I know we want to minimize the parts count, but the reader really needs a roller where it contacts the programming profile. It will save wear and be quieter. (Otherwise, won't certain notes will play slightly later after each show?)
saved my typing it in tks, oh but you don't mean in a 'with a roller there is no friction to require extra mechanisms' and just mean use both a roller and a latch mechanism lol
Martin if you read this - this is one of the best video's you've made so far in the MM's lifespan. There's not a lot of flair and polish - but there's joy. There's so much play and fun baked into this and it's great to get to see you working in a way that gets you excited about your art. Thanks for this today - we are behind you, thank you for your years of enthusiasm.
Agreed! I’ve been falling asleep looking at spreadsheets on this channel, I even stopped watching completely for awhile. So happy we’re going back to the original format!
Martin kinda fell for the same problem on MMX. Every solution should have the option to be as silent as possible, even if it sounds cool in isolation. Maybe he could take out certain sound dampening based on the song?
Thought I was the only one who thought this way. The mechanical sounds, especially when rhythmic like these machines, can actually enhance the entertainment experience for, in my opinion. It feels like something is building up, signaling that something exciting is about to occur. Almost like a metronome used as a countdown.
@@livefromhollywood194 I know your right too.. but I do love this ticking sound. But if it is creating noise in the wrong spot so it may detract. Such a shame bc I also love the ticking!
@@NobleSaintedYes. But Martin can always place a microphone directly next to something that makes a little noise. He can't remove noise from something that makes a lot of noise that gets picked up by the instrument microphones. And all the little noises will add up. If I remember correctly, he said that at some point during the MMX.
Martin, I just got a full-ride scholarship to an American college because of *you.* In my application essay, I talked about how you inspired me and the judges loved it. I've been watching you for years. Your videos, your methods, and your community have inspired me to be an engineer myself. Your experimentation, optimism, and innovation have inspired me more that you can imagine. You have literally changed the course of my life, and I can't thank you enough.
That’s amazing! I unlocked a new world, after seeing demons manifest in my room. Messing with ouija boards and seeing them come alive, sleep paralysis to many demonic experiences. I realized we live in a spiritual world more than a physical one. There is a God and His name is Jesus. Try Him with a willing heart and mind and what happens next in life is mind blowing. More than an acceptance to an Ivy League. If we harden our hearts towards His presence today, then why would we expect tomorrow to be any different?
In regards to the mechanical ticking, I actually appreciated the mechanical background noise of the first two marble machines, it gave the music a distinct sound profile from regular music. If the ticking is in association with the programming wheel then the noise should, in its own unique way, be tied directly to music and might even sound intentional
You've discovered how to integrate mechanical memory into Marble Machine - essentially MM now has a state machine - so each 'track' can be held in a number of defined states in readiness for performing with a much reduced 'noise-floor' & with much less friction. Elegant solutions are SUCH a joy. Great stuff!!!
If friction on the programming wheel interface is a big concern (500 channels, wow!), wouldn't a good idea be to use a roller cam follower on the readers, instead of the sliding contact you have now? In addition to reducing noise and mechanical load, it should drastically improve wear concerns, and keep the machine running tight with age. Even with this new design, you're still going to have friction on the actual rise of the profiles, which is still going to add up quite a bit over tons of channels. The trailing edge of the profile can be curved to match the arch of the falling follower and allow a sudden drop if that's the reason for the current knife follower. Also, for high load components like the bass fingers, would suggest looking into ways to build lots of leverage/mechanical advantage into the system, so that the load felt at the wheel is still relatively low.
It looks like the idea of reducing friction should be separated from the new toggle functionality. Rollers, of course, some obvious upsides, no downsides. But the capability to toggle the mute on or off, while still being able to put beats when needed is attractive on it's own.
@@CygnusXUnoyes there could be cases where the latching mechanism could be great. But in this case it was used to eliminate friction. You can achieve this goal with rollers/ball bearings at the tip of the reader way easier and far more efficient and have other benefits as stated above.
Have you considered using only a one or two string bass? Since hand size isn’t a factor, intervals aren’t either. Generally you’ll only have one or two strings ringing at a time. So this could save a lot of space and parts.
Why stop there? If you turn the string into some cone-shaped cardboard, it can play multiple frequencies simply by moving it at different rates. Maybe we could achieve that with electromagnets?
i think it should be possible to have a bowden cable instead of a mine retracting and releasing -> smaller form factor and could be activated by a marble hitting it
Realistically, he should be getting all these parts done in metal for the final machine. The 3D prints are the absolute best for prototyping like this, but I doubt they would hold up to the rigours of a tour, and having to check them before every show would be a huge time waster. E.g. I think the middle nub in the miniature versions of the latch channel would wear out quick, those will probably have to be made in steel. Or the connectors to the tensioning screw: I will assume Martin knows to orient the layers correctly, but even then they might fail under the constant cycles of tension and relaxation (cycles are often what kills parts). Martin is being smart about this. He will almost certainly need to tweak the geometry of the latch: angles, length, depth of the track, etc. Then add a roller tip to further reduce friction and wear, verify that still works with the geometry. All this can be done rapidly with 3D printed parts. Then version 0.9 will be made in metal and with roller tips, tested again, and finally tweaked a bit (or not if he is lucky) for version 1.0. Im sure once it is integrated into a fully functioning machine, there might need to be a v. 1.1 on some of the parts as integration and long term issues pop up. Best of luck Martin! There is still a long road ahead, but as they say "How do you eat an elephant? One bite at a time."
@@ParanoidMarvinMk2 or maybe he can get away with some fancier filament like nylon and carbon fiber reinforced filaments, etc... but just regular PLA is not really good for this job.
@@ParanoidMarvinMk2The comment wasn't about the wear and tear of the latch but for the programming wheel and its mechanism. I fear that this is a typical martin solution. You solve a problem by creating another (more complex) problem
@@biziluxgames8924 if you watch, he actually is NOT even orienting the layers correctly. The connectors for the tensioning bolts have their layers perpendicular to the stress, not parallel. They will begin to delaminate under any sustained use.
This is a great improvement! Small suggestion though. Consider the case where you unmute a channel right before a 24mm programming wheel profile, and the channel has not been placed in the locked open position. In this case, the profile shown could cause a machine crash when playing at high tempos as the pin would get stuck in the first section of the profile. I would suggest changing the angle of the outside edge of the profile to allow the pin to slide cleanly from the closed position to the fully open position in this case.
