough) half that at 15 ms!" but to me it feels unnecessary for something to be however and also crazily enough. anyway i agree with you
@@bitodd Tracks written for rhythm games seldom stray from EDM precisely because of that. Staying on beat is the whole game, and if, say, there's a song with a 6-6-7 hypertuplets, that would be the focus of the chart. Live music usually cause the "in-game BPM" to swing meaninglessly in a way that's frustrating to play. (For example, the song ってゐ！ in SDVX is marked 314-340 with 12 bpm changes that doesn't even feel like it's changing but might mess up the scroll speed.) If there's one thing that seasoned rhythm gamers are definitely better than musicians with no rhythm game experience at, it would be hitting that swing within ±16ms every single time.
As a battle DJ, editing battle samples for performance is very useful, and one thing you can do is repeat a sample over multiple rotations of the record to get "skipless samples", so if your needle jumps, you land on the same sample and not get a skip. Turntables spin at 33 1/3 RPM, so you can multiply this by integers to get a tempo at which you can place that number of beats into a revolution. The problem comes with the 1/3, or .333..., which makes fitting 3 samples at 99.99 BPM drift right as you go further into the record, and doing it at 100 BPM makes them drift left. It is a nightmare.
You made a great comment here because you were describing not just a theoretical need for decimal tempos, but a material, practical, demonstrable need for them, and not even just to the second place, but to many beyond that!
@@duon44 it really is a cool side of music manipulation. I recommend looking into the likes of Mix Master Mike, DJ Craze, the X-Ecutioners, or the many DMC and Threestyle champs for a good dive into it.
14:00 and the recipe analogy are things I really needed to hear lol I’m someone who both because of my brain and because I so thoroughly learned percussion in band, struggle with “properly” learning music and wanting to play/make music because I’m so focused on playing notes “on time” at the precisely written time above all else. The few times I’ve messed with music (beyond percussion) so far, I did best when I just kinda felt it and played by ear, not worried about perfect timing and not (as) worried about all correct notes The recipe analogy really made it click cause I’m someone who really likes cooking and constantly switches between following recipes to a T and throwing a bunch of my own stuff in there to see if that’ll make it better
A point that is implied is readability. Since the score is usually given to a performer (recipe analogy), a necessary tradeoff between readability and accuracy goes into a well made score.
As a music producer that makes their tunes by feel first, I have my tempos at decimals because I just tap tempo it in and leave it as is. Even though I could just correct the tempo to the nearest integer and it wouldn't be intelligible, it's left as it is because it still somehow feels different / "wrong" to what I envisioned for it to feel like.
9:23 One could say, then, that the notation for the audience is not the recipe, but rather the table of contents, which is often calculated. It is meaningless the the performer when the performer improvises or when music is made like the music teacher example, but gives the audience a "fun" view into the mechanics of the music (or the food)
As someone who’s not that savvy with daws but likes making mashups now and again, I cannot stress how frustrating decimal bpms are. I speed up song A slightly so it syncs with song B. All is good for 3 bars, then I hear it. The de-sync. It only grows from there. Okay, I’ll go back and adjust the speed again. 1bpm faster. Nope, it’s definitely off best this time. I can’t adjust it that finely, so I just have to scrap it since I have such a sensitive ear. Sometimes I wish it wasn’t so I could just ignore the de-sync but alas. Ignorance is bliss.
you understand😭😭😭 I’ve had to struggle syncing drum breaks and hearing them desync a few seconds into the song too many times, just to realise it’s because of those decimal bpms just now
On the rhythmic incompleteness of your bass solo transcription: That sort of "technically inaccurate" notation is fine if the performer can be reasonably expected to be familiar with stylistic conventions (or have access to an authoritative recording to supplement the notation). But part of composition is to innovate new styles (which wouldn't have established conventions) or incorporate influences from more obscure styles (which a performer should not be assumed to be familiar with). I imagine there are reasonable stylistic conventions which could simplify the notation of the opening of The Rite of Spring, for instance, but it's easier in a case like that to precisely notate than to communicate those conventions. If aimed at a performer who doesn't know the appropriate liberties to take, being explicit about landing on "e of 1" might have been more appropriate in your transcription. Similarly, lead sheets often have "quantized" versions of the melodies, the assumption being musicians familiar enough with the style to realize an accompaniment from a lead sheet will also be familiar enough to breathe life into the melody.
I am not sure if I am agreeing or disagreeing with you, but I would think that one would be better off notating music in the most basic form that fits the music and allowing performers to add their own stylistic feels and ornaments. If taken to other examples, would you even precisely notate a swung rhythm? I would think not and adding a symbol for swing at the top better indicates how to play it, since the feel choice will not always be exact to how it could be notated, and it will also hinder attempts to play the same piece in other styles. If I was playing Goldberg aria, seeing the mordents and trills notated is not helpful because you may not read it as that and instead see it as an exact thing to play, but I may play a trill with some rubato on some performances if it has the trill symbol and not precisely notated. The music would be so busy I would probably not bother to add any of my own ornaments and certainly would never cut down any written ornament. Going by Adam's example of the semi-quaver rest, if I saw that notated I would expect that to indicate some sort of syncopation in the music or purposeful rhythmic choice to not land on one, rather than a laid back note.
Couldn't agree with you more. I think if Adam delayed the E by some very difficult to quantify rhythm (5/11ths of a beat for example), then I might understand not noting the rest. In this case however, he delayed it very clearly by a 16th, and so his explanation for why he didn't notate it is kind of a head scratcher. What's worse is that he has a similar pickup-into-long-note idea at the end of the solo, (see 10:35) but that note doesn't have a delay. Like you said, if you have a recording of the solo then it all makes sense, but I feel like the notation should be able to stand alone and get you as reasonably close to the music as you can get without needing a recording, or being a part of the same social circles as the performer. A 16th note rest is very well within the limits of music notation and is much, much simpler than any of the tik tok examples. Then again, I'm not a working musician. Maybe it makes sense to other people.
I think your point ESSENTIALLY brings up a GREAT point not thoroughly addressed by Adam: there is a third, in-between, reason for transcription: Adam's transcription makes PERFECT sense to people who are used to glancing at lead sheets for music that is often improvised in nature and going "oh yeah, make sure the end of the run lands thereabouts on the chord change." But the "quantized" version is what is literally being played, whether Adam likes it or not. And, especially if learning originally improvised solos, the quantized version should be preferred if someone is learning or studying a solo. Is it busy to look at? Yes, but if I am trying to learn a piece of music (especially since I don't have a great ear) the literal/quantized version is so much more helpful... Awhile ago I was learning Dogs by Pink Floyd, and, if you know what I am talking about, the first guitar solo's tone is engulfed with that deep chorusy, swirly Yamaha rotating speaker colour... To the point that many of the syncopated notes' attacks are distorted or washed away altogether to the point that it is hard to pinpoint what was originally played. It was only until I could find a transcription done by a college student that I was better able to figure out and wrap my head around the rhythm in that solo; up until that point, tabs (which are probably as equally useful as Adam's "c'mon, we know it is actually a LAID BACK half-note" notation) could not illustrate, to me at least, the NATURE of the syncopation. In this way, I would argue that super-accurate transcribing actually can become EDUCATIONAL and useful in teaching a style. For instance, today I learned that playing "laid back" bass is playing ends of quick runs that land on a chord change about a sixteenth or so behind the beat.
@@CHEWYCHEWYQQ I thought the same thing after watching that part! I am of the school and philosophy of using your ear and intuition to make music, but when notation is concerned, specificity is paramount. It's why myself and a lot of others feel that fake books are more or less useless... you're gonna have to listen to the tune anyways, so why even bother carrying around a pile of shoddy transcriptions?
The problem I see with that is you only really need that specificity for live performance of scores, which is probably not how most scores are recorded now. If you need stuff to sync exactly to an onscreen cue, you probably just slide the notes or samples in your DAW until they line up with the footage. Plus we're already assuming that a film doesn't really follow normal music tempos, so how can you be sure your decimal point BPM will work for longer than one 3-5 second shot? At a certain point, you just have to learn the timing of the visuals themselves and not consider metronomic tempo - films, unlike scores, are typically written in Free Time.
Decimal tempo markings can be useful for Elliott-Carter style metrics modulations. If the old 5:3 eighth note is the new quarter note then decimals may be useful or necessary to be both musical and accurate.
I was faced with the dilemma you mentioned here when I was writing thousands of lead sheets for music publishers in the 1970's, mostly for copyright purposes. We were forced to fit contemporary musical performances into notation that was more suited to 1940's Tin Pan Alley, not to mention 18th Century convention. I often had to decide if I should notate the "laid back" feeling accurately or how it might best fit on paper, simplified. To do it accurately might mean an unnecessarily complex appearance but to write it in a simpler fashion meant I was second-guessing the composer's intention. 'Tis a puzzlement!
Score follower's work, to me, fits neatly into modern "effortposting", a phenomenon wherein people deliberately overdo the effort and/or precision of their work, as a joke, as a commentary on status and precision, or for any number of other reasons. You're right, they're going well beyond the limits of human perception, and that's the point, I don't think most people are confused about that.
Yeah, I'm with you. The fun is in the absurdity of the precision. I think Adam has valuable and interesting things to say regarding value and use of precision or scoring in music to different people, but seemed to never really get into this point regarding these particular kinds of scores. In a way, this video ironically misses the point by focusing on the markings rather than the feel a hyperprecise score or maybe BPM might aim for. He's become the very thing he vowed to destroy 😔
"The notation becomes another way for the audience to engage with the sound of music". Never truer words were said on music-related channels, based on my personal anecdata at least :) I am musically inept, can't play any instrument, can't catch rhythm, can't sing or dance. But boy I love to see visual cues when it comes to listening. I think it started with ProTracker visualisations back in Amiga days in early nineties, then of course Milkdrop and now all the Synthesia-like videos
I love the "7:42/7:45" analogy here. Reminds me of the Gricean maxim of quantity, as in his example: "If you are assisting me to mend a car, I expect your contribution to be neither more nor less than is required. If, for example, at a particular stage I need four screws, I expect you to hand me four, rather than two or six.". "Rounding up" seems to be human nature, in all aspects.
i like your perception of sheet music :) i think of sheet music as a story book, and me playing it is reading aloud. Eventually, i may no longer need the book to portray the story verbally anymore, and i can tell someone else the story and they can tell it after hearing me
So here's a very specific use case: Radiohead's Kid A has the peculiar quality that every song is (in my opinion) enhanced by playing back a second copy at a delay of exactly 17 seconds -- down to the hundredths of a second. The whole album is in canon with itself. But, each song is in a different time signature, different phrase lengths, different melodic contours, different tempo. So the way that it's managed is that every song syncs to a certain number of beats displaced. For Idioteque, with its 5-bar 4/4 phrases (20 beats per cycle), it repeats at exactly 39 beats --> which works out to exactly 137.647 bpm. For Kid A, 17 seconds is 32 beats at 112. 941 bpm; for the National Anthem, its 26 beats at 91.647 bpm (I hear it in double-time though - 54 beats at 183.294 bpm); and for Everything in its Right Place, it's 35 beats at 112.941 bpm. Interestingly, Kid A is the only one that lines up the 17 second delay with the cyclical phrase length (8 bars of 4/4). The National Anthem repeats halfway through the 13th bar (4-bar phrases in 4/4), meaning the off-and-on syncopation in the baseline just becomes straight 16ths, with beats 1 and 3 cycling between: C-F# > F#-A > F[nat]-A > C-F[nat] over a D pedal. Plus, the sax parts form some crazy suspended harmonies at the end. Way too much to talk through in a youtube comment, but it all lines up in a really interesting way. Bonus points in National Anthem for the lyrics -- the "it's holding on" at the end of the first verse ends up filling in a gap from earlier. "It's holding on" happens to sound a lot like "is all alone", so when both are layered on top of each other you get "Everyone [is all alone.] Everyone around here. Everyone is so clear." For Idioteque, the percussion is where you hear it most clearly, though Thom Yorke does end up harmonizing with himself at several points. The bass drum starts on a three bar cycle, split into 6+6 beats, but when the chords come in, it shifts to a 5-bar phrase, where the first 6 beats are copied over from the 3-bar version with silence filling out the rest. The snare hits on a consistent beats 2 and 4, but when the canon enters, it now hits on all four beats of every bar. It avoids feeling too chaotic though, because that alternation between a steady bass drum groove and the 6-beats + silence sections make it so that it's never overwhelming -- it just has a little extra flavor every once in a while because the bass drum pattern is so irregularly syncopated. The chords on top have such a slow attack that you don't even notice that one is coming in half a beat early, and they've got just the tiniest hint of syncopation already, so if you're just listening to the chords, you won't be able to really follow where the actual downbeat is anyways. And for Everything in its Right Place, the pattern is just a 2-bar repetition of 10/4 (at least, that's how I hear it), so a 35 beat delay puts you in half a bar early. This one also has some weird syncopation that creates interesting emergent patterns, but for me, the most evocative aspect of it -- totally unexpected, but obvious in retrospect -- is the way the vocal melody interweaves with itself. On its own, the structure for each bar is: Lyric > Long silence, gradually shifting into lyrics that fill the whole bar. But, because the canon is offset by half a bar, on the repeat you get Lyric A > Lyric B > Lyric A > Lyric B, without those gaps. It becomes a new song entirely, with a melody that, in theory should be the same, but in practice takes on an entirely different character: Album version -- Everything [pause] (repeat 4x) In its riiiiiiiiiiight place [pause] (repeat 4x) Yesterday I woke up sucking a lemon (repeat 4x) Kid 17 version -- Everything [pause] (repeat 4x) Everything in its riiiiiight place (repeat 4x) Yesterday I woke up sucking a lemon [in its riiiiight place] (repeat 4x) --- And all this magic because Radiohead decided to use extraordinarily precise tempos.
