Your videos on dithering are the clearest no-nonsense explanations on TH-cam. TH-cam needs more like this. I have watched too many popular but misinformed ones on basic digital audio: sampling, bit depth, etc. And then we have the audiophile ones...
I remember a AES article on how quantization noise derives from Bessel functions (as in FM synthesis and hence adding tones not originally present). On tonal audio with slowly varying levels, this can sound terribly obnoxious and grainy, hence term "granulation noise"
Thank you so much! I was freaking out as I render 24 bit wav and FL Studio kinda suggests there's no need to dither at 24 bit but this clarification helped me loads as I've already rendered atleast 50 tracks without dither (24 bit)
OK, I have some blocking priorities, but I'll get to this (when I wrote it, the plug-in framework I used was pretty buggy, I need it clean it up, or re-write for a different framework).
Hey Nigel....great video. I'm trying to get in contact with you. I got a chance to see some of your resources on your site and want to talk to you about some projects.
so in in 24bit recording, rendering a 24 bit mix, I shouldn't need dither? and here is a question, if i used a 16bit sample in a 24bit recording, is their any steps reverse order here, to avoid issues
@@TEN-TIMES-HARDER It never hurts to dither 24-bit truncations, but it also won't even be heard. Now, some would say, "why not just said always dither, and make it easy on everyone?". The reason I make this point is because some feel that every external send in your DAW (out to a hardware reverb, compressor, tube EQ...) must be properly dithered or you project will be ruin. This is an unnecessary complication-besides the extra setup, there is the paranoia of making sure there are no further gain changes after the truncation, etc. And it's for something that impossible to hear. The truncation distortion of at 24 bits is far below the noise floor of any electronics possible (you can't escape Johnson noise and shot noise). If that weren't enough-for people who aren't electrical engineers, and doubt that fundamental truth-your ears can't hear it (people get fooled by reading that the ear has 140 dB of dynamic range-they forget that you will have permanent hearing damage if you exercise that extreme for even a few seconds on the loud end, and on the other end you're limited by the inherent noise in your room, noise of blood pumping through your veins, and the minimum energy required to deflect your eardrum (it's not massless). If that were enough, only the most artificial cases of computed music won't already have far more than enough of a noise floor to self-dither, as you say. The bottom line is that it's unhearable for several fundamental reasons.
If it's your own high-quality lossless project, will it make most sense not to dither if converting to lower sampling rate/bitrate (example: to AAC/MP3)? Because it came from a high-quality source, so it doesn't need to interpret anything else.
Good question. The process of dithering definitely should be the responsibility of the conversion from lossless, though. Otherwise, the error in the lossy format is not a linear function, so it's hard to predict the effect relative to truncation error. But it wouldn't make sense to dither first, or after. I did a quick look on the web, and Waves says explicitly, "Don’t dither before converting to MP3 or AAC". You're right going directly from a high-quality source.
+🌴HellaHipHop🌴 The recording is 24-bit, so I assume you mean why reduce to 8-bit: it's unlikely anyone would need 8-bit these days-the 8-bit reduction in the video is for an extreme example, so you can hear the effects of dithering. It's not possible to hear at 16-bit, as the video shows, so would do no good to use that as the primary example.
At 8:09 you mention that our DACs cant pick up 24 bit audio and so on, does this mean i shouldn't bother buying 24bit music? I have a few albums that are 24bit FLAC but if I cant hear a difference then should I just convert them to 16bit? I looked this video up because I was using Foobar2000 and had a FLAC album I wanted to split into separate tracks and one of the options was dither. (Im new to the audio scene and do not know too much, I only researched audio for about 2 or 3 months)
Check out this article to get an idea of the levels of the 16th bit and 24th bit; this is a worst-case, a 16-bit dithered signal would be much better, since it's spread smoothly as background hiss: www.earlevel.com/main/2013/03/24/perspective-on-dither/
Yes, sorry about the "coming soon". A top-flight recording engineer had promised to send me some recorded material. I thought it was a better idea than supplying my own, where you might question just how "real world" (of professionally recorded and engineered audio) the recording might be. Unfortunately, he backed out. I found that he and other pros were-perhaps-reluctant to cast doubt on their work, even with the promised that only tiny bits of the material might be heard.