Exactly what I was thinking, ToT profile was independent of speed, this one has a race condition built in, hit it to fast with un-latch and it will crash
That's what I was thinking. He could build a second escape channel on the far left of the profile that allows the unlatching profile to fall through, but is to far left for the latching profile to reach it.
If you look closely on the small profile, the shape is a bit different. The back edge trails directly towards unlatch, instead of going forward towards latch as it does in the large profile. With the small profile, this issue is handled. The thing that might cause a worry to me is that the latching is done by gravity, so I would believe it might not latch properly in certain cases.
honestly I don't see the click as a minus - Many instruments have noises they make alongside their notes that aren't part of the note, but become part of the instrument's identity. The original Marble Machine (and the MMX) drew a lot of their unique sound from the rhythmic click-clacking of all those parts moving about, and I think the soft click with every movement of a marble gate will have the same effect here
Absolutely! The first Marble Machine starts and ends with the lights turning on, the song begins with the winding of the marble machine, and the latch dropping the first marbles is the very first "note"
The sound of the machine is part of its identity. It's understandable to reduce it so you can hear the music, but getting rid of it entirely removes some soul out of the marble machine.
I love your watching you enthusiasm and engineering process. I've been watching since the first Marble Machine and I'm so glad you're still going strong. Keep up the good work!
If a note is not played the entirety of a song, it would need a profile the entire diameter of the wheel (to keep that “finger” lifted). This is a set of on off profiles with a hold function, so one off profile covers it instead.
This will probably get buried, but you are using a shift cable in a brake housing; shift and brake cables are slightly different diameters, and so is the internal tube from the housing, it may work better and have slightly less slop in the system by using shift housing. Also as a secondary benefit, shift housing is a non-compressible housing, which should allow for better indexing for precise movements. The friction may increase slightly in a shift cable housing, but for very light loads you can push the cable with moderately more precision than the brake housing. Always love the content and keep up the good work!
RE: Bass Levers and weight -- double the distance from the fulcrum from the spring on the pull down side will halve the weight quadruple the distance from the spring side will quarter the weight... etc. Round figures.
re: the ticking sound that it added. You're going to put microphones on the sounds you want - you won't mic up the ticking noise. No issue with the ticking.
I love the whimsy that you have in this video. It is the most cheerful and excited you have been in a while. It’s like you’re the chef in “The Menu” and you just made a cheeseburger. Iykyk
Playful and excited Martin is back! So happy to see it! He was in there all along and I’ve enjoyed everything you’ve been doing, but seeing you happy brings me so much joy. The reason I watch your channel is I enjoy seeing you try, try and try again on something that is incredibly difficult. You’ve got this
this is cool, you've created a mechanical set reset latch, you can think of it in terms of setting and resetting whatever actuator its controlling between its active and passive position
Options to reduce the contact noise of the readers/registers: 1. A small roller bearing/wheel to turn your scraping into rolling, eliminating wear and remaining noise. 2. Coat or make the programming profiles/registers from Teflon/a similar low friction wear resistant material. Another alternative thats pretty machineable is Acetal Copoly (aka delrin). 3. Introduce a cushion area in the track where the new latching profile is clicking. Optionally, consider a dry lubricant in the groove - a passing thought is finely ground graphite. Just my two cents i wanted to voice, if its of any use to you Martin! Great work thus far!
For the necklace, you should make a heart shape out of that new latch pin channel design by getting two of them and sticking them together mirrored with one another. This new design looks pretty promising!
Could I suggest using piano bass damper felt for your string mute? It's V cut may give more surface area contact and it would last a few decades longer than your current material. If you use a whole piano bass damper assembly you would have further movement to fit flexibly, although with your plastic arm that probably already exists.
I think the toggle is a great Idea but adding it for every channel will introduce more wear on small parts which will be harder to replace. Adding a roller bearing as other suggested will reduce friction and reduce wear. I would cycle the latch until it breaks. Love the video!!! keep going!!!
Love the gleeful joy you have, it's infectious. With regards to the "fingers", moving the fulcrum closer to the fingertips would require less force on the wheel to press them down. Then of course the travel length of the tips would shorten but as long as they completely disengage the strings everything is fine. Now you are a lot smarter than I am so you probably have the optimal dimensions for the fingers but I thought I would give my 2 cents. Keep up the awesome work! Can't wait for the finished product.
I don't like being negative, especially in the face of such enthusiasm and optimism, but given that he never got either of the previous two machines to work as such and he's now moved onto this mega machine with hundreds of channels... Personally I think that the best approach would have been to build a simple 1 or 2 octave basic marble vibraphone and get it working. Then start looking at bells and whistles. Stephenson didn't start by building Mallard, the Wright brothers didn't start with Concorde and NASA didn't send their first rocket to the moon...
17:14 - the best advice for those of us who try and solve all the problems before we start. I don't know why this popped into my feed, but it is the advice I needed right now.
I am excited to see all three parts(mute, finger, and dropper) come together to play a note in your test rig. The string instruments in your machines always seem to bring out such creativity and passion!
i feel like the mechanic fingers would be cooler being pressed by hand, like a piano keyboard. This makes the bass a hybrid instrument plucking the note with the marbles but hammering fingers on the fret with a keyboard played by martin. With that bike cable you can transmit the mechanical force far away, so put it next where you are playing! An automatic machine is cool, but a hybrid one is even cooler, especially for concerts (human touch, yay!)
"...Mechanical latch..." "...We have a slide!" "...don't wait for inspiration..." Martin's entire character growth; music, engineering, and philosophy woven into one seemingly adlibbed video. What an incredible time we're at in the journey!
This is such a great and underated video. You are a very intelligent man. This is what technology was supposed to be used for. Coming up with ideas to solve problems, then rapid prototyping to see what works and what doesn't. Over design, under design, redesign, and come up with something really cool, it's fun. And now anyone can take this idea and come up with something of their own. Brilliant!
