this is the BEST explanation of this I've seen!! i finally felt the concept really click in my brain. thank you so much! 7 years later and this video is still so helpful to many people
Thanks so much for taking the time to organize this video with audio examples because when I read an article on dithering, I understood what it was saying, but I couldn't visualize what it was saying. Now I can!
Wow! What an amazing video to explain this using extreme examples so we can actually both hear and see what's going on! I love this and also think it's the best video on Dithering out there that I've come across so far!
very well made video. great explanation. Never understood dithering like this before. thanks a lot. Could you make one on summing of audio and its effects in ITB mix downs?
not really.. few opt for print method and others are fine with a simple bounce. Have read a lot of reviews. But i have found that daw bounces sometimes loose quality which can be perhaps retained by print.
A great video to understanding one of the most important "bits" of digital recording, especially if you are going from analog to digital and things just aren't working like they used to! I think this is the first lesson anyone entering digital recording needs to understand and your video did that quite nicely! Thanks for the helpful information.
Great illustration on Dither. I viewed this to refresh myself as a mixer and mastering engineer. I got because I was being a second ear for a song and low and mid level random noises in a supposively mastered version. I a MTSU Recording graduate 2014.
Ive still got loads to learn and loads to understand, but this was so great and informative for a beginner like me. I can hear the differences and sort of see what causes the magic. I just need to dig in with time and practice. Thanks for making it conceptually graspable!
I barely make comments on the net but this vidéo is really the best one to understand dither I ever saw on the net (And I am a french language guy). Congratulations Sir Tarr !
Very, very good tutorial for newbies like me who just discovered the concept of dithering. It's something not talked about alot but it' so seems to be very useful. Thanks for the tutorial!
Very informative. I learnt to add dither when reducing bit depth as a kinda rule but I never really understood why or what it's doing. Now I do. Thanks a mil :)
When it comes to 24 bit and dithering, I like how Dan Worrall did put it: „If you have to amplify a signal by 100dB to hear a problem, it’s not really a problem.“ But yea, for 16 bit, dithering is kinda mandatory.
Very good video-I'm sorry I have to complain about one detail. You probably noticed that the shaped-dither demo starting around 13:57 doesn't really pay off as planned. Unlike the earlier example, it doesn't suck the distortion peaks down by decorrelation, it just buries them in noise. In fact, the lowest distortion peak still sticks out above the noise. The problem is that "noise shaped dither" isn't "dither with shaped noise", it's "dither that has been noise-shaped". The shaping happens as part of the truncation, not to the noise addition before the truncation. It's really "error shaping", to differentiate is from the broadband noise that's added, but to engineers, error and noise are synonyms.
Absolutely brilliant, thanks so much for this. Finally understand what dither does (and why) and I've been working with digital audio for over ten years!
you said we intentionally add noise but when the bit is reduced the new bit reduction noise will be added and there is a previous one too what if both frequencies crash with each other, so here i recommend always invert the polarity of signal of the noise we add intentionally so it cancels the new noise which is added by bit reduction , what do you all think
Excellent explanation and tutorial, except for the minor technical error - what you described as 2-bit must have been at least 3-bit as 2 bits would only have a maximum of 4 levels. But top marks for making a technically obscure issue easy to understand
2mn of Video I already understand broadly what is dither and WHY we do it. Nice. Lets pick some desert and watch the rest. I understand dithering in image and basically the principle is the same. You add noise to actually obtain a more pleasing result. In imaging dithering is handy when you reduce the colour count of an image. Take any 16 bit color image and reduce it to 256 colours. You'll notice clear surfaces of colour that does not look nice and often change the original shape. Take dithering and there will be a noise of colour that is noticeable yet more pleasing than the non dithered image. Nowadays dithering images have become pointless. Or maybe in some specific use cases it is still relevant but almost every image nowadays can have true colours. Disk space and bandwidth are no more bottlenecks to take in the equation.