Good request-I'm pretty backlogged after being busy for the past few years, but I do intend to make the plugin available at some point. First up is a new video, in progress...
True, but if you can't hear the artifacts, then they don't exist, effectively. And that's always the case truncating to 24-bit. You can dither whenever you want, no harm, I'm just trying to take the mysticism out of the process.
Yes. Each bit increase in sample size adds 6.02 dB of dynamic range, and reduced the noise by the same amount, normalized to full scale (as audio always is). So, about 48 dB difference between 8- and 16-bit.
OK, then this will look familiar: Adding one bit doubles the range by a factor of 2 (7 bits gives you 128 steps, 8 gives 256). dB_gain = 20 * log10(2) = 6.0206 dB. In audio, we really call max 1.0 and the extra bits are added to the right side of decimal point, so adding a bit cuts the step size in half (instead of doubling the range). So it would be more accurate to say that adding a bit multiplies the minimum step size by 0.5: dB_gain = 20 * log10(0.5) = -6.0206 dB, whichever way you want to look at it. :-)
I just released my own dithering and bit reduction plug-in. It can be found at github.com/datajake1999/C1Bitcrusher BTW, Can you please upload the song you used in the video? Thanks.
It was an unfinished snippet of a song in progress, pretty arbitrary. I actually wanted to use something professional mastered, a couple of professional friends had said they could give me something. But I think they got nervous that my video would somehow highlight faults, or something-didn't happen.
Seems to me that you're completely ignoring the *cumulative* impact of DSP - including the very normalizing you're adding - rather than just the playback of a 32 float or 24 bit source. Truncation distortion simply *accumulates* at a far greater level than dither noise applied to bit depth reductions. In other words, this is less about what's natively audible, and more about minimizing all too easy accumulative losses. Self-noise of a source is not dither. - mastering engineer.
Not at all. The discussion here is about dithering the final mix. primarily-it shows that right up front in the video. Your average person does not have control over truncation in their DAW and its plugins, unless they created them, so it's pretty futile to address that in a video like this. Disagree that self noise is not dither - digital signal processing engineer ;-)
Half way through the video, thank you for sharing your knowledge, you are brilliant. "...So i created a plug-in". Like what??? xD Is this plug-in available anywhere? Would love it! Thank you!
@@nigel_redmon I appreciate the response.. new question and unfortunate development... is it possible to mix and master with tinnitus? Or is it much harder since you can't tell what's real and what's phantom sound from the ringing?
@@HealthyBodyForLife I'll bet that's not an uncommon problem at all, as hazard of the industry. I don't have any experience with it, but I see many discussion threads pop up in a search ("mixing audio with tinnitus"). Good luck!