”Ooo oohh…. ÅÅÅÅÅÅÅÅÅÅÅ SLAJDAR (🇸🇪) OHHHH YES!….det, uhm, I can slide, OK I was so nervous I couldn’t slide, listen to this That is soo cool, I can slide We are going to be able to make that BOW choopa khugha wheeooooow Maybe not……yeah, whaaaooww. WHAAAAAAAAAAAAAAAOOOWW!! Ok, this makes me really happy WWOOOOOOOOOWWW!! This is going to be the BEST mechanical instrument ever built, ok. Like…, EASILY!” This section is TH-cam gold right there 😂😂🤣
I'm so excited for this! As far as the clicking sound, its not a big deal at all. As some others have said, adding a roller at the contact point will reduce wear. Given that it is on a bowline cable, you can move the clicking mechanism as far away as you want in a soundproof box. Also, I love how you are concerned with clicking now, after you designed and built the fish stair marble lift for the MMX. I can't wait for what is next!
6:00 I'm happy to see that you use an improved version of the variant on the neclace. Meaning it can push through the mecanism when hitting the large profile from the resting position. ❤
Print the corners where the pins hit with TPU inserts. Also eventually cnc the part from teflon or some other slidy stuff of choice. Congrats Martin, I'm happy to see you smile again.
17:20 your comments about just doing reminds me of the aperiodic monotile, the mathematicians kept imagining but some random guy started doing and stumbled upon the answer!
I can see one small quirk remaining: The necessary force needed to push the mechanism past the disengagement (and the other) thresholds hasn't changed. You have minimized the time, but the friction is still momentarily present and will probably grind down your profiles and cost you energy per state-change/note played. But you need the force at the instrument to operate it, just not at the wheel... Idea: Get the grandfather clock anchors back in! Have weights that exert additional force to each finger, _but only in the 'down' position_ Operation: When a 'lift and lock' command (medium profile) comes from the wheel, have it first disengage the weight, then lift the arm. When the command to 'press' (high profile) comes, connect the weights to the fingers as soon as, but only after, you pass the threshold so they do the work, *meaning you can use softer springs!* When a single 'lift momentarily' command (lowest profile) comes - do nothing with the weight - I guess that quick response is more important in that case and low profile means low strain on the programming wheel anyway, so it is an acceptable tradeoff. Integrate this control mechanism into "the shape" as well, so it all works regardless of BPM or state of global 'mute instrument' lever.
Love the concept, it is so elegant! A small issue I see is: Let's say the mechanism latches in the middle state with a mid size block, and you mute it. Then, a tall block comes, which is supposed to get the muting arm down on the string. However, since the mechanism is muted, the tall block does not engage, and the mechanism stays in the middle state. So when you unmute, suddenly you have a string that is supposed to be muted, which isnt. So hard to convey in a TH-cam comment but hope you get my drift?
That spring tension mechanism is used in most drum kick pedals! The Mapex Falcoln has a niftier version where you don't need to pull on the string to adjust it, which is quite useful with stronger springs.
I can see why you're so excited about this! I was commenting last video about some ideas to make the bass play more human like ... and this implaments one of the things I sugested, but with a whole other dimention of possabilities with the latching mechanism you've created. Nice!
This is the way to go. You can search single parts extremes and take all the time you want. Sometimes you might think its too much, but then again you explore what is possible and whats not. Then decide how far you want to go with the design. Also with this method if you find something new you dont have the whole machine built. You can easily re-iterate the old designs and prototype the new version individually, without actually have built anything.
I had made a post about adding a preload spring to the activating side of the gate, but I realize that would just reduce the tension being used to hold down the finger, which you don't want. BUT, there is another way to reduce the friction on the programming wheel: leverage. The higher the resistance of the "instrument", the more leverage you need to overcome that friction without putting excess downward force on the programming wheel. What you would need is an adjustable pivot arm on the gate, with an adjustable pivot position, or just different sets of them, since their locations wouldn't need to change once the machine is built. The higher the pressure, the longer you'd want the lower half of the arm to be vs the upper half. The trick would be that the programming profiles would need to be taller, since changing the arm lengths would reduce the travel. That's really how a lever works anyway, it spreads out the force required over a longer distance (and time). There's clearly a limit to this, since you only have so much time for the profile to push up on the lever and activate the gate, and making the profile steeper will counteract the longer lever, again putting more force on the programming wheel, only now you'd be trading downward pressure for rearward pressure, which is almost certainly worse. But, for things like the bass fingers, unless the tempo is REALLY fast, you probably have a fair amount of time after the pressure is initially released to get the gate to the locked open position. Obviously, you could do the same thing on the bass fingers, using as long of a finger pivot arm as possible. There will probably be construction limitations of that, plus you need some minimum distance to lift the finger, and it would increase the preload tension required on the spring to hold up a longer lever while keeping enough tension on the Bowden cable. Carbon fiber finger pivot arms perhaps?
One of the best "non workshop" tools I've ever purchase was a couple of different sizes of Crocodile Forceps - easily purchased from medical supply websites. They're AMAZING for retrieving lost items inside guitars, reaching fiddly small things inside of complex mechanisms, or merely holding something small and fragile securely with a variable grasping force
Normally I'm a stickler for those cute little mechanical noises, but I'm with Martin on this one. That scraping is wearing out parts on a touring instrument, and it's an unnecessary expense. Great design!! Oh, also I think lengthening the top channel could help keep the channel from being damaged by the "unlatch" pin when it's not already latched - like just a place for it to glide to that won't hurt the mechanism.
I love the mechanical ticking of the machine. It makes it part of the melody. With the very first machine video, every single tick, rolling sound, falling marble, it is all part of the melody. Yes it wasn't perfect, and this is what people loved.
I'd suggest you should check the force required to lift that reader up the ramps - the friction of 500 channels rubbing the profiles can be reduced with the right material selection, but the force required to lift that profile against all the return springs, bowden cable friction etc doesn't change - so the latch might actually on average mean more force required than rubbing on most channels of the machine. The latch is rather neat, clearly should be used on some instrument parts that will actually want to stay latched down for a prolonged time, but I suspect the marble droppers are better off being allowed to rub - you are only rubbing for the duration the note is primed but not released yet, a time you can make almost arbitrarily small by changing the profile lengths.
That's a huge eureka moment! Goes to show how much effort you and everyone involved put into this beautiful project ❤ Maybe a fan that's a jeweler can make you a more practical, robust version of that necklace for your day to day use. Or get together with a designer and sell them to grow your project funds if needed. People could wear them as a token of gratitute and a reminder, that there is a solution to everything, we just have not found them all, yet!