@@rottingcorpse6565 it's not unheard of for a track that can loop on itself unintentionally, especially if the whole track is a repeated chord progression. But its exceedingly unlikely to be coincidence for an entire album to do it at exactly a 17-second delay, especially where the melodic lines from each song at that delay interweave seamlessly with each other. Only in a few instances is it Thom Yorke harmonizing with himself -- like in Kid A or parts of Idioteque. The rest of the time, the melodies all have space left in them that maps to that 17 seconds, so that the only things happening are a) direct harmonization, or b) emergent melodic lines from fragments played in canon. There are no instances where the counterpoint creates out of place harmonic or rhythmic dissonance -- which is not easy to do when you have a canon at the unison across all melodic, harmonic, and rhythmic components, where the second instance starts not just partway through the phrase, but on a weak subdivision of the midpoint of the phrase. That doesn't happen by accident. Not for an hour straight.
This is fascinating but it also highlights that an important factor here is just kind of semantical - precise decimals look cringey because of their format, but a lot if these examples are just decimal expressions of fractions and only look cringe because you are forcing everything into base 10. If you just kept the bpm as the original fraction it would look fine
@@jackdavinci They are still misleading. If I told you that one song at 100 bpm played 3 beats in the same amount of time as another song played 4 beats, that doesn't mean the second song was played at 133.333 bpm. It could well have been 132 bpm or something else close by. It's a rounding error. You can look at the audio from a specific performance and calculate exactly the average bpm (assuming a given time signature), but that's just an empirical average for one performance; it's not by design. It's rather like using a high-precision laboratory laser measuring device to measure the length of a track after a race and then declaring it a 99.99983125 meter dash.
As someone who has both played traditional instruments and done some DJ stuff, I think appropriate level of precision definitely does depend on what kind of music you're making and what you're trying to do rhythmically. If I'm playing the piano or guitar, I want to let it flow naturally and let it be slightly imperfect. I don't want it to sound "inorganic." On the other hand, precision is much more important when I'm making a remix with a looping instrumental track, especially when I'm working with electronic genres, hip hop, etc. It's still creative and artistic, but it's more like building mechanical/kinetic art pieces than painting a scene with oil on canvas.
I love Phonon's work, but i felt similar feelings as you did at the way his strange rhythms were given legitimacy by their ability to *technically* be notated in sheet music. Like when I hear Polyriddim I'm taken in by the shifting off-center wubs, but there's lots of electronic music that deals in more generative organic rhythms like that. The only innovation that Phonon brought was sheet-music novelty. (btw adam please listen to Confield you'll love it)
Fun fact about the Just Noticeable Difference: Weber's law states that it's proportional to the existing stimulus. That's why the metronome goes from 69 to 72 (a difference of 3), but then it goes from 138 to 144 (a difference of 6).
only mostly true. you can't treat metronome markings as whole numbers. as stated in the video, 69bpm is one click every ~870ms, 72bpm is every ~833ms, a difference of ~37ms. the difference between 138 and 144 is the difference between 435ms and 417ms, which is actually only 18ms. but it's still difficult to tell between 138 and 152, which at a period 395ms gives a difference of 40.
This is related to the fact that half- and double-tempo relationships are privileged among the standard metronome markings. I imagine this came about to permit musicians a choice between setting the metronome to beat "in 4" or "in 2" over a range of tempi where either might make sense (especially for practice purposes)--and/or to allow for practice at exactly half speed.
Note that Weber's law was apparently modelled after testing people's perception of light intensity. It does not necessarily hold up for other things. For example, it most likely does not for loudness or pitch difference.
@@ithinkpaulmightbehavinastr8878 Not only that, but our perception of loudness is really complicated overall. It also depends on the sound's frequency, for instance. Search for Equal-loudness contour.
Well, in terms of DJing, the point is also that even small differences will make the tracks run noticeably out of sync within a short amount of time. And you might not mist want to transition, you might want to leave them running in parallel for a while and play around with switching back and forth or mixing in different elements. It helps if you don't need to correct all the time.
Adam. Incredible way of explaining notation so simply. It reminds of that quote written by who knows who that says, "If you can explain something simply, you understand." You humor and intellect are prime my friend.
3:59 quick note, as someone that’s going to college to study meteorology next year, i normally wouldn’t measure the temperature down to the decimal point, however close to the 32 degree mark, extremely small differences can mean the difference between a rain shower and an ice storm. i wouldn’t make a distinction between 68.5F and 69F, but I would make it between 31.5F and 32F
Hearing "I got to the party at 7:42" just makes me think of reports from investigations of deadly industrial accidents, where the timestamps increase in precision as you get closer to the explosion (or the meltdown or fire or toxic release or crash or whatever else). When the report tells you what an employee said at some exact second, you know something terrible is extremely imminent.
Love this, and Adam’s editing quality is absurdly good. I could listen to this man talk about anything. A thought that this brings to my mind is ritardandos and accelerandos (sorry if I spelt either wrong). You could, technically, write a new tempo for each note (which could get very complex if you have a complex rhythm happening) instead of writing “rit.” Just thought that was another cool musical example of this idea Adam was talking about.
I recently imported a MusicXML file of mine across softwares and had exactly this happen, instead of the ritardando I had notated there were about 32 new tempo markings within the one bar 😅 I guess that's how a notation software generates a ritardando for MIDI playback
@@tonywilliamspiano One interesting example of this is the 2nd movement from Thomas Ades’ piano concerto. He notated the ritardando with subdivided rhythms, because otherwise in this case it’d be difficult to put orchestral musicians in one exact way of rubato. I recommend you to have a listen to the beginning, guess how it’s actually written down, then check the score (the preview of it is available on Faber Music website).
There's an aspect I'm missing here - the fact that the human conscious isn't able to pinpoint a subtle difference from a certain point, does by far not mean that that difference doesn't affect the subconscious, which in turn alters your perception a lot. As in, the fact that you hear a song in a tempo that you have never before heard a song in. It's not too different of a discussion from different keys or tunings to people without perfect pitch, where they couldn't tell the difference without direct comparison
I would be very surprised if there hadn't been studies made to find out if people without perfect pitch "felt" different about imperceivably small tuning changes. My guess is there's no there there
George Collier is great because the people in the videos aren’t exactly playing what was transcribed but the transcription sums up what you hear really well
Hey man love your exposition skills. You're a great orator and really good at deconstructing topics into concise digestible segments (you seem to have a good understanding of the perspective of the listener/viewer)! Also I love your music bro keep it up you're an S-tier creative imo!
In experimental sciences there is a similar concept called "significant figures". If your measurement is only accurate to the ms, don't write down the microseconds. It's not just useless, it's misleading because someone reading the data could assume the accuracy of the numbers based on that. It is also very important when calculating the error bars.
I refer to this as spurious accuracy, but it is apparently more often called False precision. You often see it when people pass numbers through formulas, even something as simple as an average. Averaging three numbers might end up with an answer ending in .3, and that's fine, but ending it in .333, when it was a whole number to begin with, sets my teeth on edge.
@@Alan_Duval typically called false precision, i imagine, because a sometimes used definition of precision is the number of significant figures or somewhat equivalently the ratio of the max (random) error to the whole (reported) value, where accuracy is whether or not the reported value equals, up to the given precision/sig figs, the actual value (for intrinsic/systematic error). Thus, if you give too many digits, that may not affect the accuracy much (small places being wrong doesn’t matter as much as large places being wrong) but it implies your random error is way lower than it actually is, i.e. false precision.
Scientists, mathematicians, engineers, programmers, and technologists - at least those who learned and practice their discipline properly - are very careful to not confuse precision with accuracy. But artists usually lack this understanding. And usually don't need to know it anyhow. Sometimes high precision or high accuracy are critical, some forms of creative expression require it. But mostly it makes the art more complicated without adding any meaning or any quality.
This is why I find it funny, and a bit sad, when a trail to the lake is marked as 7 miles or 11.265 km long, or a radio's range is stated as 492 ft (obviously translated from rough estimate of 150m). The former is becoming a bit less common, tho - maybe they got tired of fitting all those decimals on the sign.
Part of the appeal of transcription videos is that it's a sort of subtitles for music. To an untrained ear, it might be hard to pick apart a piece of music to understand why it sounds the way it does, but with the transcription on screen, it's harder to miss something (as long as you understand it), and it gives you an amount of mental advanced notice for a sound you're about to hear so you know to be listening for it.
Decimals are very useful in gradual changes in tempo. they are also useful in soundtracks when you want the music to go well with the scene. Also sampling something might be easier if you have decimals in your tempo
For gradual changes in tempo, you're better off writing accelerando/rallentando and just marking the initial and final tempos. For your last point, he's not talking about the tempo you put in a DAW, he's talking about tempo in sheet music notation.
I don't know the first thing about music theory, but as someone with a linguistics background, this is very relatable. A lot of phonetics is literally played by ear until you reach the applied science (read: voice recognition), and even then, personal biases come into play quite a bit. Also, anyone transcribing in narrow notation is either trying to make an academic point or is being unhelpfully specific.
I think I agree with everything Adam said in this video! I have thought a lot about the relationship of western sheet music with sounds that are being retrospectively transcribed. I have worked as a transcriber for Michael Paouris for his contemporary pieces for greek bouzouki and we had discussed many times about how I should write some sounds that had never been written on sheet music. My humble opinion, that I have formed through this work and also my TH-cam video transcription, is that most times the sheet music should serve the sounds-music and not the other way around. I would love to hear some other opinion if you disagree. Have a good day!
I tend to use Decimal BPMs when I make drum covers, specially those that were not recorded at a metronome beat, like the old time Sonata Arctica, where I need to more or less match the "whatever" feel they were recording to
The first music I ever made were beats on an MPC 1000 in my high school because one of the interns was big into hip hop. He told me to ALWAYS set your bpm to a decimal number so that if a dj is using it in a set they have to work for it instead of just putting two songs together they know are the same. Of course, with bpm match on CDJs nowadays that doesn’t matter but I still do it just for fun!
I now tune my BPM with the key of the track. For instance, if I compose in A minor, my BPM will be 103,125 ( 440 HZ dropped 8 octaves). With a saw wave, you can clearly hear each click of the saw is in time. This allows to move from rythm to harmony witth ease. In this case decimals are useful.
Agreed, but for a different reason: I am more musically fascinated by co-prime relationships. For instance, in 120 BPM, I may set one chorus effect to 113 and an identical chorus to 131; the co-prime relationship to the tempo can add an organic feel to improvised music. Not saying your way is not useful, as it completely is, I just wanted to expand on your point to include that your reasoning used in an opposite context could be useful for a somewhat intuitive reason.