@@HealthyBodyForLife I'm just a random 40y/o guy with tinnitus, but whether mine is self inflicted (concerts, shooting guns) or whether it's genetic (my aunt has it, but quotient never been around anything loud), is an unknown... Either way, I wanted to share this: With my tinnitus, my hearing is actually not affected -hence thinking perhaps generic vs damage induced- and while I can hear it at almost any time, it's not actually *_masking_* anything in hearing! _(while I've toyed with tone generator apps on my phone, I've never managed to dial in the _*_exact_*_ frequency of mine, to say whether it might be causing that specific tone to be missed by me...)_ That isn't to say it doesn't sometimes make me hear phantom sounds, but that's only when the environment is almost entirely silent. Example, my computer is on and so is an oscillating fan, so there's some audible white noise fan whirring; when I'm outside my room and 20ft away in the kitchen, there are times when I've *_thought_* I heard my computer make the noise indicating I've been messaged (via Discord)... Sometimes I have been messaged, but sometimes not lol. OR, and this case is where it genuinely impacts what I'm doing: deer hunting. Sitting, trying to hear the faintest leaf crinkle and crunch from a deer... It's nearly impossible because I can't break focus from my tinnitus. If there's some kind of quiet, ambient noise going on in the distance though, that helps a ton. All of this is to say, *_for me specifically,_* I'm completely fine, *so long as* I'm able to hear _something_ or I'm listening to anything that's above my (arbitrary) "focus" threshold. Which if there's headphones or speakers involved, if I've missed something audio related, it's *not* because of my tinnitus! 😁 (ie: hardware or audio mix was too low) Unfortunately, tinnitus is a rather vague catch-all for _multiple_ conditions, so even asking 50 people would likely get such a wide range of responses, that you'd be unable to draw any relevant/usable conclusion from them. I would say your best way to tell would be to have an audiologist give a hearing test. Also ask what frequencies they're testing with. Then, sit down with a frequency/tone generator and crank out the sound while wearing quality headphones, and see if you can find any absolute limits! (Samsung's Galaxy S phones have very capable audio out hardware [via specs, experience], so an app and non-Bluetooth headphones *should* suffice)
Great presentation. Hands down the best video about dithering on youtube. Looking forward to your next video!
Your videos on dithering are the clearest no-nonsense explanations on TH-cam. TH-cam needs more like this. I have watched too many popular but misinformed ones on basic digital audio: sampling, bit depth, etc. And then we have the audiophile ones...
Thanks Jack!
I remember a AES article on how quantization noise derives from Bessel functions (as in FM synthesis and hence adding tones not originally present). On tonal audio with slowly varying levels, this can sound terribly obnoxious and grainy, hence term "granulation noise"
hey nigel, this video series is REALLY helping me understand stuff! :) youre doing an awesome job at explaining things, thank you very much!
Thank you so much! I was freaking out as I render 24 bit wav and FL Studio kinda suggests there's no need to dither at 24 bit but this clarification helped me loads as I've already rendered atleast 50 tracks without dither (24 bit)
Glad it helped, and thanks for letting me know!
Hi Nigel. Great videos! Why didn't you continue? Are you going to?
Glad you like the videos, JF. I've been tied up on a project, but intend to get back to it soon and finish up a video that's been on hold...
Thank you. Very educational. I like to learn new things and you have taught me something new.
Love your work! Thanks
@Nigel Redmon, thanks. Will you share this plug-in of yours? Couldn't find it on your blog site.
OK, I have some blocking priorities, but I'll get to this (when I wrote it, the plug-in framework I used was pretty buggy, I need it clean it up, or re-write for a different framework).
Hey Nigel....great video. I'm trying to get in contact with you. I got a chance to see some of your resources on your site and want to talk to you about some projects.
Very useful, thanks!
so in in 24bit recording, rendering a 24 bit mix, I shouldn't need dither? and here is a question, if i used a 16bit sample in a 24bit recording, is their any steps reverse order here, to avoid issues
oh, self dither, so it seems to me dither is only for electronic or direct imput instrument recordings...
@@TEN-TIMES-HARDER It never hurts to dither 24-bit truncations, but it also won't even be heard. Now, some would say, "why not just said always dither, and make it easy on everyone?". The reason I make this point is because some feel that every external send in your DAW (out to a hardware reverb, compressor, tube EQ...) must be properly dithered or you project will be ruin. This is an unnecessary complication-besides the extra setup, there is the paranoia of making sure there are no further gain changes after the truncation, etc. And it's for something that impossible to hear. The truncation distortion of at 24 bits is far below the noise floor of any electronics possible (you can't escape Johnson noise and shot noise). If that weren't enough-for people who aren't electrical engineers, and doubt that fundamental truth-your ears can't hear it (people get fooled by reading that the ear has 140 dB of dynamic range-they forget that you will have permanent hearing damage if you exercise that extreme for even a few seconds on the loud end, and on the other end you're limited by the inherent noise in your room, noise of blood pumping through your veins, and the minimum energy required to deflect your eardrum (it's not massless). If that were enough, only the most artificial cases of computed music won't already have far more than enough of a noise floor to self-dither, as you say. The bottom line is that it's unhearable for several fundamental reasons.