I think the note on note off system is amazing for the cyber bass but I would argue that it's not very good for the marbel drops since the marbel is gonna fall anyway no matter the length of the profile, and the latching and unlatching could prevent you from playing faster repeated notes
programming wheel and the finger riding it reminds me of a camshaft in a car, overhead cams to be precise, there the spring tension is extreme and oil is used to keep friction low. If the profile of the programming pins made sure there is no "impact" but rather the finger is pushed up, friction and noise could be reduced even more. So, if the leading edge of the pins was convex instead of concave, for example.
I strongly suggest you to watch the channel from the beginning, it's a wild ride indeed, I follow this guy closely since 2016. Martin Molin is the biggest brag Sweden has =) Also check out Wintergatan's music, it is SO cool, you might like it
Not sure what I just bumped into. But it’s great! What I love with TH-cam is that it shows me great material that I did not search for - made by complete nerds who put a lot of effort into solving fascinating problems. 🙂 Subscribed.
Real Engineering recently released a video where they interviewed an engineer in an aeronautics startup and one thing that stuck out for me is that engineers are creative people who are good with numbers. I had to think of Martin with his struggle to balance the creative and engineering side of the MM. I'm glad to see this project is gaining traction again and going in the right direction!
I'm really excited for this new MM chapter. Your new approach seems really promising and exciting. With your plans encompassing so many features into a single machine, have you considered applying these techniques to designing and building individual smaller marble machines first (maybe a designated bass playing machine, a vibraphone machine, a drum machine, etc. ) and eventually using what you have learned to design your magnum opus Marble Machine. I feel like this would allow for faster turnarounds on smaller projects, and a series tangible and functional finished machines, rather than trying to squeeze every feature into a single machine straight away.
oh man this so genuine and its literally art and engineering but fun! And there's even some nice mindset advice that I wish everyone (and me some time ago) agreed with, I love this channel so much :D
Not to rain on your parade, Martin, but the magic-shape latch interfares with the self-latching override hooks (I'm making theese terms up as I type this, hope it's legible) causing droppers/dempers remain in the state they where in when their programming channel was muted instead of finishing their note and then getting disabled. Not to worry though, switching the order of the components (programming wheel reader -> magic-shape latch -> muting hooks -> dropper/damper/finger) will restore that finish-but-don't-start-anything-new functinallity. No new components needed! Fantastiskt jobbat med basen och lycka till!
The "giant" differs from the "small" shape in an important manner. The small one allows the tall wedge to move all the way to the end; the giant sample requires that to pause. (The difference between a click-pen which must latch, and a half-press/full press choice.)
It's a joy to see how happy you are about how dynamic the bass can now sound.
'''Bwaowmm Chikaboonga Vwoooom'''🎵🎶
"weeeeaaaAAAOOOooww"
I love a dynamic bass, but I see the feature creep monster around the corner.
so awesome, love how he enjoys what he is doing.
One thing that struck me since a few weeks Martin is that you're having fun again with this. That is absolutely awesome to see.
Yes this absolutely. The spark is back
This! 🤩
struck?
edit: Yeah. It's great.
what do you mean he somehow didn't before?
Exactly
I know we want to minimize the parts count, but the reader really needs a roller where it contacts the programming profile. It will save wear and be quieter. (Otherwise, won't certain notes will play slightly later after each show?)
saved my typing it in tks, oh but you don't mean in a 'with a roller there is no friction to require extra mechanisms' and just mean use both a roller and a latch mechanism lol
Yeah I was going to say the same thing, it seems like a lot of the friction problem could be taken care of by a wheel on the reader.
came here to write this and it turned out to be a top comment XD
Teflon pads would be a lot simpler and might work well enough.
Millionaire tape PTFE tape.
Martin if you read this - this is one of the best video's you've made so far in the MM's lifespan. There's not a lot of flair and polish - but there's joy. There's so much play and fun baked into this and it's great to get to see you working in a way that gets you excited about your art. Thanks for this today - we are behind you, thank you for your years of enthusiasm.
Agreed! I’ve been falling asleep looking at spreadsheets on this channel, I even stopped watching completely for awhile. So happy we’re going back to the original format!
Agreed! So happy to see it
I personally like the ticking noise of the latching channel. It also feeds into the aesthetics of the machine looking like clockwork!
Martin kinda fell for the same problem on MMX. Every solution should have the option to be as silent as possible, even if it sounds cool in isolation. Maybe he could take out certain sound dampening based on the song?
Thought I was the only one who thought this way.
The mechanical sounds, especially when rhythmic like these machines, can actually enhance the entertainment experience for, in my opinion. It feels like something is building up, signaling that something exciting is about to occur. Almost like a metronome used as a countdown.
@@livefromhollywood194 I know your right too.. but I do love this ticking sound. But if it is creating noise in the wrong spot so it may detract. Such a shame bc I also love the ticking!
@@NobleSaintedYes. But Martin can always place a microphone directly next to something that makes a little noise. He can't remove noise from something that makes a lot of noise that gets picked up by the instrument microphones. And all the little noises will add up. If I remember correctly, he said that at some point during the MMX.
sometimes a metronome is a drum :D
Martin, I just got a full-ride scholarship to an American college because of *you.* In my application essay, I talked about how you inspired me and the judges loved it. I've been watching you for years. Your videos, your methods, and your community have inspired me to be an engineer myself. Your experimentation, optimism, and innovation have inspired me more that you can imagine. You have literally changed the course of my life, and I can't thank you enough.
Congratulations, friend! All the best over there!
grats! Welcome to the states from a fellow engineer!
Congrats
That’s amazing!
I unlocked a new world, after seeing demons manifest in my room. Messing with ouija boards and seeing them come alive, sleep paralysis to many demonic experiences. I realized we live in a spiritual world more than a physical one. There is a God and His name is Jesus. Try Him with a willing heart and mind and what happens next in life is mind blowing. More than an acceptance to an Ivy League.
If we harden our hearts towards His presence today, then why would we expect tomorrow to be any different?
In regards to the mechanical ticking, I actually appreciated the mechanical background noise of the first two marble machines, it gave the music a distinct sound profile from regular music.