I feel like this sums it up perfectly. You might need to tell a DAW to run at 164.31 BPM to synchronize with a recording, but a musician can match the tempo on their own - and if they're not playing with a recording, then something like "Fast" might be all the advice they need.
Another thing which isn't really about musicality, but about giving access to a producer who would like to use your music: If a song is following a metronome, but the BPM isn't a whole number, if I wanna manipulate it in a DAW I would like to match its BPM. If the BPM isn't a whole number, I'm gonna have to go to audacity, pick two beginnings of a measure, count the number of measures, and calculate the BPM. Then, the more precise I got the more measures it will take for a desyncing to occur, so then I have to fudge with 2nd or 3rd decimal place like an idiot. Obviously if the original recording wasn't closely tied to a metronome, or three song changes tempo frequently then this is unavoidable, but if you are controlling the BPM, then be considerate. Musically you won't be able to tell the difference between 135bpm and 135.269, but the poor sod that would want to manipulate it would
How good is this guy’s content that despite talking deep musical topics to the point of making them almost philosophical he’s got 1’6 million subs. Hats off Mr. Neely, hats off
Decimal point tempos also might be necessary for calculating/programming metric modulations, (polyriddim, lol) but honestly, you could round up or down and nobody would notice.
Watch this video ad-free on Nebula!
nebula.tv/videos/adam-neely-this-tempo-is-trash-and-should-never-be-used
From a different perspective you could say it's similar to the increasing complexity of Cooking, Baking, Gastronomy... or wood floor, cement floor and linoleum... there's always good reasons for increased complexity, but yes... if it doesn't create a noticeable change then you have to ask Why you're doing it.
All systems need 3 base components 2:1. Water H2O 2:1 electricity current resistance voltage. All this knowledge yet you still can't see how it relates to you?
Sun earth moon???
Ever noticed when you take all letters of the alpha that rhythm with 3 you get a symmetry? BCDE-G-PTVZ
If Sungazer were playing on a space ship, an interstellar Titanic, who instead of crashing into an iceberg collided with a black hole: as time itself slowed to a stop, would you steadily increase your tempo to preserve the audience's experience or maintain tempo and allow the song to slow to a stop, drawing attention to the strange environmental physics?
I think perhaps the ultimate example of western notation just not being able to represent something is SWING! Of course, you can try triplets, dotted eighth notes or other tuplets, but the best solution is just to notate the music as if it were straight, and just tell the performer “swing!”.
In other words, if you micromanage the swing, the swing dies.
Should we direct the swing percentage at the start of the piece? (Genuine Q) if the feel is a component of your intention for the performer
My way of notating that is to denote the entire piece as 12/8, which is convenient because sheet music software will play it correctly (sometimes the recipe isn't for a Human)
I sing with choirs professionally and you get the same thing when more traditional white choirs try and do gospel or spirituals. They read it instead of feeling it and just doesn’t grove right. Sometimes they even screw up otherwise really easy syncopated sections because they are reading the notes, but they should just be feeling the rhythm.
It's even better to write it straight and tell the musicians to swing than trying to accurately describe it in notation. That way they can use their ears to find the ration that better suits the piece and even vary it across its duration.
Also, I'm not from a black community or even American, but I'd consider the second option to be disrespectful and ignorant about the nature of swing itself, it's a result of the embodiment of rhythm and movement derived from African traditions and not suitable for a fundamentally European form of notation.
The amount of swing is always improvised depending on the player/group. Same way tempo is a suggestion and the conductor will follow their own beat note to note.
Theres one very good reason I sometimes use decimal points in BPMs. I hit the tap tempo button 4 times and it was 93.2 and I said "cool whatever"
at that point wouldnt you just round down to 93? .2 isnt much of a difference
@@cy-bernet-ix Maybe. Or you could just ignore it because "cool whatever". On the very off chance that I actually can feel that decimal point (almost never) then I'll probably prefer how I felt it to whatever changes I make. And in the reality that I probably didn't feel it, I won't be bothered to change it in my DAW because I won't notice a difference either way, so why do the extra step of manually changing the tempo? I already got it in like, two seconds, I'm not gonna add two more to erase then type in more numbers.
At least, that's my thought process if I'm recording something random and do the same thing. I'll often round up or down, but I imagine OP's thought process is the same as mine on the days where I don't even do that. Nobody will notice the difference anyway unless they're trying to record a cover at the exact same tempo or something. I could be wrong, but I vaguely remember hearing that Andrew Huang does something similar instead of deciding "this must be a whole number because Reasons".
In all other circumstances I definitely avoid the decimal point BPMs, though I had no thought process for that before this video beyond a vague sense of wrongness or whatever.
@@cy-bernet-ix exactly it's not a big difference so why bother pushing buttons and twisting knobs to turn it down?
@@RobertKnutzen so this is something i was actually thinking about in general during the vid. one reason you might make micro corrections is either youre "repitching" samples and trying to pitch / tempo match, or because the performer could not physically play the riff / melody etc at that tempo, IE shredder guitarists doing extremely fast sweeps etc. as a composer i totally agree with just rounding up or down to a whole digit, but tons of people definitely just hit the tap tempo and move on.
@@Irapotato this depends entirely whether you see yourself as a composer or a guy just playing with an MPC
That opening one-take is an incredibly effective intro, Adam. Crazy how the lack of a cut makes such a big subconscious difference in our attention.
agreed!! contrasting Charles Cornell, if you watch him, who makes cuts at almost every sentence
As a whole, the cuts in this video are very minimal and/or well hidden.
@@Allen-rv9jz I think this is in large part due to the structure of the videos themselves. Adam scripts these out, so he can read large chunks at a time, whereas Charles speak more off the cuff (especially in his reaction videos of course) so they have to edit it down to make it a usable video
@@kevinwells9751 that’s a good point. I guess I prefer a more thought out video then
Good one-takes always remind me of the show Good Eats
Weird fact: in rhythm heaven fever the music for the minigame air rally is played at 162 bpm, but when it returned in rhythm heaven megamix, it was sped up to 162.01 bpm
god… just that lil bit faster…
Sounds like the producer accidentally tapped their mouse wheel and exported the new mega mix all with that mistake 😂😂😂
there are also many instances of games being played originally at, for example 150 bpm and in megamix theyre at 149.999
floating point error maybe?
@@nappeywappey How does anyone even notice that kind of stuff? People just go around comparing versions?
I DJ electronic music and I always told people and beginners, that you can easily change the tempo of a track by about 3 BPM without the crowd noticing and even 3-5 is ok. I never had any sientific base for that, it was just feel. But it makes total sense
You do need to be careful with sudden shifts, though. Especially with older tempo changing methods that would adjust the pitch as well. A semitone is a 6% change in note frequency (aka pitch), and so a sudden change of 1% is noticable, as that 1% is about twice the Just Noticable Difference in pitch (generally quoted as being betwen 6 and 10 cents).
I absolutely feel 5bpm - that's the number I go up when I'm practicing technique. 5bpm is a big jump when you're at the top of your level. But even without trying to play it, you can easily hear a 5bpm increase just listening to the metronome.
@@jada90 for musicians for sure I agree. Even 2-3 is noticeable if it's something you've spent time playing. But for the general audience that can't clap on 2 and 4, I'd argue you could push it abt 6 bpm lmao
Yep. 3 BPM increments will never be noticeable and you can change genres over the course of 6 or 7 songs due to this naturally
changing tempo in dj sets can actually be pretty cool if done right. it feels weird if you do a small amount (like 3 bpm) slowly, because the groove is lost. but if you cut the previous track entirely and then play the next track +/- 3 bpm, it sounds super dope. also it can be a good way to transition into slower/faster stuff when you dont have enough room on the pitch slider to beatmatch it with what youre currently playing
I loved this. As a linguist, I wanted to say "yes, in language, too!" to literally every section of this essay. Phonetic transcription is similarly lossy and sometimes overly detailed without being able to tell you the difference between whether it sounds like Steve Urkel or Darth Vader. The term for overspecificity changing the meaning of something is called "Gricean Maxim violation" in linguistics/pragmatics, and it might be interesting to look into more with respect to music.
Had the same experience! Also how transcription is inherently interpretation
(There's a great video by Tom Scott on Grice's maxims, if anyone is curious)
Ryan, thank you so much for pointing out this term. I’ve been interested in the subject for about two years now, and haven’t heard that.
Yeah, that's true. The IPA is basically like music notation. Sure, it's possible to annotate how a vowel is lengthened or fronted or aspirated or nasalized, but that'd just be like saying you went to the party at 7:42. There's a reason for the lossiness; that's required of any system to model something.
@@vigilantcosmicpenguin8721 the IPA vowel notation already includes information about height, frontness and roundness though, without any additional diacritis and markings used, /i/ for instance is different from /o/ on the basis of these features, so they are not redundant and it's far from being similar to the 7:42 situation that Adam talks about
I'd agree with nasality and aspiration, but only as far as English goes, since there are languages that have minimal pairs where only nasality or aspiration changes (ancient greek for the latter, for example)
"If it *sounds* right, it *is* right" - groundbreaking British recording engineer and producer Joe Meek. I wonder what the *precise* tempos (and how unsteady they are) of many of the most highly regarded songs are? That would be an interesting analysis.
A long while back I did a remix of "Single Ladies" alongside "Liquidator" by the Harry J All Stars, and... that took a lot of tweaking. And there's nothing wrong with either track, but the one that was clearly recorded to a steady clicktrack was a lot easier to work with. This is where decimal tempos come into their own though - when you want to sync two tracks that aren't the same tempo, you need to find their precise bpm and then find a happy medium between them.
yes! I'd like to see an analysis of most heard or awarded songs tempo
@@gabrielv.4358 I made a spreadsheet for myself, of the 1100 most popular tracks in the past 20 years, all with precise BPMs specified. I won't share it cause... hey that's a lot of work xD But if you want I can give you averages or stuff, anything you wanna know?
Pantera used to sometimes tune their guitars to something that felt right. No tuners or nothing.
Dew it
4:41
Adam: "It represents more concerns with being precise and right, rather than being musical."
All Rhythm Game music composers: *_*sweating*_*
Kobaryo my beloved
Frums my beloved
@@HokoraYinphine Aegleseeker my beloved
@@RJVZBloxerYTXNOR XNOR XNOR my beloved
Sotarks my beloved
Absolutely agree with you regarding notation, but within the DAW I’m an ardent decimal points tempo person since I use tap tempo to “feel” where I’d like my recording to be. I’d also point out that if you’re only measuring the just noticeable difference of quarter notes, you overlook that we can perceive differences in durations of longer notes within that tempo :)
i think thats the most popular usage of decimal point tempos yeah
I don't recall asking you Andrew.
I tap tempo, then round to the nearest even number because it musically makes no difference but I'm bothered by seeing decimals in the tempo slot.
@@Yokai.Wakukhan I don't recall asking you to ask Andrew
I love to see that other musicians/youtubers get things out of each others content. Also you guys should collab and make a sick jazz pop song
To me it seems kind of like comparing 29.97fps to 30fps. Your eye can't tell the difference between the two and it generally doesn't matter which frame rate a video is in, but it becomes extremely important when combining and syncing audiovisual content.
It is the same for music, an accurate bpm is sometimes very useful. If the source audio is at 100.25bpm and you add 100bpm audio to it, it gets out of sync and sounds absolutely awful. This is something that I occasionally come across when creating mashups and it is terrible to see people mark down the bpm incorrectly without the decimal point.
the decimal in fps is actually standardized in cinema though because people do recognize the difference
@@budswayzeniger3473 You really can't tell that difference by eye. The reason that decimal exists is actually because of the color TV and the way it was implemented on top of black-and-white television.
@@MLGaeming standards are standards. it'd be the same if 30 was the standard.
@@budswayzeniger3473 Both 29.97fps and 30fps are standards and commonly used. Mixing them up can lead to issues, hence it is important to know which one a video clip is using in some technical contexts.