So, now that the mix has been printed as say, 16bit @ 44.1kHz utilizing the dither plugin, do we just find the file in the audio audio folder and use that for all intents and purposes? If I understand correctly, when we want to export to mp3 instead, using 3rd party software, should we print the mix as a non-dithered file and then use whatever 3rd party software to create the mp3 since it also applies dither to the final file?
thanks this helps a lot! what if I export a master file in 24bit adding some dither noise and it gets bitreduced at a later stage when being converted. will the dither noise help preventing the quantization noise as well?
What was the use of adding noise in the case of 8 bit cuantization? You literally can't hear the cuantization noise, all you did was add random noise to a decent signal.
At about 4:18 in the video you show a signal you say has been reduced to 2-bit resolution, but the visual doesn't make sense to me. If it were two-bit, there would be a maximum of 4 levels in the displayed signal, played signal, but there are at least 7 levels shown. Is the really a 3-bit signal?
Thanks for the video! I have a question, what if you put an EQ with a high cut in the master, let's say around 16kHz, does dithering still works? Also, how do you know how much dither amount to apply?
It was questionable which noise was more unpleasant when you lowered the bit depth to 2 bits. And when you raised it back to 4 or 8 bits, I couldn't really hear the quantization noise. But when you added white noise, I could definitely hear it. So I'm still not sure I understand everything correctly. The concept of the series is good, though...
very great explanation thanks a lot.. my question is this; i do parallel comp, with high shelf eq on my low and high freq just as u did in this video for dithering, its a way of adding harmonics to my mixes, please do i stand a chance of destroying my mix if i apply dithering? are there differences btw parallel comp with parallel eq and dithering
Thank you so much Eric. This is amazingly helpful to really understand what this process is about. But I do have some questions. Why should we lower resolution when we bounce to a stereo track in mastering (or mixing)? What if we leave it as it is or lower it as less as possible? Are there playback devices that do not work on higher resolutions or are all of them different? What is the criteria to choose the right bit depth? Thanks a lot man. Cheers
Wonderful video, thank you very much. I am left with one question, though: The harmonics, which you bring in at 13:54, I can see very well, but hardly hear. The noise, which you bring in at 14:08, I can hear very well. So, my question is - why dither, when the quantification error is much more silent and hardly as disturbing as the dithering noise, which is being used to mask the error?
Pretty complicated topic, but I think, I finally got it. Is this maybe also partially an explanation as to why older mp3s, which had like 96, perhaps 192 kbps data rate max seemed to sound a little louder than they would do nowadays with 320 kbps? I think, I had such a feeling in the past, when I was shopping mp3s or just listening to all my files on my pc, where you could compare songs with various data rates to one another... I just could never explain it...
Great video! I have a quick question - the nature of harmonics is to get down in amplitude as the frequency goes up, so why would we like to add more noise there? It seems more rational to add the noise at the mid frequencies so we can mask them better no?
Guy Shilo i'm no expert on the topic and the harmonics indeed reduce in amplitude as you go up the spectrum, but human hearing is a different thing, there's the fletcher-munson curve, which kind of "allows" you to boost the higher frequency noise to maximize the masking effect and it still won't be audible unlike in the upper mids. so my guess would be, you do it because "you can"
Nice video! Love it! I have some question that just curios to me. What's the point of adding noise at 15-20khz as you not gonna hear it anyway, Isn't just belling out some middle enough for noise shaping? Thanks!
![Compression Plugins [Analyzing Mixing Effects]](http://i.ytimg.com/vi/AKS7EdK0dPY/mqdefault.jpg)
![Compression Plugins [Analyzing Mixing Effects]](/img/tr.png)
this is the BEST explanation of this I've seen!! i finally felt the concept really click in my brain. thank you so much! 7 years later and this video is still so helpful to many people
thanks dude, literally no one talks about this topic so well done.
The best explanation that can be find for that !
Tnx !
Thanks so much for taking the time to organize this video with audio examples because when I read an article on dithering, I understood what it was saying, but I couldn't visualize what it was saying. Now I can!
Wow! What an amazing video to explain this using extreme examples so we can actually both hear and see what's going on! I love this and also think it's the best video on Dithering out there that I've come across so far!
Eric Tarr's tutorials are always excellent
This was godlike. Thank you, sir.
you are a GEM MY FRIEND! world won't need schools if RUclips is utilized in such an exceptional way!