If the ticking is in association with the programming wheel then the noise should, in its own unique way, be tied directly to music and might even sound intentional
You've discovered how to integrate mechanical memory into Marble Machine - essentially MM now has a state machine - so each 'track' can be held in a number of defined states in readiness for performing with a much reduced 'noise-floor' & with much less friction. Elegant solutions are SUCH a joy.
Great stuff!!!
If friction on the programming wheel interface is a big concern (500 channels, wow!), wouldn't a good idea be to use a roller cam follower on the readers, instead of the sliding contact you have now? In addition to reducing noise and mechanical load, it should drastically improve wear concerns, and keep the machine running tight with age. Even with this new design, you're still going to have friction on the actual rise of the profiles, which is still going to add up quite a bit over tons of channels. The trailing edge of the profile can be curved to match the arch of the falling follower and allow a sudden drop if that's the reason for the current knife follower.
Also, for high load components like the bass fingers, would suggest looking into ways to build lots of leverage/mechanical advantage into the system, so that the load felt at the wheel is still relatively low.
I though about the leverage too, I think it would really help
It looks like the idea of reducing friction should be separated from the new toggle functionality. Rollers, of course, some obvious upsides, no downsides. But the capability to toggle the mute on or off, while still being able to put beats when needed is attractive on it's own.
@@CygnusXUnoyes there could be cases where the latching mechanism could be great. But in this case it was used to eliminate friction. You can achieve this goal with rollers/ball bearings at the tip of the reader way easier and far more efficient and have other benefits as stated above.
0:34 "Once in a while, something happens that changes everything"
The fire nation attacked?
Eating bacon for the first time.
@@sworddomo1951 Truth.
I see you are a man of culture and taste. 😁
*DUNN* dun Dun dunnn...
no
Ragarding the friction on the programming wheel, could you not attach a roller/bearing on the reader head to instead have rolling friciton?
I was thinking the same.
It has to be considered yes
Great idea, just like a roller cam follower on an engine vs a a plain tappet
You have to remember it's a prototype, so only the essential and easy to find parts 😉
@@Irilia_nekoI think ball bearings work perfect for that and are mass produced parts. You'd only need to screw them in
Martin: *making random cat noises while playing with a piano-guitar hybrid*
Me: "back to his roots, good, very good"
Boyish enthusiasm activated.
weeeeOOOOOwew
Have you considered using only a one or two string bass? Since hand size isn’t a factor, intervals aren’t either. Generally you’ll only have one or two strings ringing at a time. So this could save a lot of space and parts.
Chicken in the corn!
The bass also looks cool
He could just have one string and change the tension at that point. Maybe have 2/3, 1/2, 3/5 and 3/4 frets and then just change the string tension.
Why stop there? If you turn the string into some cone-shaped cardboard, it can play multiple frequencies simply by moving it at different rates. Maybe we could achieve that with electromagnets?
Nah, strings have different overtones due to their thickness and it's always cool to have a choice.
I just realized this is how a clicky pen work
WOAEW
It's a design worth looking at carefully as it's axial (not linear like Martin's), very clever.
i think it should be possible to have a bowden cable instead of a mine retracting and releasing -> smaller form factor and could be activated by a marble hitting it
in a way yes, it's the same idea wrapped around a cylinder
variations are what are locking latches are. in your kitchen cupboard clicker closer, you sdcard slot etc..
Probably already mentioned but usyally the pickups tracing the programming wheel have small rolls on the end to reduce ware.
Realistically, he should be getting all these parts done in metal for the final machine. The 3D prints are the absolute best for prototyping like this, but I doubt they would hold up to the rigours of a tour, and having to check them before every show would be a huge time waster. E.g. I think the middle nub in the miniature versions of the latch channel would wear out quick, those will probably have to be made in steel. Or the connectors to the tensioning screw: I will assume Martin knows to orient the layers correctly, but even then they might fail under the constant cycles of tension and relaxation (cycles are often what kills parts).
Martin is being smart about this. He will almost certainly need to tweak the geometry of the latch: angles, length, depth of the track, etc. Then add a roller tip to further reduce friction and wear, verify that still works with the geometry. All this can be done rapidly with 3D printed parts. Then version 0.9 will be made in metal and with roller tips, tested again, and finally tweaked a bit (or not if he is lucky) for version 1.0. Im sure once it is integrated into a fully functioning machine, there might need to be a v. 1.1 on some of the parts as integration and long term issues pop up.
Best of luck Martin! There is still a long road ahead, but as they say "How do you eat an elephant? One bite at a time."
@@ParanoidMarvinMk2 or maybe he can get away with some fancier filament like nylon and carbon fiber reinforced filaments, etc... but just regular PLA is not really good for this job.
@@ParanoidMarvinMk2The comment wasn't about the wear and tear of the latch but for the programming wheel and its mechanism. I fear that this is a typical martin solution. You solve a problem by creating another (more complex) problem
@@biziluxgames8924 if you watch, he actually is NOT even orienting the layers correctly. The connectors for the tensioning bolts have their layers perpendicular to the stress, not parallel. They will begin to delaminate under any sustained use.
You could make the contact in the system smaller, by making it thinner and maybe have a ^^^^^ surface instead of a --- surface
This is a great improvement! Small suggestion though. Consider the case where you unmute a channel right before a 24mm programming wheel profile, and the channel has not been placed in the locked open position. In this case, the profile shown could cause a machine crash when playing at high tempos as the pin would get stuck in the first section of the profile.
I would suggest changing the angle of the outside edge of the profile to allow the pin to slide cleanly from the closed position to the fully open position in this case.
Really great suggestion. Hope Martin sees this!
Exactly what I was thinking, ToT profile was independent of speed, this one has a race condition built in, hit it to fast with un-latch and it will crash
That's what I was thinking.
He could build a second escape channel on the far left of the profile that allows the unlatching profile to fall through, but is to far left for the latching profile to reach it.
Could make a 'special' combined unlatch & play note profile for those few songs where tempo is high enough for a crash to become a significant risk...
If you look closely on the small profile, the shape is a bit different. The back edge trails directly towards unlatch, instead of going forward towards latch as it does in the large profile. With the small profile, this issue is handled. The thing that might cause a worry to me is that the latching is done by gravity, so I would believe it might not latch properly in certain cases.
honestly I don't see the click as a minus - Many instruments have noises they make alongside their notes that aren't part of the note, but become part of the instrument's identity. The original Marble Machine (and the MMX) drew a lot of their unique sound from the rhythmic click-clacking of all those parts moving about, and I think the soft click with every movement of a marble gate will have the same effect here
Absolutely! The first Marble Machine starts and ends with the lights turning on, the song begins with the winding of the marble machine, and the latch dropping the first marbles is the very first "note"
The sound of the machine is part of its identity. It's understandable to reduce it so you can hear the music, but getting rid of it entirely removes some soul out of the marble machine.