Yeah as a mashup artist at times the stems just go off beat when they are like tenth decimald off which makes things unnecessary hard
2:58 A slight detail here: we can't actually say that the just noticeable threshold is at 30 bpm based on Shawn's experiment. His experiment asked participants to correctly indicate the pulses were changed. This is finer information than just that the pulse was changed in some (perhaps unidentifiable) way, which is what the just noticeable threshold measures.
I haven’t watched Shawn’s video, so I can state this without bias:
That’s technically true if there is no control. But the distinction isn’t meaningful.
If people can’t tell 500 ms didn’t shift to 499 ms (the smallest “earlier” at this resolution) when it actually shifted to 530 ms, then the “actual” just noticeable threshold is 31ms.
You are describing something other than the just noticeable threshold, though. We want the pulse frequency where, when altered, 75% of participants can indicate that it has been changed.
Shawn's experiment had three responses: late, early, or on time. By misidentifying those who noticed a change but misidentified the type of change, you are introducing a systematic statistical bias and you will misidentify the just noticeable threshold.
Now for Shawn's purposes, this is fine because he wasn't looking for the just noticeable threshold, as explained in his video. But this does make the experiment unsuitable for Adam's purposes.
@@tracyh5751 It's easy to imagine that you're hearing a change if you're actively listening for it. I think a fourth response option would be better ("could hear change but not identify late or early"), than to rephrase the question to be just "Did you hear a change or not?", as that would introduce much more uncertainty into the study with only two responses "yes" or "no". Also, by adding a response option, you could measure at what point (average ms) did the fourth response option ("could hear change but not identify late or early") start to get used more and more frequently.
There is also the issue of consistency: How often do we actually play a piece in EXACTLY the same tempo from start to finish? I find that I often have to change the tempo on the playback from say 80 to 79 to 82 in another passage, but I see no point in writing it down that exactly. The thing is that the "right" tempo mgiht also depend on the performers and how they percieve the piece, accentuations, dynamics etc.
I don't think I even play a piece in the exact written tempo. I tend to approximate using the pulses of pieces I do know well that I also know the tempo for, and there are some large jumps in that list.
@@papercraftcynder5430 BPM is also a strange way of measuring, since the difference between each step becomes smaller the faster tempo we are dealing with. In a tempo of 190, a change to 191 is barely noticable if at all, but in a tempo of 22, a change to 23 is huge - it becomes a completely different piece.
"They're more like guidelines anyway!"
Yeah who cares about a metronome
@@NidusFormicarum This is actually true for all things sound related.
We measure it in frequencies, which tell us how many times it cycles per *second*. So a 20 Hz tone cycles 20 times per second. A 20,000 Hz tone cycles 20,000 times per second. That's our audible range (at birth and during childhood), and humans can be pretty good at hearing a difference between 20 Hz and 21 Hz. They just absolutely cannot tell any notable difference between 10,000 and 10,001 Hz though.
So we devolved scales that compress these into formats that match our perception better. The exact same thing is done with the amplitude (or "volume") of a sound. The dB scale gets wild, because 10dB is barely perceptible...but 100dB is quieter than a car horn...and is has 10 BILLION times more energy behind it. That isn't a typo.
So the BPM measurement is...fine.
In college composition class I used a tempo that wouldn't be found on a metronome marking, and my professor called it pretentious.
He was right.
Sounds similar to my experience intentionally overusing tritones.
I wouldn't even know what markings are on a real metronome, I've never seen one in person. I didn't know they 'skipped' numbers til this video. When I make music I just start in a ballpark BPM number and maybe adjust a few up or down from there over the course of making the music. Rigidly adhering to the specific numbers on a centuries old device seems far more pretentious to me.
Literally the same thing happened to me a few weeks ago, I marked it at 90 and was called out for it. Wild
@@robinlydian4452 Personally, I could understand decimal points drawing remarks or possibly odd numbers such as 91, but being called out for something like 90 to me as a failure to accept and adapt to modern sensibility. Electronic music is often written in multiples of 5 (not always obviously) and thus I would argue is common practice, just not traditional practice
Sheet music never lies. I remember getting a sheet from another student (Rachmaninov Prelude), and there was a note made with a pencil: 'You can't play this'
Guess what... they were right
😂😂😂
To be fair, most people need to omit the top and/or bottom notes of Rachmaninoff's chords to even start playing his pieces
@@kurumi394 I've seen other scores that are not possible for people with less than six fingers on a hand, or look like they are reductions of a duet that were rendered sloppily as to what is actually possible.
I think your recipe analogy is right on the money. I'm part of the chef team of a restaurant and we have discussions all the time about how much detail to put into our recipes. It seems intuitively obvious that the more detail you put in, the more consistent and accurate the final dish will be, but if you make the recipe too detailed it gets in the way of people using their experience, judgement, and ultimately taste. The feel of a decimal tempo is kinda the same as the feel of an exact number of grams of salt; I appreciate the effort but I'm mostly going to use this as a rough guide and adjust to make it the best I possibly can.
I hope to some day own a restaurant and am studying at a cooking school atm, I agree that too many details get in the way of a personal touch and can even make it a bit bland
Also, got any tips for this field of work?
Yes exactly! I see the decimal notation things as goofy recipes that are intentionally outrageous. Like Allegra chicken. No one wants to make that, especially not via the recipe. But it's really entertaining to make jokes about it and write unusable recipes for it.
@@amora1246 This is going to sound really glib and shallow, but honestly just find a good chef and work for them. School is useful for teaching the general principles but nothing can really show you how those principles apply in the real world besides doing it in the real world under a chef that wants to see you grow. When it's time to move on, find another chef to learn from and work for them. And the whole time make sure you're making it worth their while to teach you. Best of luck!
I recently listened to a lecture given by Brian Eno, where he spoke of his collaboration with Italian tenor, Pavoratti. The first time that Pavoratti entered a studio with Brian Eno, Eno and the musicians present started to attempt to throw something together, to write something new using the studio techniques and equipment they had available. Each musician probably arrived with a few ideas that they had developed on their own, and the group started to try and see how they could get these ideas to gel together. Eno said he looked at Pavoratti, who had a look of surprise and incredulity on his face. He said something to Eno like “You are making this music right now?” He had never in his life experienced anything like that. A recording session to him was where you got together with a symphony orchestra, broke out some old warhorse, e.g. Bizet’s Carmen, and started to rehearse and record. He just never realized that it was possible to make music on the fly. And he was amazed, and Eno and Pavoratti began to have deep discussions on the topic.
Pavarotti, not PavOratti...
@@psykeeq …
Pavo is turkey in Spanish, and Pavorotti was a turkey whenever he attempted to lipsync a concert instead of rendering a live performance like the people were paying for…
Segovia had the same reaction upon witnessing Django play - 'You made it up off the top of your head'?
@@MichaelHarvill98 …
And was thinking “… and with only three fingers?”…
@@tribudeunotwo
I've missed video essays like this. I like your performance diaries and Q&A's, but these more esoteric topics are what really get me going.
one piece of “transcription notation” that i love is just writing “laid back” or “rushed” instead of accurate notation as it allows for you to worry less about accuracy in notation and more about accuracy in performance
Same thing applies as when Adam speaks about BPM. Including the exact rhythm of rushed or pulled back notates is useless because it’s a pain in the ass to read and it’s not how the performer was hearing it. Whilst notating the exact rhythm is more ‘accurate’ - its useless and helps nobody
That's just "rit..." or "accel..." (if you mean gradual change) or "meno mosso" or "piu mosso" (if you mean sudden change), but using English because you want to look cool.
@@rosiefay7283 rit. or accel. implies the band underneath is changing tempo too which is why it's not used in this case
@@rosiefay7283 You're using Italian to look cool so let's agree to disagree
In fighting games it's common to have combos that require one or multiple 1 frame links, meaning you have a 1/60th of a second time window to hit the combo. Always found the rhythms of combos interesting especially when played on an arcade stick
And speedrunners on almost any game do that too
Have you seen the account that turns infinites into music? it's super satisfying
They can make an interesting rhythm, definitely. You can literally hear if somebody's combo will come out, based on the noise their button presses and stick make.
Unless it's SFV lol
@@pedroscoponi4905 whats the name of the account?
a few years ago when I first getting into music production, I made this melody that had a super wack temp like 135.27 or something like that. it happened after I accidentally messed with something and I was to inexperienced to know the difference. A year later I revist the project because I like the melody. Noticing the wack tempo, I go to fix it, and for some reason the synth sound I was using on the melody sounded super different and not nearly as good when I fixed the tempo. So somehow I have this melody that only sounds good at a wack tempo lol
Probably due to an LFO which is synced to the tempo.
It either 1. Is a mistake somehow or 2. You just got so used to it that you don’t like how it sounds. Once you listen to a song enough it can be impossible to be objective about it
@@melody3741Yes, exactly. I listen to my own stuff in the car, cuz I often get ideas for revisions or new songs there, and take it from me, if you want to listen to your own stuff regularly, you HAVE to be good at fighting this feeling. It's possible, but not easy sometimes lol
No tempo is whack it's the stuff u put in the track that's good or whack
@8:35 "Man, this food analogy is going too far" - No, thank you for it. By making that analogy, you helped me link aspects of music (which I'm bad at) to cooking (which I'm good at). The examples you gave demonstrated playing by ear vs by practice vs by sight as ways we think about many things in life, and I appreciate the depth you put into the analogy :).
@13:49 "Because it's inherently a lossy format" - Now you're speaking my kind of nerd-speak! Musical notation is like JPEG - our minds fill in the missing bits :)
@@Tawnos_ I'm surprised your mind jumped to JPEG instead of MP3 (or any other lossy audio compression format), but I had the same basic reaction. I was like "AHA! Now I understand!". 😆😆
Yes, in fact he kind of didn't go too far enough - there's a salient point about good cooking and good music that the recipe analogy highlights - one major advantage of the lack of precision is that it allows for interpretation - such as adjusting ingredient ratios or cooking temperature to get a unique taste/texture, or micro-variations in tempo or volume to highlight the feel of the music. Precision recipes/sheet music are great for mass production, but stifle the creative performance aspect.
@@Tawnos_ bmgj
bm 6
He wasn't saying the analogy was bad just that it had been going on too long
Adam's channel, at this point, is an extended series convincing people that music is generally overanalyzed in all the wrong ways and underanalyzed in all the actual interesting ways.
That's a great point, TheDoffin.
…to him.
I love the channel and appreciate the insight, but you’re just describing what any (music theory) channel or podcast amounts to.
truuuue
Which is true.
@BOP I disagree. Many use the usual methods to draw their conclusions and do it very well.
Strange example of decimal BPM: Hurricane by Kanye West is 160 BPM on the dot, but its earlier leaked versions are all 157.61 BPM, since they all use the older Yandhi beat. Since Moon was initially created as an outro to those earlier versions, it, too, is 157.61 BPM, even though Hurricane isn't anymore.
The idea is that, although it can be played in decimal BPMs, when composing it's almost never that specifc, as that's pointless. Supposedly the only exception is for composing for films when audio/video synchonization is needed, but I personally doubt that justifies using decimal BPM either, as it's also possible to just mark down a few key notes that are required to be synced with the video, and everything else can shift a bit without getting noticed.
@@FTZPLTC Film scores are often recorded live I think, aka performing in real time along with the movie being played and the conductor can control the pace to meet those key points (or at least that's how I would imagine things to go). Alternatively it's also possible to stretch the audio in post to meet those key points.
And for live mixing, yes, exact BPM is important. But just relying on the BPM reading is not enough, you got to have a device to send the metronome across all devices, and then turn on BPM syncing for it to work, otherwise everything would still drift away from each other.
So in summary, although decimal BPM technically exists in recorded or played back tracks, they should be perceived as whole number BPM with a margin of error as well.
@@FlameRat_YehLon You're right that scores are sometimes recorded with the movie playing. I don't think it's super-common for music to be played to sync up precisely with the action though. Sometimes it's done, but if anything I'd say the metre of the music is more likely to be used to set the pace of the edit. It would be unusual for a film editor to just happen to making edits that formed a coherent pulse for the music, after all.
I'm rambling though, I find this subject really interesting. And, again, watch Cats to see this done fantastically badly and arse-about-face.
@@FTZPLTC Thus I said key points. Movies can be designed to have a few key points that are synced to specific notes on a piece of music, and every other notes can be played in a more flexible way, just to fill in the duration between key points.