Definitely gave me a new insight about dithering. I really had no idea how it worked, just that it was necessary. Thank you!
very well made video. great explanation. Never understood dithering like this before. thanks a lot. Could you make one on summing of audio and its effects in ITB mix downs?
Hey, did you ever get a answer on bouncing methods? Print?
not really.. few opt for print method and others are fine with a simple bounce. Have read a lot of reviews. But i have found that daw bounces sometimes loose quality which can be perhaps retained by print.
A great video to understanding one of the most important "bits" of digital recording, especially if you are going from analog to digital and things just aren't working like they used to! I think this is the first lesson anyone entering digital recording needs to understand and your video did that quite nicely! Thanks for the helpful information.
Great illustration on Dither. I viewed this to refresh myself as a mixer and mastering engineer. I got because I was being a second ear for a song and low and mid level random noises in a supposively mastered version. I a MTSU Recording graduate 2014.
This is a super-clear discussion of an abstruse topic, and makes a very good case for noise shaping!! :) Thank you!!
That abstruse was an abstruse autological word.
Ive still got loads to learn and loads to understand, but this was so great and informative for a beginner like me. I can hear the differences and sort of see what causes the magic. I just need to dig in with time and practice. Thanks for making it conceptually graspable!
Superb video..... so well explained... its amazing. Thank you soo much
I barely make comments on the net but this vidéo is really the best one to understand dither I ever saw on the net (And I am a french language guy). Congratulations Sir Tarr !
This is one of the most useful videos I have seen in a long time Thank You!
Very, very good tutorial for newbies like me who just discovered the concept of dithering. It's something not talked about alot but it' so seems to be very useful. Thanks for the tutorial!
Brilliant dude! I had no idea what dithering was. Makes complete sense now! Thanks! You have excellent examples!
New video from Eric Tarr all about dithering.
Very informative. I learnt to add dither when reducing bit depth as a kinda rule but I never really understood why or what it's doing. Now I do. Thanks a mil :)
this was just interesting from a technical standpoint, too... great video
Now I understand!! Superbly explained, thank you very much. Excellent presentation.
probably my second comment on youtube ever; but i think you did an incredible job of explaining this concept. Great job man, subscribed!
When it comes to 24 bit and dithering, I like how Dan Worrall did put it: „If you have to amplify a signal by 100dB to hear a problem, it’s not really a problem.“ But yea, for 16 bit, dithering is kinda mandatory.
The cops came in the end. They wanted to bust you for being awesome. Brilliant explanation. You made dithering sound a bit less boring!
Very good video-I'm sorry I have to complain about one detail. You probably noticed that the shaped-dither demo starting around 13:57 doesn't really pay off as planned. Unlike the earlier example, it doesn't suck the distortion peaks down by decorrelation, it just buries them in noise. In fact, the lowest distortion peak still sticks out above the noise. The problem is that "noise shaped dither" isn't "dither with shaped noise", it's "dither that has been noise-shaped". The shaping happens as part of the truncation, not to the noise addition before the truncation. It's really "error shaping", to differentiate is from the broadband noise that's added, but to engineers, error and noise are synonyms.
Absolutely brilliant, thanks so much for this. Finally understand what dither does (and why) and I've been working with digital audio for over ten years!
Fantastic job. Absolutely brilliant demonstration.
Cool video. I'd love to see an example of an actual song tho, since I couldn't hear the harmonics of the sine wave from 8.0 bit depth up.
Excellent visualization, great job!
Finally an awesome explanation with an awesome example!!!!! Thanks!!
Thank you very much! Your teaching is very good, congratulations!
I guess you already know that you're a genius. If it wasn't for you, I would go to my grave never understanding what dithering was. Thank you 😊
Stunning to see the effect in real-time, cool vid man!
One of the best videos I've ever watched, thanks!
Great video. Well spoken, informative, and visually satisfying. Thank you.
Thank you for taking your time to explain! Very well presented. Clear and to the point! subscribed.
Excellent explanation! Clear and to the point. Cheers
Thank you so much for explaining so visually
Thanks, great explanation!
dithering is so cool lol. i dont think it’s boring. thank you fam
Wow! Killer video!