I love your watching you enthusiasm and engineering process. I've been watching since the first Marble Machine and I'm so glad you're still going strong. Keep up the good work!
Have to admit, seems like a problem that a small bearing on the programming finger woudl solve handily .
The latch is still useful for really really long sections, at very least saving a lot of time not 3D printing profile segments
If a note is not played the entirety of a song, it would need a profile the entire diameter of the wheel (to keep that “finger” lifted). This is a set of on off profiles with a hold function, so one off profile covers it instead.
That's not true, because muting of channels is a different function. @@amaze2708
That shape is also what's used in ballpoint pens to latch the lead in the outward position on one press and let it snap back inside on the next press.
And some bathroom sink pop-ups. Fascinating and genius mechanism
This will probably get buried, but you are using a shift cable in a brake housing; shift and brake cables are slightly different diameters, and so is the internal tube from the housing, it may work better and have slightly less slop in the system by using shift housing. Also as a secondary benefit, shift housing is a non-compressible housing, which should allow for better indexing for precise movements. The friction may increase slightly in a shift cable housing, but for very light loads you can push the cable with moderately more precision than the brake housing. Always love the content and keep up the good work!
RE: Bass Levers and weight -- double the distance from the fulcrum from the spring on the pull down side will halve the weight
quadruple the distance from the spring side will quarter the weight... etc. Round figures.
re: the ticking sound that it added. You're going to put microphones on the sounds you want - you won't mic up the ticking noise. No issue with the ticking.
Yeah just make sure the ticking is not in the same place as the instruments playing.
Regardless, the ticking is nothing compared to the marble on marble sounds.
"The playfulness is the point." It's nice to see you finding joy with this project again.
VERIFY VERIFY VERIFY. Don't get complacent!! You need to measure friction!!
Correct. What is the total friction of 50 or 100 channels playing on the same beat?
"we- we're going to be able to make like beeoww chapa goonga bwwyow maybe no- yeah weow.. WEEOOOWWW" ~Wintergatan 13:04
I love the whimsy that you have in this video. It is the most cheerful and excited you have been in a while. It’s like you’re the chef in “The Menu” and you just made a cheeseburger. Iykyk
Playful and excited Martin is back! So happy to see it! He was in there all along and I’ve enjoyed everything you’ve been doing, but seeing you happy brings me so much joy.
The reason I watch your channel is I enjoy seeing you try, try and try again on something that is incredibly difficult. You’ve got this
The click noice can even be a feature 😉, you just have to verify that the it can hande the max bpm
0:17 Me explaining to the car mechanic what is wrong with my car
Couldn't you fix the friction and noise by adding a wheel to the tip of the reader?
yes, but it would also add moving parts & complexity.
An off-the-shelf bearing.
@@egg5802I get jokes.
Now bring us some this changes everything merch featuring the shape!
this is cool, you've created a mechanical set reset latch, you can think of it in terms of setting and resetting whatever actuator its controlling between its active and passive position
Options to reduce the contact noise of the readers/registers:
1. A small roller bearing/wheel to turn your scraping into rolling, eliminating wear and remaining noise.
2. Coat or make the programming profiles/registers from Teflon/a similar low friction wear resistant material. Another alternative thats pretty machineable is Acetal Copoly (aka delrin).
3. Introduce a cushion area in the track where the new latching profile is clicking. Optionally, consider a dry lubricant in the groove - a passing thought is finely ground graphite.
Just my two cents i wanted to voice, if its of any use to you Martin! Great work thus far!
For the necklace, you should make a heart shape out of that new latch pin channel design by getting two of them and sticking them together mirrored with one another.
This new design looks pretty promising!
Could I suggest using piano bass damper felt for your string mute? It's V cut may give more surface area contact and it would last a few decades longer than your current material. If you use a whole piano bass damper assembly you would have further movement to fit flexibly, although with your plastic arm that probably already exists.
I guess Wilson has a new friend now, THE shape
I think the toggle is a great Idea but adding it for every channel will introduce more wear on small parts which will be harder to replace. Adding a roller bearing as other suggested will reduce friction and reduce wear. I would cycle the latch until it breaks. Love the video!!! keep going!!!
Starting is Always the hardest Part, but when you do, never Stop,!
Love the gleeful joy you have, it's infectious. With regards to the "fingers", moving the fulcrum closer to the fingertips would require less force on the wheel to press them down. Then of course the travel length of the tips would shorten but as long as they completely disengage the strings everything is fine. Now you are a lot smarter than I am so you probably have the optimal dimensions for the fingers but I thought I would give my 2 cents. Keep up the awesome work! Can't wait for the finished product.
"WOW!" - Owen Wheelson
The Wheelson Family:
Wilson Wheelson
Steelson Wheelson
Woodson Wheelson
Plastison Wheelson
Someone needs to handcraft a Glassson Wheelson.
your doing the thing again.... the thing where you make things super over complicated.... that whole mechanism to mute the string is crazy
I don't like being negative, especially in the face of such enthusiasm and optimism, but given that he never got either of the previous two machines to work as such and he's now moved onto this mega machine with hundreds of channels...
Personally I think that the best approach would have been to build a simple 1 or 2 octave basic marble vibraphone and get it working. Then start looking at bells and whistles. Stephenson didn't start by building Mallard, the Wright brothers didn't start with Concorde and NASA didn't send their first rocket to the moon...
This Old Tony is a worldwide treasure
His channel seems dead.
@@thomasbecker9676 You summoned him, he just posted a new video
17:14 - the best advice for those of us who try and solve all the problems before we start. I don't know why this popped into my feed, but it is the advice I needed right now.
Nice to see new stuff :)
I am excited to see all three parts(mute, finger, and dropper) come together to play a note in your test rig. The string instruments in your machines always seem to bring out such creativity and passion!