Sup yikes do you prefer dogs or cats
2:05 I love how you made the subtitles so they don't spoil it, thanks!
I've heard it said of scientific models that all models are wrong, but some models are useful. Maybe something similar applies here. All notation is wrong. Some notation is useful. And we can even take the parallel even further. Each scientific model has a certain regime under which it useful and outside of which it's not. Some notation is useful in one context, but not in others.
Bach seems pretty writable. I do feel Jesu Joy of ...as triplets, not as 12/9 or whatever it is written as
"I've heard it said of scientific models that all models are wrong, but some models are useful."
That's a quote often attributed to statistician George Box (though it predates him as a concept)! The more you know 🌈
Yes, it’s a common quote regarding statistics (which is a science). Statistical models provide best guesses with ranges of certainty.
@@danparish1344could apply to modeling in general though because all models are constructed based on summerising as many variables as can be accounted for in order to abstract, reflect or predict the shape or pattern of something. For intents and purpose, if we consider enough variables, we have a functionally usable model. But in order to stay consistent, it must be open to unforseen variables and adapt to new information. No model can be absolute; therefore all models are still "wrong", no matter how functional and useful they are at the time.
Regarding wrong models, that's probably true. Take for example all physics calculations: they all ignore extreme variables, like the inner flexibility of an object. There's a video about how long does it take for one end of a metallic bar to move after you push the other end. In plain physics it would be 0.
Talking about tempo and perception, it might be fun to look into the music genre "Extratone", where the BPM is so fast it becomes an, you guessed it, extra tone
this is called granulation. welcome to the club
Isnt that called weirdness?
Thats not why its called extratone. DJ Heinrich invented the term, from german words "extrahieren" (to extract) and tone which is the same in german.
@@rykehuss3435 I've never heard this before, interesting. I was just going about what was told to me before, I stand corrected
damn that's a lot of purple, tossing some yellow in here for some contrast
The issue here is known in statistics as "overprecision", or more fittingly, "drama digits". It is the misguiding use of precision that simply makes no sense in a specific context.
I haven't heard "drama digits" before, but I may have to find an excuse to use it now. (I do bioinformatics things, so tools giving incredibly fine grained results when you put in a handful of vague data points is basically standard.)
basically, funny. that's why
@@dnebdal haha, indeed, I like that word, too. I can well imagine that in your field you will have amble opportunity to tell people to drop the “drama digits”. Maybe at least sometimes opt for calling it something more scientific-sounding, like “spurious accuracy“, because … well, just because. ;-)
Yes, but 94.732% of the time, adding digits makes the statistic you are quoting sound more authentic to neophytes.
The food recipe analogy was so perfect, I absolutely love it. I think it’s so true that the notation is merely a recipe for us musicians to make the musical story come to life, and it is incomplete as well. And I completely agree that there are times when notation is not at all helpful, and it would be used as analysis or even something more trivial. However, I find myself so grateful for music notation and desiring to keep using it all the time, I love it so much. I don’t feel the need to rely on notation, but just the sight of a five-line staff, clefs, key signatures, time signatures, and notes on that staff bring me such a joy that I cannot describe. I love music notation, I just love it.
another time to use decimal points in bpms is when you have a tempo change that's fractional to the base tempo! like in a song i'm writing where quintuplets in 4/4 become 16th notes in 5/4. it goes from 94.4 bpm to a clean 118
That's technically true, but most of the time it's better to just notate what change happened instead of putting the exact tempos. Like, start the song in 95 bpm and at the transition put something that says quintuplet = 16th note. That way, it's obvious what you actually intend to happen instead of having to check the new tempo on an actual metronome.
you mean, like what Adam Neely did on Tuplets for Todlers? Yes, he finds that to be a legitimate use ^.^
The beginning could just be written as 20/16 time signature w 118 bpm no? The main pulse felt would change from every 5 16ths to every 4, but AAL do exact same thing in their song The Brain Dance. A lot easier to see the correlation between the 2 that way imo. I understand recording with a click would be trickier tho, could also use some midi instrument to make your own click emphasizing every 5 16ths.
a similar issue i had was trying to make a quintuplet swing groove 'last' for two more notes, (essentially as though it were 5/16 with occasional bars of 7/16, but i'd already established that it was quintuplets so didn't want to change the whole system). I found the fractional bpm thing way too technical for how easy it was to feel.
@@goatmagic5220 You could do that, but it would be a bit harder to read than quintuplets in 4/4. I think just writing it as metric modulation is nice, but either way works as long as you don't try to make someone read 94.4.
A perfectly detailed 1:1 terrain map would be totally unusable despite being correct.
The lack of unnecessary details is what gives notation meaning.
I disagree. What you're describing is a recreation. They create perfect 1:1 reproductions of terrain for the purpose of forensics, or military training. They're less common but sometimes you really need to re-create a sense of scale.
@@nathanjasper512 yeah but it wouldn't work as a map, that's the point Three Colored Squares was making
@@nathanjasper512 No you've missed the point as another user has pointed out. The point of sheet music is not to recreate, only to provide a useful abtraction of the music - much like how a map is an abtraction of an area; in the reduction of details comes the utility.
Sure, you can have a 1:1 reconstruction but that's is not what we are discussing today. You're disagreeing with the paticulars of Three Coloured Squares' analogy but not his point
The point of sheet music is to be able to hear your ideas from the playback function of the notation software before you go on to orally explain them to other musically illiterate musicians
The closest we have is a recording, which cannot be replicated by musicians - spectrogram sheet music doesn't sound fun to read...
3:43 Adam is explaining the linguistic theory of Relevance. It's sometimes known as Grice's Maxim of Quantity. Always pleased when Adam sneaks some linguistics into music theory videos
Pragmatics!
Is this the concept Tom Scott talked about in that one video? He used an example like "asbestos-free cereal"
@@boblobgobstopper13214 Tom Scott is great! Didn't realize he had done a video on Grice's Maxims.
@@boblobgobstopper13214 Indeed it is.
As an admirer of the existential philosophy, I would say, this thought he is sharing is actually saving music from mechanical objectivity and justifying the randomness of human actions, like a slightly delayed note that brings all the beauty.
Yes! I'm actually now more inclined to produce music with decimal BPMs because (a) it's not noticable and (b) it will absolutely be felt. Taping a BPM in a DAW always made me dread the decimals but now I see it in a new light, I'm keeping the decimals in.
Having worked in music scoring for film and TV for 15 years, I can say you don’t always need fractional tempo to match musical beats to picture, but sometimes you do. Especially for cartoons with a lot of action or comedic beats, or to hit picture cut, etc. And in those instances decimals are sometimes required. Without them the piece is simply incorrect and does not contain enough information to achieve its core purpose which is to match picture. But if you want to adapt that music for public performance, by all means strip off the decimal points and round it off because of course no one will perceive the difference.
I have to call you out for this. Especially with cartoons, they're originally close to 11 frames per second, so I can't see why you would need to have hundredth of a second precision with sound. Even when you have 1 minute long scene with one score using 120bpm vs 120.5bpm you have roughly .2 second drift. In one minute you have skipped literally one frame. Half a BPM I can somehow understand in very long scenes with very long scores, but going any more precise than that is just not very useful. Also, many composers just play watching the scene organically, like it was used to do before computers. People didn't use fractions of a BPM then and I hardly believe many does so today.
@@eliteextremophile8895 Yes but the viewer would easily notice a 200ms delay between the score and the animation.
Also let's not forget physics. 30ms is the time that sound waves in air need to travel roughly ten meters. 5ms is the delay of a sound played 1.7m away.
Have you ever stood 1.7m away from your fellow musician (or your amplifier box) and felt the delay, or that you couldn't get the beat right? I haven't.
I was literally INVESTED in the tweet thread on this topic you were involved in, so it's great to have an actual video on it!
i saw it too lol
You were literally invested? You were traded in return for stock?
@@smergthedargon8974 For knowledge.
@@segmentsAndCurves What an unfortunate opinion.
@@noonehere0987 Yes, and?
I freaking love your videos. The preparation, the info, the delivery, the editing, the everything…it’s all so enjoyable and interesting every single time.
As a musician, I was glad, that you revealed, that the fifth note was slightly late. I "felt" it. It is like the ear trainings for pitch. I can tell if the next note played is 0,7 hz higher or lower. But I don't hear it anymore. It is just a feeling. But in the end, I am right. I love these challenges!
I think it's always connected to what you're doing and how often you do it. I didn't notice the change in tempo... but by zeus I'll see small dips in framerate where my buddies say "you're imagining things, looks fine to me"
I've been playing rhythm games for nearly 15 years now. When the difference between a perfect score and a "choke" can be as little as 10 milliseconds, the late note was heard loud and clear.
DJs heard it.
If he'd not told me, I'd have obsessed. I'm still not sure if I really noticed it, or I guessed.
@@oblivion2k490 I used to play mungyodance and osu shile back and I thought it was late, but I don't know if it was a wild guess.
0:13 nice
this is why i love this man
lmao
This is the sort of stuff I addressed in my Grad Dip thesis, "The Communication of Type-Setting". I came up with:
"The purpose of typeset music is to let the reader recreate the composer’s intent under the worst performance conditions. Printed music is a set of instructions for the performer, not the composer. The printed document can be used in two distinctly different ways: either for slow time-relaxed consumption (such as for practice, memorisation or analysis), or for time-critical events, such as sight-reading or live performance. The music has to be able to meet both uses."
But. Maybe there's another purpose for written music - to record aural objects. But I think a sound recording would be better.
Thanks for the thought-provoking discussion.
--------
I was at a rehearsal this week and the drummer wondered why ♩=69 rather than ♩=70. We all agreed that while 69 was "nice", 70 would probably be simpler to comprehend. I thought it was either to do with gears in mechanical metronomes (numbers of teeth on cogs, etc) or to make simple ratios for changes between 2-based time and 3-based times, such as 3/4 to 6/8, or hemiolas. Now I know it's the minimum perceptual distance from the previous marking. I presume the datum is ♩=60, and the rest were derived - probably empirically - from there.
--------
I sometimes add stuff to amuse future-me. In my thesis, in the section on timbre:
"Timbre also seems to miss out in the typesetting world. This is paradoxically because there’s too little to say, or too much to say. It would be extremely difficult to get a clarinet quartet where:
· One clarinet should sound like the second flautist of the Berlin Radio Orchestra in 1963.
· One clarinet should sound as though it’s using a split number 1½ reed where the player is used to playing on a number 4.
· One clarinet should sound like it was invented on Neptune, at Easter.
· One clarinet should sound like it’s being played by a trombone player, who had to quickly borrow any instrument that was available, because his own instrument was stolen by a gang of fake portrait artists on a tram in Rio di Janeiro.
It would be impossible to get a second quartet to play the same music.
Although rather extreme, the example shows that timbre cannot be made too specific. Rather, “four clarinets” has to suffice."
i enjoyed reading this, thank you
One would have to approximate with terms like "sweet," "cronchy," and so forth. There are even some standard Italian musical terms for this. If you really want to capture a timbre, then find a way to record a sample of it.
"You got to use your ears, to hear music" - Adam Neely 2022 😂
@Ke baX damnnnnnnnn
Beethoven and Helen Keller disagree.
@@EmpiricalPragmatist well... they dont hear music ....
@@SreenikethanI Depends on your definition of the word "hear". They could both perceive vibrations well enough to distinguish variations in both pitch and rhythm. Does that _have_ to be done using the ears to count as hearing?
@@EmpiricalPragmatist ah gotcha, I had the same thought in mind, just that I wasn't sure to consider that as "hearing"... maybe "perceive" is a better word here?
3:06 "69 bpm, nice"
This video essay really benefited from that, Adam.
He said it cause funneh
69 likes
I love this video and I agree wholeheartedly with it... apart from one little caveat: When creating a video animation that moves to the beat, I compose a piece of music (or time-stretch an existing one) to match a beat to a whole number of frames. That means, at 25fps, if I want a beat every 12 frames, I need 125bpm, or for every 15 frames it's 100bpm. But if I want a beat every 14 frames, it's 107.142857bpm to stay in time. 16 is 93.75, 18 is 83.333333 (or 166.666666). 🎶
I feel that was mentioned (as one of the quoted reasons that wasn't further engaged with). For such purposes though, wouldn't it be more informative to write 25/14*60 bpm (or even 25/14 bps?), rather than 107.143?