If it's your own high-quality lossless project, will it make most sense not to dither if converting to lower sampling rate/bitrate (example: to AAC/MP3)? Because it came from a high-quality source, so it doesn't need to interpret anything else.
Good question. The process of dithering definitely should be the responsibility of the conversion from lossless, though. Otherwise, the error in the lossy format is not a linear function, so it's hard to predict the effect relative to truncation error. But it wouldn't make sense to dither first, or after. I did a quick look on the web, and Waves says explicitly, "Don’t dither before converting to MP3 or AAC". You're right going directly from a high-quality source.
Great stuff!
with most software studios having higher bit resolutions why record at an 8-bit level at all?
+🌴HellaHipHop🌴 The recording is 24-bit, so I assume you mean why reduce to 8-bit: it's unlikely anyone would need 8-bit these days-the 8-bit reduction in the video is for an extreme example, so you can hear the effects of dithering. It's not possible to hear at 16-bit, as the video shows, so would do no good to use that as the primary example.
Excelent videos, thank you very much. Can I ask, What DAW are you using?
R2 Official TV Thank you. I'm using Digital Performer.
At 8:09 you mention that our DACs cant pick up 24 bit audio and so on, does this mean i shouldn't bother buying 24bit music? I have a few albums that are 24bit FLAC but if I cant hear a difference then should I just convert them to 16bit? I looked this video up because I was using Foobar2000 and had a FLAC album I wanted to split into separate tracks and one of the options was dither. (Im new to the audio scene and do not know too much, I only researched audio for about 2 or 3 months)
Check out this article to get an idea of the levels of the 16th bit and 24th bit; this is a worst-case, a 16-bit dithered signal would be much better, since it's spread smoothly as background hiss: www.earlevel.com/main/2013/03/24/perspective-on-dither/
Coming soon.. :')
Yes, sorry about the "coming soon". A top-flight recording engineer had promised to send me some recorded material. I thought it was a better idea than supplying my own, where you might question just how "real world" (of professionally recorded and engineered audio) the recording might be. Unfortunately, he backed out. I found that he and other pros were-perhaps-reluctant to cast doubt on their work, even with the promised that only tiny bits of the material might be heard.
Ahaaaa ...
More, please! :)
Great video. Can you please upload the song you used in the video and the plugin you created? Thanks.
Good request-I'm pretty backlogged after being busy for the past few years, but I do intend to make the plugin available at some point. First up is a new video, in progress...
Artefacts of quantization and real noise are something totally different... I think we should always dither for converting to lower bit rates...
True, but if you can't hear the artifacts, then they don't exist, effectively. And that's always the case truncating to 24-bit. You can dither whenever you want, no harm, I'm just trying to take the mysticism out of the process.
did you work out its 48dB quieter from the dynamic range ?
Yes. Each bit increase in sample size adds 6.02 dB of dynamic range, and reduced the noise by the same amount, normalized to full scale (as audio always is). So, about 48 dB difference between 8- and 16-bit.
Nigel Redmon thanks ! Iv always worked it out using a logarithmic equation
OK, then this will look familiar: Adding one bit doubles the range by a factor of 2 (7 bits gives you 128 steps, 8 gives 256). dB_gain = 20 * log10(2) = 6.0206 dB. In audio, we really call max 1.0 and the extra bits are added to the right side of decimal point, so adding a bit cuts the step size in half (instead of doubling the range). So it would be more accurate to say that adding a bit multiplies the minimum step size by 0.5: dB_gain = 20 * log10(0.5) = -6.0206 dB, whichever way you want to look at it. :-)
Thanks :)
I just released my own dithering and bit reduction plug-in. It can be found at github.com/datajake1999/C1Bitcrusher
BTW, Can you please upload the song you used in the video? Thanks.