Profile ine your bass? Operation!
i feel like the mechanic fingers would be cooler being pressed by hand, like a piano keyboard. This makes the bass a hybrid instrument plucking the note with the marbles but hammering fingers on the fret with a keyboard played by martin. With that bike cable you can transmit the mechanical force far away, so put it next where you are playing! An automatic machine is cool, but a hybrid one is even cooler, especially for concerts (human touch, yay!)
"...Mechanical latch..."
"...We have a slide!"
"...don't wait for inspiration..."
Martin's entire character growth; music, engineering, and philosophy woven into one seemingly adlibbed video. What an incredible time we're at in the journey!
I love the happiness of martin in this video. Missed this emotions from him since mmx
"How his shape just changed Everything" 😍
😏
I bloody adore the unbridled excitement about the sliding-possibility in the intro. Makes me so happy
"His shape"? Or "this shape"?
please change the title, Martin 😂
His shape being this old Tony's shape. As in the path. He mentions it less than 5 minutes into the video
@@harmonic5107Exactly. Either title works, given the context.
Wilsons shape... Thats where the trouble begin... These gears ... These damn gears
Maybe the real shape was the friends we made along the way
This is such a great and underated video. You are a very intelligent man. This is what technology was supposed to be used for. Coming up with ideas to solve problems, then rapid prototyping to see what works and what doesn't. Over design, under design, redesign, and come up with something really cool, it's fun. And now anyone can take this idea and come up with something of their own. Brilliant!
Sorry to be that guy. But can we have one video where you don't claim that everything has changed
He's planning to use a new mechanism on like 500 mechanical channels. I think it's pretty accurate. Also are you sure you don't want to be that guy?
^ exactly this particular mechanism literally did change everything.
I remember the watching the original marble machine video, and this video reminded me of that. I love the joy and enthusiasm martin has
Did it change everything and make you realize you should build something instead of just making YT videos?
Just like it's nice to see the brush strokes on a painting I also think it's beautiful to hear the mechanical noises of the machine.
”Ooo oohh….
ÅÅÅÅÅÅÅÅÅÅÅ SLAJDAR (🇸🇪)
OHHHH YES!….det, uhm, I can slide, OK
I was so nervous I couldn’t slide, listen to this
That is soo cool, I can slide
We are going to be able to make that
BOW choopa khugha wheeooooow
Maybe not……yeah, whaaaooww.
WHAAAAAAAAAAAAAAAOOOWW!!
Ok, this makes me really happy
WWOOOOOOOOOWWW!!
This is going to be the BEST mechanical instrument ever built, ok.
Like…, EASILY!”
This section is TH-cam gold right there 😂😂🤣
I'm so excited for this! As far as the clicking sound, its not a big deal at all. As some others have said, adding a roller at the contact point will reduce wear. Given that it is on a bowline cable, you can move the clicking mechanism as far away as you want in a soundproof box.
Also, I love how you are concerned with clicking now, after you designed and built the fish stair marble lift for the MMX.
I can't wait for what is next!
6:00 I'm happy to see that you use an improved version of the variant on the neclace. Meaning it can push through the mecanism when hitting the large profile from the resting position. ❤
Print the corners where the pins hit with TPU inserts. Also eventually cnc the part from teflon or some other slidy stuff of choice.
Congrats Martin, I'm happy to see you smile again.
I love this channel cause its a beautiful story of human ingenuity and perseverance
....as well as: how to run a scam.
17:20 your comments about just doing reminds me of the aperiodic monotile, the mathematicians kept imagining but some random guy started doing and stumbled upon the answer!
Also, a small firm rubber roller on the latching mechanism should help some to reduce clicking as well.
I can see one small quirk remaining: The necessary force needed to push the mechanism past the disengagement (and the other) thresholds hasn't changed. You have minimized the time, but the friction is still momentarily present and will probably grind down your profiles and cost you energy per state-change/note played. But you need the force at the instrument to operate it, just not at the wheel...
Idea: Get the grandfather clock anchors back in! Have weights that exert additional force to each finger, _but only in the 'down' position_
Operation:
When a 'lift and lock' command (medium profile) comes from the wheel, have it first disengage the weight, then lift the arm.
When the command to 'press' (high profile) comes, connect the weights to the fingers as soon as, but only after, you pass the threshold so they do the work, *meaning you can use softer springs!*
When a single 'lift momentarily' command (lowest profile) comes - do nothing with the weight - I guess that quick response is more important in that case and low profile means low strain on the programming wheel anyway, so it is an acceptable tradeoff.
Integrate this control mechanism into "the shape" as well, so it all works regardless of BPM or state of global 'mute instrument' lever.
Love the concept, it is so elegant! A small issue I see is: Let's say the mechanism latches in the middle state with a mid size block, and you mute it. Then, a tall block comes, which is supposed to get the muting arm down on the string. However, since the mechanism is muted, the tall block does not engage, and the mechanism stays in the middle state. So when you unmute, suddenly you have a string that is supposed to be muted, which isnt. So hard to convey in a TH-cam comment but hope you get my drift?
That spring tension mechanism is used in most drum kick pedals! The Mapex Falcoln has a niftier version where you don't need to pull on the string to adjust it, which is quite useful with stronger springs.
I can see why you're so excited about this! I was commenting last video about some ideas to make the bass play more human like ... and this implaments one of the things I sugested, but with a whole other dimention of possabilities with the latching mechanism you've created. Nice!
This is the way to go. You can search single parts extremes and take all the time you want. Sometimes you might think its too much, but then again you explore what is possible and whats not. Then decide how far you want to go with the design.
Also with this method if you find something new you dont have the whole machine built. You can easily re-iterate the old designs and prototype the new version individually, without actually have built anything.
I love seeing you just glow with joy! This entire project perfectly satifies my musical intrique aswell as my love for weird mechanical stuff!
It is so good to see you so energized. For someone that deals with chronic pain and severe depression it is inspirational. Thank you.
I had made a post about adding a preload spring to the activating side of the gate, but I realize that would just reduce the tension being used to hold down the finger, which you don't want. BUT, there is another way to reduce the friction on the programming wheel: leverage. The higher the resistance of the "instrument", the more leverage you need to overcome that friction without putting excess downward force on the programming wheel. What you would need is an adjustable pivot arm on the gate, with an adjustable pivot position, or just different sets of them, since their locations wouldn't need to change once the machine is built. The higher the pressure, the longer you'd want the lower half of the arm to be vs the upper half. The trick would be that the programming profiles would need to be taller, since changing the arm lengths would reduce the travel. That's really how a lever works anyway, it spreads out the force required over a longer distance (and time).