I also would be that specific if I needed someone to record their part separately. However, I'd also be including a guide track, and the tempo is just in case they want to DAW it up themselves before they submit it.
As others have mentioned, fractions seem more useful in this scenario as they say where you got that tempo from, and are actually _more_ precise!
or you could use fbp (frames per beat)
@@adisander My DAW doesn't accept vulgar fractions, only decimal ones.
This video is pure entertainment and learning combined.
Great video Adam!
The first thing that came to my mind when I saw the decimal points was synchronisation of different tracks. I quickly tested a 120 bpm with a 120.5 bpm and after about 5 seconds (or 10 beats) the asychronicity became noticeable to me, a few seconds later it was quite obvious. This kind of corresponds to visual perception, where fine differences in structures might be unnoticeable, but overlaying these structures makes the differences become easily visible (think of a Vernier scale or Moiré pattern).
An interesting and relaxed video as always. :)
I thought of such music structures which go in and out of sync periodically. However it's probably more convenient to use fractions rather than decimals so you could tell the pattern by looking at them (even if you're a mere human). Sadly DAWs don't usually support fraction tempo and we have to stick to decimals. And decimal values cannot be precisely represented using binary values. Game over :-D
@@AntonNidhoggr You mean beat, the interference phenomenon? Yeah, it's a good example. And I agree, fractions would often times be more convenient.
I guess that's where the just noticeable difference comes in. If you put the two against each other, sure, you can notice it, but seems like no one would ever really play anything like that.
@@vigilantcosmicpenguin8721 A friend who was DJ for electronic music for a few years explained to me that the problem can arise during a transition between two tracks. With a fading transition the two beats inevitably overlap for a few seconds and thus the bpm must match perfectly (at least good enough so the asynchronicity would become noticeable just after the transition).
So there is something off...
On from minute 4:06 to 4:19, Adams says that he arrived at the Party at 6:42, but would say that he arrived at 6:45 to avoid seeming suspicious.
But from minute 15:01 to 15:09, he refers to arriving at the party at 7:42 or 7:45...
And this sounds suspcious, like a massive deception in need for an alibi!
omg
14:44 - "I like the sound of the push and the pull of different downbeats not lining up"
Oh boy, would you love listening to me practice then.
This is just insanely good editing. Well formed and easy to understand. This man loves his job.
Another justification for a decimal BPM for a performer could be when that performer isn't human. Currently a machine can't feel a tune so precise notation is required for the "performer" to create the music as intended.
Otherwise it is a BPM of how things ended up lining up because a machine was running at such a speed
The way you described the inherent impossibility of perfectly complete and unbiased sheet music reminded me of something I've heard once about maps. It is impossible to fit all the detail of a perfect aerial view onto a map, and if there was a marker for every detail the map would literally be nothing but markers. So, it's up to the artist to decide what is important, what can be simplified, what details can be left out, and what should be marked. Thus, there are an infinite number of ways to make a map of the exact same plot of land, no matter how big or small, and each map is inextricably tied to the mind of the person that made it.
This kind of reminds me of this little "amazing facts" book I had in my childhood. It listed the height of Mt Everest down to the cm well before we had the technology to do stuff like that. Eventually I figured out they had simply taken the height in whole feet from the original edition and done a conversion that resulted in that false precision.
Significant Figures!
@@onyxv.o.4830 now we just need to include error bars in our music notation
@@janmelantu7490 I mean, that would work pretty well if you want precision but also replicability. "Play this music at 70±2 BPM."
If I remember it right, the original measured height of Mt. Everest was exactly 29,000 feet (to the nearest foot), but it was then declared 29,002 feet to prevent people from thinking "29,000 feet was just a rounded estimate".
Decimal point tempos is actually a really big problem for people who do mashups because for a little while it seems normal and then it becomes a mess…… #JusticeForMashupMakers
Also i love the music that comes before you talk about a topic, i want a full version please!
Lol hope this blows up
Hilarious 🤣
Was once told a story by a percussionist who was playing a piece written mainly on a computer. The percussion part was all over the place, no identifiable pulse, etc. So he asked the composer how it should go. Composer couldn't tell him. He'd essentially just clicked his mouse randomly into the software. So the next question was "would you notice if I got it wrong?" and the answer was "no".
precise notation doesn't always help convey what you want.
"Everyone knows poly-rhythm's get the clicks!"
Couldn't say it better myself.
@Kevin L See also: metronomes
Microtonality is the BOMB
I'm a classical composer, which means I almost always deal with live players and conductors and almost never with computers.
I always remember a thing one performer said to a classmate during a composition masterclass: people aren't metronomes. Unless you have a clicktrack in their ears with the precise 133 BPM you want, no performer is gonna perfectly adhere to that. In fact, since in most professional settings performers have very little time prepping for a piece, metronome markings will be practically pointless unless you specifically bring in a metronome.
Because of that, I always use words to convey tempo (adagio, presto etc.) and even if I add a tempo marking, it's always in brackets, always marked as approximate, and always rounded to the nearest multiple of 10 or 4.
Isn't the conductor a "metronome" that they can follow?
@@gyroninjamodder first of all, in chamber music it's rare to have a conductor
second, even if you have a conductor, they don't go on stage with a metronome or a click track, and so an exact metronome mark wouldn't be of use once again.
I sometimes put in an approximate metronome marking in case there's a soloist or a conductor who studies the piece in advance and can practice it with a metronome, but even then there's really no use for a mark like "131", as their actual performance will always be somewhere around that and not the exact value. so I write "~130" or "~84" etc, because the exact value really doesn't matter all that much.
Not at the level of precision of 132 vs. 130 bpm, but some conductors, such as Toscanini or more recently Benjamin Zander, made powerful arguments that Beethoven meant for performers to take his metronome notations seriously, and that the fashion for playing the symphonies at much slower tempi made them lugubrious and robbed them of the energy Beethoven intended for them to have. The point isn't to hit a specific tempo as though it were a click track, but that the metronome notation conveys very important information about how the piece should feel when it's played.
th-cam.com/video/L_bd1J98jNQ/w-d-xo.html
@@jonathangilligan6019 hard agree. I actually believe there's also pieces that didn't have a metronome marking initially that are misplayed, e.g. Moonlight Sonata mv. 1 (there's no way that when he wrote "adagio sostenuto" he meant the almost "grave" that people play nowadays)
Perfect rhythm does exist like perfect pitch, ya know?
04:30 you might want to visit Switzerland. Here every minute is important when telling the time. So if you ask someone what time is it hey mostly reply with the exact minute.
It also depends wich part in Switzerland. People living in the mountains usually round up time too because it is easier and they don’t need to do everything by the minute.
Why is every minute important?
@@Anonymous-df8it because time is money. If the train is a few minuted late then chances are high you won’t catch the next one. Here trains are very punctual and that’s why we have only small window to go from one track to another and it works very well, like a Clockwork.
@@Nanomaroni
😂 Okay. No. Switzerland is nowhere near Japan in it's punctuality and nowhere on the same planet as America in it's economic significance.
You don't NEED to tell the exact minute. It's far more likely a quirk of the language.
@@Kloppin4H0rses aint no way bro just corrected a swedish person about their own culture
Polyrhythms get the clicks - best description of “the algorithm” ever. Great video (from a DJ who started without software and decimal tempo displays).
When it comes to rhythm games, typically the size of the timing window to hit each note at the highest accuracy of judgement (e.g. hitting 'Perfects') is around 30 ms. The timing window for Dance Dance Revolution, however, is (crazily enough) half that at 15 ms! What is crazier still is that there are players out there who strive to clear songs hitting every note precisely to within that 15 ms window: the coveted "Marvelous Full Combo". In fact, top DDR player iamchris4life in particular is well-known within the rhythm dance game community for his quest to get Marvelous Full Combos on as many songs as he can.
This video does bring this premise of hitting notes super-precisely into another perspective for me, since having predetermined rhythmic challenges for players to attempt to exactly recreate while getting scored on their accuracy to such an insane degree feels almost robotic and absolute. Even swung notes are precisely laid out for the player to try to hit exactly, laid out in a way not too dissimilar to the precise transcriptions from these music social media posts.
I love the inherent contradiction in that. Can you groove/swing *exactly*.
Alright I suppose that’s also grooving/swinging *like someone else*, which is both what makes sounding exactly like other musicians hard, but is all the more impressive when someone can do that for multiple performers.
recommend replacing "however, is crazily enough" with "crazily enough, is". i understand "however" is a transition word which is used to express contrast, but "crazily enough" also serves more or less the same purpose. as it is now, it'd be drowning in commas as, like, "...Revolution, however, is, crazily enough, half that at 15 ms!". might be salvageable with parenthesis instead like "...Revolution, however, is (crazily enough) half that at 15 ms!" but to me it feels unnecessary for something to be however and also crazily enough.
anyway i agree with you
@@badmanjones179 haha thanks for the tip
I was actually low-key agonizing about the commas there myself while writing it
@@bitodd Tracks written for rhythm games seldom stray from EDM precisely because of that. Staying on beat is the whole game, and if, say, there's a song with a 6-6-7 hypertuplets, that would be the focus of the chart. Live music usually cause the "in-game BPM" to swing meaninglessly in a way that's frustrating to play. (For example, the song ってゐ！ in SDVX is marked 314-340 with 12 bpm changes that doesn't even feel like it's changing but might mess up the scroll speed.)
If there's one thing that seasoned rhythm gamers are definitely better than musicians with no rhythm game experience at, it would be hitting that swing within ±16ms every single time.
@@TheLaxOne But can you place the commas super-precisely to exactly recreate that sentence??
As a battle DJ, editing battle samples for performance is very useful, and one thing you can do is repeat a sample over multiple rotations of the record to get "skipless samples", so if your needle jumps, you land on the same sample and not get a skip.
Turntables spin at 33 1/3 RPM, so you can multiply this by integers to get a tempo at which you can place that number of beats into a revolution. The problem comes with the 1/3, or .333..., which makes fitting 3 samples at 99.99 BPM drift right as you go further into the record, and doing it at 100 BPM makes them drift left. It is a nightmare.
"As a battle DJ" might be one of the coolest introductions i've ever read
You made a great comment here because you were describing not just a theoretical need for decimal tempos, but a material, practical, demonstrable need for them, and not even just to the second place, but to many beyond that!
@@duon44 it really is a cool side of music manipulation. I recommend looking into the likes of Mix Master Mike, DJ Craze, the X-Ecutioners, or the many DMC and Threestyle champs for a good dive into it.
@@duon44 I've never heard of a battle DJ but just from reading it I can tell that's the coolest description.
14:00 and the recipe analogy are things I really needed to hear lol
I’m someone who both because of my brain and because I so thoroughly learned percussion in band, struggle with “properly” learning music and wanting to play/make music because I’m so focused on playing notes “on time” at the precisely written time above all else. The few times I’ve messed with music (beyond percussion) so far, I did best when I just kinda felt it and played by ear, not worried about perfect timing and not (as) worried about all correct notes
The recipe analogy really made it click cause I’m someone who really likes cooking and constantly switches between following recipes to a T and throwing a bunch of my own stuff in there to see if that’ll make it better
A point that is implied is readability. Since the score is usually given to a performer (recipe analogy), a necessary tradeoff between readability and accuracy goes into a well made score.
As a music producer that makes their tunes by feel first, I have my tempos at decimals because I just tap tempo it in and leave it as is. Even though I could just correct the tempo to the nearest integer and it wouldn't be intelligible, it's left as it is because it still somehow feels different / "wrong" to what I envisioned for it to feel like.
I was going to say the same thing!