It was an unfinished snippet of a song in progress, pretty arbitrary. I actually wanted to use something professional mastered, a couple of professional friends had said they could give me something. But I think they got nervous that my video would somehow highlight faults, or something-didn't happen.
I actually like the song you used in the video, and I would like to hear the full version if possible.
Seems to me that you're completely ignoring the *cumulative* impact of DSP - including the very normalizing you're adding - rather than just the playback of a 32 float or 24 bit source. Truncation distortion simply *accumulates* at a far greater level than dither noise applied to bit depth reductions. In other words, this is less about what's natively audible, and more about minimizing all too easy accumulative losses. Self-noise of a source is not dither. - mastering engineer.
Not at all. The discussion here is about dithering the final mix. primarily-it shows that right up front in the video. Your average person does not have control over truncation in their DAW and its plugins, unless they created them, so it's pretty futile to address that in a video like this. Disagree that self noise is not dither - digital signal processing engineer ;-)
Half way through the video, thank you for sharing your knowledge, you are brilliant. "...So i created a plug-in". Like what??? xD Is this plug-in available anywhere? Would love it! Thank you!
Yes, I need to do this, but first working to finish a new video, it's been a long time...thanks for asking about it.
@@nigel_redmon I appreciate the response.. new question and unfortunate development... is it possible to mix and master with tinnitus? Or is it much harder since you can't tell what's real and what's phantom sound from the ringing?
@@HealthyBodyForLife I'll bet that's not an uncommon problem at all, as hazard of the industry. I don't have any experience with it, but I see many discussion threads pop up in a search ("mixing audio with tinnitus"). Good luck!
@@HealthyBodyForLife I'm just a random 40y/o guy with tinnitus, but whether mine is self inflicted (concerts, shooting guns) or whether it's genetic (my aunt has it, but quotient never been around anything loud), is an unknown... Either way, I wanted to share this:
With my tinnitus, my hearing is actually not affected -hence thinking perhaps generic vs damage induced- and while I can hear it at almost any time, it's not actually *_masking_* anything in hearing!
_(while I've toyed with tone generator apps on my phone, I've never managed to dial in the _*_exact_*_ frequency of mine, to say whether it might be causing that specific tone to be missed by me...)_
That isn't to say it doesn't sometimes make me hear phantom sounds, but that's only when the environment is almost entirely silent.
Example, my computer is on and so is an oscillating fan, so there's some audible white noise fan whirring; when I'm outside my room and 20ft away in the kitchen, there are times when I've *_thought_* I heard my computer make the noise indicating I've been messaged (via Discord)... Sometimes I have been messaged, but sometimes not lol. OR, and this case is where it genuinely impacts what I'm doing: deer hunting. Sitting, trying to hear the faintest leaf crinkle and crunch from a deer... It's nearly impossible because I can't break focus from my tinnitus. If there's some kind of quiet, ambient noise going on in the distance though, that helps a ton.
All of this is to say, *_for me specifically,_* I'm completely fine, *so long as* I'm able to hear _something_ or I'm listening to anything that's above my (arbitrary) "focus" threshold. Which if there's headphones or speakers involved, if I've missed something audio related, it's *not* because of my tinnitus! 😁 (ie: hardware or audio mix was too low)
Unfortunately, tinnitus is a rather vague catch-all for _multiple_ conditions, so even asking 50 people would likely get such a wide range of responses, that you'd be unable to draw any relevant/usable conclusion from them.
I
would say your best way to tell would be to have an audiologist give a hearing test. Also ask what frequencies they're testing with.
Then, sit down with a frequency/tone generator and crank out the sound while wearing quality headphones, and see if you can find any absolute limits!
(Samsung's Galaxy S phones have very capable audio out hardware [via specs, experience], so an app and non-Bluetooth headphones *should* suffice)