There's clearly a limit to this, since you only have so much time for the profile to push up on the lever and activate the gate, and making the profile steeper will counteract the longer lever, again putting more force on the programming wheel, only now you'd be trading downward pressure for rearward pressure, which is almost certainly worse. But, for things like the bass fingers, unless the tempo is REALLY fast, you probably have a fair amount of time after the pressure is initially released to get the gate to the locked open position.
Obviously, you could do the same thing on the bass fingers, using as long of a finger pivot arm as possible. There will probably be construction limitations of that, plus you need some minimum distance to lift the finger, and it would increase the preload tension required on the spring to hold up a longer lever while keeping enough tension on the Bowden cable. Carbon fiber finger pivot arms perhaps?
One of the best "non workshop" tools I've ever purchase was a couple of different sizes of Crocodile Forceps - easily purchased from medical supply websites. They're AMAZING for retrieving lost items inside guitars, reaching fiddly small things inside of complex mechanisms, or merely holding something small and fragile securely with a variable grasping force
Normally I'm a stickler for those cute little mechanical noises, but I'm with Martin on this one. That scraping is wearing out parts on a touring instrument, and it's an unnecessary expense. Great design!!
Oh, also I think lengthening the top channel could help keep the channel from being damaged by the "unlatch" pin when it's not already latched - like just a place for it to glide to that won't hurt the mechanism.
A tiny foam pad at the connection point can help reduce the clicking noise. Super excited about the new part, and what it means for you.
I love the mechanical ticking of the machine. It makes it part of the melody. With the very first machine video, every single tick, rolling sound, falling marble, it is all part of the melody. Yes it wasn't perfect, and this is what people loved.
Fabulous playing, great ingenuity and just a pleasure to watch your genuine enthusiasm! ❤
I'd suggest you should check the force required to lift that reader up the ramps - the friction of 500 channels rubbing the profiles can be reduced with the right material selection, but the force required to lift that profile against all the return springs, bowden cable friction etc doesn't change - so the latch might actually on average mean more force required than rubbing on most channels of the machine.
The latch is rather neat, clearly should be used on some instrument parts that will actually want to stay latched down for a prolonged time, but I suspect the marble droppers are better off being allowed to rub - you are only rubbing for the duration the note is primed but not released yet, a time you can make almost arbitrarily small by changing the profile lengths.
The year is 2048, and Wintergatan X still isnt finished
His constant need for perfection is his biggest weakness. At this rate his machine will require his children's children to complete.
That's a huge eureka moment! Goes to show how much effort you and everyone involved put into this beautiful project ❤
Maybe a fan that's a jeweler can make you a more practical, robust version of that necklace for your day to day use. Or get together with a designer and sell them to grow your project funds if needed.
People could wear them as a token of gratitute and a reminder, that there is a solution to everything, we just have not found them all, yet!
I think the note on note off system is amazing for the cyber bass but I would argue that it's not very good for the marbel drops since the marbel is gonna fall anyway no matter the length of the profile, and the latching and unlatching could prevent you from playing faster repeated notes
That “wow” in the beginning that was just like Owen Wilson!!!! Hahahahahaha
My condolences in advance for the future music programmer :P
programming wheel and the finger riding it reminds me of a camshaft in a car, overhead cams to be precise, there the spring tension is extreme and oil is used to keep friction low. If the profile of the programming pins made sure there is no "impact" but rather the finger is pushed up, friction and noise could be reduced even more. So, if the leading edge of the pins was convex instead of concave, for example.
Dunno how this turned up on my feed , but man,is this hellacool or what???
I strongly suggest you to watch the channel from the beginning, it's a wild ride indeed, I follow this guy closely since 2016. Martin Molin is the biggest brag Sweden has =) Also check out Wintergatan's music, it is SO cool, you might like it
Not sure what I just bumped into. But it’s great!
What I love with TH-cam is that it shows me great material that I did not search for - made by complete nerds who put a lot of effort into solving fascinating problems. 🙂
Subscribed.
Real Engineering recently released a video where they interviewed an engineer in an aeronautics startup and one thing that stuck out for me is that engineers are creative people who are good with numbers. I had to think of Martin with his struggle to balance the creative and engineering side of the MM. I'm glad to see this project is gaining traction again and going in the right direction!
Your enthusiasm for the slide is great, even on its own.
I'm really excited for this new MM chapter. Your new approach seems really promising and exciting. With your plans encompassing so many features into a single machine, have you considered applying these techniques to designing and building individual smaller marble machines first (maybe a designated bass playing machine, a vibraphone machine, a drum machine, etc. ) and eventually using what you have learned to design your magnum opus Marble Machine. I feel like this would allow for faster turnarounds on smaller projects, and a series tangible and functional finished machines, rather than trying to squeeze every feature into a single machine straight away.
Did he just make a set/reset circuit machnicly? OMG. Genius
I love this, it feels like the old marble machine videos!
oh man this so genuine and its literally art and engineering but fun! And there's even some nice mindset advice that I wish everyone (and me some time ago) agreed with, I love this channel so much :D
Not to rain on your parade, Martin, but the magic-shape latch interfares with the self-latching override hooks (I'm making theese terms up as I type this, hope it's legible) causing droppers/dempers remain in the state they where in when their programming channel was muted instead of finishing their note and then getting disabled. Not to worry though, switching the order of the components (programming wheel reader -> magic-shape latch -> muting hooks -> dropper/damper/finger) will restore that finish-but-don't-start-anything-new functinallity. No new components needed!
Fantastiskt jobbat med basen och lycka till!
Wow, the last two minutes were so insightful and actually the piece of advice I needed for my own creative block
The "giant" differs from the "small" shape in an important manner. The small one allows the tall wedge to move all the way to the end; the giant sample requires that to pause. (The difference between a click-pen which must latch, and a half-press/full press choice.)
13:14 Never have I heard a human so perfectly approximate the classic anime ✴wow 🤩 sound effect.
Honestly amazing by this project I love how it's developing! My fingers are triple crossed hoping you get it all figured out
As whom supported martin since mm1, it's such a pleasure to see martin with joy and excitement with him.keep up!