9:23 One could say, then, that the notation for the audience is not the recipe, but rather the table of contents, which is often calculated. It is meaningless the the performer when the performer improvises or when music is made like the music teacher example, but gives the audience a "fun" view into the mechanics of the music (or the food)
As someone who’s not that savvy with daws but likes making mashups now and again, I cannot stress how frustrating decimal bpms are. I speed up song A slightly so it syncs with song B. All is good for 3 bars, then I hear it. The de-sync. It only grows from there. Okay, I’ll go back and adjust the speed again. 1bpm faster. Nope, it’s definitely off best this time. I can’t adjust it that finely, so I just have to scrap it since I have such a sensitive ear. Sometimes I wish it wasn’t so I could just ignore the de-sync but alas. Ignorance is bliss.
you understand😭😭😭 I’ve had to struggle syncing drum breaks and hearing them desync a few seconds into the song too many times, just to realise it’s because of those decimal bpms just now
@@emmycheeto303fucking relatable
On the rhythmic incompleteness of your bass solo transcription: That sort of "technically inaccurate" notation is fine if the performer can be reasonably expected to be familiar with stylistic conventions (or have access to an authoritative recording to supplement the notation). But part of composition is to innovate new styles (which wouldn't have established conventions) or incorporate influences from more obscure styles (which a performer should not be assumed to be familiar with). I imagine there are reasonable stylistic conventions which could simplify the notation of the opening of The Rite of Spring, for instance, but it's easier in a case like that to precisely notate than to communicate those conventions. If aimed at a performer who doesn't know the appropriate liberties to take, being explicit about landing on "e of 1" might have been more appropriate in your transcription. Similarly, lead sheets often have "quantized" versions of the melodies, the assumption being musicians familiar enough with the style to realize an accompaniment from a lead sheet will also be familiar enough to breathe life into the melody.
I am not sure if I am agreeing or disagreeing with you, but I would think that one would be better off notating music in the most basic form that fits the music and allowing performers to add their own stylistic feels and ornaments.
If taken to other examples, would you even precisely notate a swung rhythm? I would think not and adding a symbol for swing at the top better indicates how to play it, since the feel choice will not always be exact to how it could be notated, and it will also hinder attempts to play the same piece in other styles.
If I was playing Goldberg aria, seeing the mordents and trills notated is not helpful because you may not read it as that and instead see it as an exact thing to play, but I may play a trill with some rubato on some performances if it has the trill symbol and not precisely notated. The music would be so busy I would probably not bother to add any of my own ornaments and certainly would never cut down any written ornament.
Going by Adam's example of the semi-quaver rest, if I saw that notated I would expect that to indicate some sort of syncopation in the music or purposeful rhythmic choice to not land on one, rather than a laid back note.
Couldn't agree with you more. I think if Adam delayed the E by some very difficult to quantify rhythm (5/11ths of a beat for example), then I might understand not noting the rest. In this case however, he delayed it very clearly by a 16th, and so his explanation for why he didn't notate it is kind of a head scratcher. What's worse is that he has a similar pickup-into-long-note idea at the end of the solo, (see 10:35) but that note doesn't have a delay. Like you said, if you have a recording of the solo then it all makes sense, but I feel like the notation should be able to stand alone and get you as reasonably close to the music as you can get without needing a recording, or being a part of the same social circles as the performer. A 16th note rest is very well within the limits of music notation and is much, much simpler than any of the tik tok examples.
Then again, I'm not a working musician. Maybe it makes sense to other people.
I think your point ESSENTIALLY brings up a GREAT point not thoroughly addressed by Adam: there is a third, in-between, reason for transcription:
Adam's transcription makes PERFECT sense to people who are used to glancing at lead sheets for music that is often improvised in nature and going "oh yeah, make sure the end of the run lands thereabouts on the chord change."
But the "quantized" version is what is literally being played, whether Adam likes it or not. And, especially if learning originally improvised solos, the quantized version should be preferred if someone is learning or studying a solo. Is it busy to look at? Yes, but if I am trying to learn a piece of music (especially since I don't have a great ear) the literal/quantized version is so much more helpful...
Awhile ago I was learning Dogs by Pink Floyd, and, if you know what I am talking about, the first guitar solo's tone is engulfed with that deep chorusy, swirly Yamaha rotating speaker colour... To the point that many of the syncopated notes' attacks are distorted or washed away altogether to the point that it is hard to pinpoint what was originally played. It was only until I could find a transcription done by a college student that I was better able to figure out and wrap my head around the rhythm in that solo; up until that point, tabs (which are probably as equally useful as Adam's "c'mon, we know it is actually a LAID BACK half-note" notation) could not illustrate, to me at least, the NATURE of the syncopation. In this way, I would argue that super-accurate transcribing actually can become EDUCATIONAL and useful in teaching a style. For instance, today I learned that playing "laid back" bass is playing ends of quick runs that land on a chord change about a sixteenth or so behind the beat.
@@CHEWYCHEWYQQ I thought the same thing after watching that part! I am of the school and philosophy of using your ear and intuition to make music, but when notation is concerned, specificity is paramount. It's why myself and a lot of others feel that fake books are more or less useless... you're gonna have to listen to the tune anyways, so why even bother carrying around a pile of shoddy transcriptions?
@@mss11235 Agreed!
Decimal points in BPM are good if you are scoring a film (the music needs to line up exactly), but the sheet won't show the decimal point.
Sure, but click tracks exist. 3rd trombone doesn't really need to know the tempo is exactly 114.57bpm as long as the conductor is locked in.
The problem I see with that is you only really need that specificity for live performance of scores, which is probably not how most scores are recorded now. If you need stuff to sync exactly to an onscreen cue, you probably just slide the notes or samples in your DAW until they line up with the footage. Plus we're already assuming that a film doesn't really follow normal music tempos, so how can you be sure your decimal point BPM will work for longer than one 3-5 second shot? At a certain point, you just have to learn the timing of the visuals themselves and not consider metronomic tempo - films, unlike scores, are typically written in Free Time.
That's done with time signature changes, not tempo. If the tempo needs to be adjusted it's in post-production, not while composing.
Decimal tempo markings can be useful for Elliott-Carter style metrics modulations. If the old 5:3 eighth note is the new quarter note then decimals may be useful or necessary to be both musical and accurate.
yeah good point
this whole video reminded me of videotape being 154.78 bpm for no reason at all lmao.
I was faced with the dilemma you mentioned here when I was writing thousands of lead sheets for music publishers in the 1970's, mostly for copyright purposes. We were forced to fit contemporary musical performances into notation that was more suited to 1940's Tin Pan Alley, not to mention 18th Century convention. I often had to decide if I should notate the "laid back" feeling accurately or how it might best fit on paper, simplified. To do it accurately might mean an unnecessarily complex appearance but to write it in a simpler fashion meant I was second-guessing the composer's intention. 'Tis a puzzlement!
Score follower's work, to me, fits neatly into modern "effortposting", a phenomenon wherein people deliberately overdo the effort and/or precision of their work, as a joke, as a commentary on status and precision, or for any number of other reasons. You're right, they're going well beyond the limits of human perception, and that's the point, I don't think most people are confused about that.
Yeah, I'm with you. The fun is in the absurdity of the precision.
I think Adam has valuable and interesting things to say regarding value and use of precision or scoring in music to different people, but seemed to never really get into this point regarding these particular kinds of scores. In a way, this video ironically misses the point by focusing on the markings rather than the feel a hyperprecise score or maybe BPM might aim for. He's become the very thing he vowed to destroy 😔
"The notation becomes another way for the audience to engage with the sound of music". Never truer words were said on music-related channels, based on my personal anecdata at least :)
I am musically inept, can't play any instrument, can't catch rhythm, can't sing or dance. But boy I love to see visual cues when it comes to listening. I think it started with ProTracker visualisations back in Amiga days in early nineties, then of course Milkdrop and now all the Synthesia-like videos
I love the "7:42/7:45" analogy here. Reminds me of the Gricean maxim of quantity, as in his example: "If you are assisting me to mend a car, I expect your contribution to be neither more nor less than is required. If, for example, at a particular stage I need four screws, I expect you to hand me four, rather than two or six.". "Rounding up" seems to be human nature, in all aspects.
i like your perception of sheet music :) i think of sheet music as a story book, and me playing it is reading aloud. Eventually, i may no longer need the book to portray the story verbally anymore, and i can tell someone else the story and they can tell it after hearing me
15:45 "everybody knows that polyrhythms get the clicks"
best line of the video
Exactly!
Using the boom whacker thing as a record-scratch was SUCH a good idea!!
So here's a very specific use case:
Radiohead's Kid A has the peculiar quality that every song is (in my opinion) enhanced by playing back a second copy at a delay of exactly 17 seconds -- down to the hundredths of a second. The whole album is in canon with itself.
But, each song is in a different time signature, different phrase lengths, different melodic contours, different tempo. So the way that it's managed is that every song syncs to a certain number of beats displaced. For Idioteque, with its 5-bar 4/4 phrases (20 beats per cycle), it repeats at exactly 39 beats --> which works out to exactly 137.647 bpm. For Kid A, 17 seconds is 32 beats at 112. 941 bpm; for the National Anthem, its 26 beats at 91.647 bpm (I hear it in double-time though - 54 beats at 183.294 bpm); and for Everything in its Right Place, it's 35 beats at 112.941 bpm.
Interestingly, Kid A is the only one that lines up the 17 second delay with the cyclical phrase length (8 bars of 4/4).
The National Anthem repeats halfway through the 13th bar (4-bar phrases in 4/4), meaning the off-and-on syncopation in the baseline just becomes straight 16ths, with beats 1 and 3 cycling between: C-F# > F#-A > F[nat]-A > C-F[nat] over a D pedal. Plus, the sax parts form some crazy suspended harmonies at the end. Way too much to talk through in a youtube comment, but it all lines up in a really interesting way.
Bonus points in National Anthem for the lyrics -- the "it's holding on" at the end of the first verse ends up filling in a gap from earlier. "It's holding on" happens to sound a lot like "is all alone", so when both are layered on top of each other you get "Everyone [is all alone.] Everyone around here. Everyone is so clear."
For Idioteque, the percussion is where you hear it most clearly, though Thom Yorke does end up harmonizing with himself at several points. The bass drum starts on a three bar cycle, split into 6+6 beats, but when the chords come in, it shifts to a 5-bar phrase, where the first 6 beats are copied over from the 3-bar version with silence filling out the rest. The snare hits on a consistent beats 2 and 4, but when the canon enters, it now hits on all four beats of every bar. It avoids feeling too chaotic though, because that alternation between a steady bass drum groove and the 6-beats + silence sections make it so that it's never overwhelming -- it just has a little extra flavor every once in a while because the bass drum pattern is so irregularly syncopated. The chords on top have such a slow attack that you don't even notice that one is coming in half a beat early, and they've got just the tiniest hint of syncopation already, so if you're just listening to the chords, you won't be able to really follow where the actual downbeat is anyways.
And for Everything in its Right Place, the pattern is just a 2-bar repetition of 10/4 (at least, that's how I hear it), so a 35 beat delay puts you in half a bar early. This one also has some weird syncopation that creates interesting emergent patterns, but for me, the most evocative aspect of it -- totally unexpected, but obvious in retrospect -- is the way the vocal melody interweaves with itself. On its own, the structure for each bar is: Lyric > Long silence, gradually shifting into lyrics that fill the whole bar. But, because the canon is offset by half a bar, on the repeat you get Lyric A > Lyric B > Lyric A > Lyric B, without those gaps. It becomes a new song entirely, with a melody that, in theory should be the same, but in practice takes on an entirely different character:
Album version --
Everything [pause] (repeat 4x)
In its riiiiiiiiiiight place [pause] (repeat 4x)
Yesterday I woke up sucking a lemon (repeat 4x)
Kid 17 version --
Everything [pause] (repeat 4x)
Everything in its riiiiiight place (repeat 4x)
Yesterday I woke up sucking a lemon [in its riiiiight place] (repeat 4x)
---
And all this magic because Radiohead decided to use extraordinarily precise tempos.
Hmm what's more likely that Radiohead intended this nonsense or that Radiohead fans are turbonerds
@@rottingcorpse6565 it's not unheard of for a track that can loop on itself unintentionally, especially if the whole track is a repeated chord progression. But its exceedingly unlikely to be coincidence for an entire album to do it at exactly a 17-second delay, especially where the melodic lines from each song at that delay interweave seamlessly with each other. Only in a few instances is it Thom Yorke harmonizing with himself -- like in Kid A or parts of Idioteque. The rest of the time, the melodies all have space left in them that maps to that 17 seconds, so that the only things happening are a) direct harmonization, or b) emergent melodic lines from fragments played in canon. There are no instances where the counterpoint creates out of place harmonic or rhythmic dissonance -- which is not easy to do when you have a canon at the unison across all melodic, harmonic, and rhythmic components, where the second instance starts not just partway through the phrase, but on a weak subdivision of the midpoint of the phrase. That doesn't happen by accident. Not for an hour straight.
@@rottingcorpse6565 also, I wholeheartedly accept the term 'turbonerd'.
This is fascinating but it also highlights that an important factor here is just kind of semantical - precise decimals look cringey because of their format, but a lot if these examples are just decimal expressions of fractions and only look cringe because you are forcing everything into base 10. If you just kept the bpm as the original fraction it would look fine
@@jackdavinci They are still misleading. If I told you that one song at 100 bpm played 3 beats in the same amount of time as another song played 4 beats, that doesn't mean the second song was played at 133.333 bpm. It could well have been 132 bpm or something else close by. It's a rounding error. You can look at the audio from a specific performance and calculate exactly the average bpm (assuming a given time signature), but that's just an empirical average for one performance; it's not by design. It's rather like using a high-precision laboratory laser measuring device to measure the length of a track after a race and then declaring it a 99.99983125 meter dash.
As someone who has both played traditional instruments and done some DJ stuff, I think appropriate level of precision definitely does depend on what kind of music you're making and what you're trying to do rhythmically. If I'm playing the piano or guitar, I want to let it flow naturally and let it be slightly imperfect. I don't want it to sound "inorganic." On the other hand, precision is much more important when I'm making a remix with a looping instrumental track, especially when I'm working with electronic genres, hip hop, etc. It's still creative and artistic, but it's more like building mechanical/kinetic art pieces than painting a scene with oil on canvas.
I love Phonon's work, but i felt similar feelings as you did at the way his strange rhythms were given legitimacy by their ability to *technically* be notated in sheet music. Like when I hear Polyriddim I'm taken in by the shifting off-center wubs, but there's lots of electronic music that deals in more generative organic rhythms like that. The only innovation that Phonon brought was sheet-music novelty. (btw adam please listen to Confield you'll love it)
Fun fact about the Just Noticeable Difference: Weber's law states that it's proportional to the existing stimulus. That's why the metronome goes from 69 to 72 (a difference of 3), but then it goes from 138 to 144 (a difference of 6).
only mostly true. you can't treat metronome markings as whole numbers.
as stated in the video, 69bpm is one click every ~870ms, 72bpm is every ~833ms, a difference of ~37ms. the difference between 138 and 144 is the difference between 435ms and 417ms, which is actually only 18ms. but it's still difficult to tell between 138 and 152, which at a period 395ms gives a difference of 40.
This is related to the fact that half- and double-tempo relationships are privileged among the standard metronome markings. I imagine this came about to permit musicians a choice between setting the metronome to beat "in 4" or "in 2" over a range of tempi where either might make sense (especially for practice purposes)--and/or to allow for practice at exactly half speed.
Note that Weber's law was apparently modelled after testing people's perception of light intensity. It does not necessarily hold up for other things.
For example, it most likely does not for loudness or pitch difference.
@@Trirosmos is it because loudness behaves in an exponential way? I recall reading that a 10x increase in loud is perceived as 2x
@@ithinkpaulmightbehavinastr8878 Not only that, but our perception of loudness is really complicated overall. It also depends on the sound's frequency, for instance. Search for Equal-loudness contour.
Well, in terms of DJing, the point is also that even small differences will make the tracks run noticeably out of sync within a short amount of time. And you might not mist want to transition, you might want to leave them running in parallel for a while and play around with switching back and forth or mixing in different elements. It helps if you don't need to correct all the time.
Adam. Incredible way of explaining notation so simply. It reminds of that quote written by who knows who that says, "If you can explain something simply, you understand." You humor and intellect are prime my friend.
Absolutely amazing that you’re music lessons get me teary-eyed almost every time. 🥺
3:59 quick note, as someone that’s going to college to study meteorology next year, i normally wouldn’t measure the temperature down to the decimal point, however close to the 32 degree mark, extremely small differences can mean the difference between a rain shower and an ice storm. i wouldn’t make a distinction between 68.5F and 69F, but I would make it between 31.5F and 32F
Hearing "I got to the party at 7:42" just makes me think of reports from investigations of deadly industrial accidents, where the timestamps increase in precision as you get closer to the explosion (or the meltdown or fire or toxic release or crash or whatever else). When the report tells you what an employee said at some exact second, you know something terrible is extremely imminent.
Das ist sehr interessant. Dankeschön und alles gute weiterhin ☀️
Love this, and Adam’s editing quality is absurdly good. I could listen to this man talk about anything.
A thought that this brings to my mind is ritardandos and accelerandos (sorry if I spelt either wrong). You could, technically, write a new tempo for each note (which could get very complex if you have a complex rhythm happening) instead of writing “rit.” Just thought that was another cool musical example of this idea Adam was talking about.
You can't say ritardando anymore
I recently imported a MusicXML file of mine across softwares and had exactly this happen, instead of the ritardando I had notated there were about 32 new tempo markings within the one bar 😅 I guess that's how a notation software generates a ritardando for MIDI playback
@@tonywilliamspiano One interesting example of this is the 2nd movement from Thomas Ades’ piano concerto. He notated the ritardando with subdivided rhythms, because otherwise in this case it’d be difficult to put orchestral musicians in one exact way of rubato. I recommend you to have a listen to the beginning, guess how it’s actually written down, then check the score (the preview of it is available on Faber Music website).
There's an aspect I'm missing here - the fact that the human conscious isn't able to pinpoint a subtle difference from a certain point, does by far not mean that that difference doesn't affect the subconscious, which in turn alters your perception a lot. As in, the fact that you hear a song in a tempo that you have never before heard a song in. It's not too different of a discussion from different keys or tunings to people without perfect pitch, where they couldn't tell the difference without direct comparison
I would be very surprised if there hadn't been studies made to find out if people without perfect pitch "felt" different about imperceivably small tuning changes. My guess is there's no there there
100%. Humans can’t differentiate between .01 and .02 inch lines either, but 4K video still looks noticeably better than 1080p.
George Collier is great because the people in the videos aren’t exactly playing what was transcribed but the transcription sums up what you hear really well
I once used a 160.1 tempo because I thought it'd be funny. It was until I forgot the bpm and had to find it all over again
Hey man love your exposition skills. You're a great orator and really good at deconstructing topics into concise digestible segments (you seem to have a good understanding of the perspective of the listener/viewer)! Also I love your music bro keep it up you're an S-tier creative imo!
In experimental sciences there is a similar concept called "significant figures". If your measurement is only accurate to the ms, don't write down the microseconds. It's not just useless, it's misleading because someone reading the data could assume the accuracy of the numbers based on that.
It is also very important when calculating the error bars.
I refer to this as spurious accuracy, but it is apparently more often called False precision.
You often see it when people pass numbers through formulas, even something as simple as an average. Averaging three numbers might end up with an answer ending in .3, and that's fine, but ending it in .333, when it was a whole number to begin with, sets my teeth on edge.
@@Alan_Duval I made a similar post regarding designed dimensions. Random sizes given tight tolerances.
@@Alan_Duval typically called false precision, i imagine, because a sometimes used definition of precision is the number of significant figures or somewhat equivalently the ratio of the max (random) error to the whole (reported) value, where accuracy is whether or not the reported value equals, up to the given precision/sig figs, the actual value (for intrinsic/systematic error).
Thus, if you give too many digits, that may not affect the accuracy much (small places being wrong doesn’t matter as much as large places being wrong) but it implies your random error is way lower than it actually is, i.e. false precision.
Scientists, mathematicians, engineers, programmers, and technologists - at least those who learned and practice their discipline properly - are very careful to not confuse precision with accuracy.
But artists usually lack this understanding. And usually don't need to know it anyhow. Sometimes high precision or high accuracy are critical, some forms of creative expression require it. But mostly it makes the art more complicated without adding any meaning or any quality.
This is why I find it funny, and a bit sad, when a trail to the lake is marked as 7 miles or 11.265 km long, or a radio's range is stated as 492 ft (obviously translated from rough estimate of 150m). The former is becoming a bit less common, tho - maybe they got tired of fitting all those decimals on the sign.
Part of the appeal of transcription videos is that it's a sort of subtitles for music. To an untrained ear, it might be hard to pick apart a piece of music to understand why it sounds the way it does, but with the transcription on screen, it's harder to miss something (as long as you understand it), and it gives you an amount of mental advanced notice for a sound you're about to hear so you know to be listening for it.
Decimals are very useful in gradual changes in tempo. they are also useful in soundtracks when you want the music to go well with the scene. Also sampling something might be easier if you have decimals in your tempo
he makes these points in the video
For gradual changes in tempo, you're better off writing accelerando/rallentando and just marking the initial and final tempos. For your last point, he's not talking about the tempo you put in a DAW, he's talking about tempo in sheet music notation.
You don’t write out decimal tempos in soundtrack scores my guy.
I don't know the first thing about music theory, but as someone with a linguistics background, this is very relatable. A lot of phonetics is literally played by ear until you reach the applied science (read: voice recognition), and even then, personal biases come into play quite a bit. Also, anyone transcribing in narrow notation is either trying to make an academic point or is being unhelpfully specific.
I think I agree with everything Adam said in this video! I have thought a lot about the relationship of western sheet music with sounds that are being retrospectively transcribed. I have worked as a transcriber for Michael Paouris for his contemporary pieces for greek bouzouki and we had discussed many times about how I should write some sounds that had never been written on sheet music. My humble opinion, that I have formed through this work and also my TH-cam video transcription, is that most times the sheet music should serve the sounds-music and not the other way around. I would love to hear some other opinion if you disagree.
Have a good day!
I tend to use Decimal BPMs when I make drum covers, specially those that were not recorded at a metronome beat, like the old time Sonata Arctica, where I need to more or less match the "whatever" feel they were recording to
The first music I ever made were beats on an MPC 1000 in my high school because one of the interns was big into hip hop. He told me to ALWAYS set your bpm to a decimal number so that if a dj is using it in a set they have to work for it instead of just putting two songs together they know are the same. Of course, with bpm match on CDJs nowadays that doesn’t matter but I still do it just for fun!
I now tune my BPM with the key of the track. For instance, if I compose in A minor, my BPM will be 103,125 ( 440 HZ dropped 8 octaves). With a saw wave, you can clearly hear each click of the saw is in time. This allows to move from rythm to harmony witth ease. In this case decimals are useful.
smart
Agreed, but for a different reason: I am more musically fascinated by co-prime relationships. For instance, in 120 BPM, I may set one chorus effect to 113 and an identical chorus to 131; the co-prime relationship to the tempo can add an organic feel to improvised music.
Not saying your way is not useful, as it completely is, I just wanted to expand on your point to include that your reasoning used in an opposite context could be useful for a somewhat intuitive reason.
interesting comment!
Precision is for computers.
Western notation is for humans.
When you cross between them, wonky results can occur.
I feel like this sums it up perfectly. You might need to tell a DAW to run at 164.31 BPM to synchronize with a recording, but a musician can match the tempo on their own - and if they're not playing with a recording, then something like "Fast" might be all the advice they need.
Another thing which isn't really about musicality, but about giving access to a producer who would like to use your music:
If a song is following a metronome, but the BPM isn't a whole number, if I wanna manipulate it in a DAW I would like to match its BPM.
If the BPM isn't a whole number, I'm gonna have to go to audacity, pick two beginnings of a measure, count the number of measures, and calculate the BPM. Then, the more precise I got the more measures it will take for a desyncing to occur, so then I have to fudge with 2nd or 3rd decimal place like an idiot.
Obviously if the original recording wasn't closely tied to a metronome, or three song changes tempo frequently then this is unavoidable, but if you are controlling the BPM, then be considerate.
Musically you won't be able to tell the difference between 135bpm and 135.269, but the poor sod that would want to manipulate it would
How good is this guy’s content that despite talking deep musical topics to the point of making them almost philosophical he’s got 1’6 million subs. Hats off Mr. Neely, hats off