Thank you, Erik. YOU ARE A GOOD MAN!
Finally a good one explanation. Than u so so much
this is amazing. so easy to understand
Very well expanation !!!! Now I understand what dithering mean. Thank you, sir
do you ever dither form 32 bit floating point to 24 bits? I usually don't bother but wonder if it is affecting my quality
Fantastic explanation! Thanks so much
Very informative, Great work!
Wow that was very informative!! Thank you!!! 😁
Best Dither video ever!
Excellent video! Very easy to follow!
you said we intentionally add noise but when the bit is reduced the new bit reduction noise will be added and there is a previous one too what if both frequencies crash with each other, so here i recommend always invert the polarity of signal of the noise we add intentionally so it cancels the new noise which is added by bit reduction , what do you all think
Best explanation ever for this topic!
Thanks buddy.That was very well explained and very helpful!
thank you for your detailed explanation
great video Eric, concise and very informative, thanks :)
You're brilliant. So thorough
godlike tutorial, thank you
I love this video
Excellent video! Thanks very much for this.
thanks. such an amazing explanation
Excellent explanation and tutorial, except for the minor technical error - what you described as 2-bit must have been at least 3-bit as 2 bits would only have a maximum of 4 levels. But top marks for making a technically obscure issue easy to understand
Thank you very much :) that was quite educational and the visual demonstrations helped a lot :)
2mn of Video I already understand broadly what is dither and WHY we do it. Nice. Lets pick some desert and watch the rest. I understand dithering in image and basically the principle is the same. You add noise to actually obtain a more pleasing result.
In imaging dithering is handy when you reduce the colour count of an image.
Take any 16 bit color image and reduce it to 256 colours. You'll notice clear surfaces of colour that does not look nice and often change the original shape. Take dithering and there will be a noise of colour that is noticeable yet more pleasing than the non dithered image.
Nowadays dithering images have become pointless. Or maybe in some specific use cases it is still relevant but almost every image nowadays can have true colours. Disk space and bandwidth are no more bottlenecks to take in the equation.
Excelent demonstration! Thank you!!
So, now that the mix has been printed as say, 16bit @ 44.1kHz utilizing the dither plugin, do we just find the file in the audio audio folder and use that for all intents and purposes? If I understand correctly, when we want to export to mp3 instead, using 3rd party software, should we print the mix as a non-dithered file and then use whatever 3rd party software to create the mp3 since it also applies dither to the final file?
Nicely done video, thank you!
Very helpful! Thx Eric
thanks this helps a lot! what if I export a master file in 24bit adding some dither noise and it gets bitreduced at a later stage when being converted. will the dither noise help preventing the quantization noise as well?
What was the use of adding noise in the case of 8 bit cuantization? You literally can't hear the cuantization noise, all you did was add random noise to a decent signal.
Thanks for this important topics 😍
Fantastic! I finally get it. Thank You!
At about 4:18 in the video you show a signal you say has been reduced to 2-bit resolution, but the visual doesn't make sense to me. If it were two-bit, there would be a maximum of 4 levels in the displayed signal, played signal, but there are at least 7 levels shown. Is the really a 3-bit signal?
You are correct, it can't be two bit
Great presentation!
Thanks for the video! I have a question, what if you put an EQ with a high cut in the master, let's say around 16kHz, does dithering still works? Also, how do you know how much dither amount to apply?
Great Work! Thanks!
Very nice tutorial
It was questionable which noise was more unpleasant when you lowered the bit depth to 2 bits. And when you raised it back to 4 or 8 bits, I couldn't really hear the quantization noise. But when you added white noise, I could definitely hear it. So I'm still not sure I understand everything correctly. The concept of the series is good, though...
very great explanation thanks a lot.. my question is this; i do parallel comp, with high shelf eq on my low and high freq just as u did in this video for dithering, its a way of adding harmonics to my mixes, please do i stand a chance of destroying my mix if i apply dithering? are there differences btw parallel comp with parallel eq and dithering
Thank you so much Eric. This is amazingly helpful to really understand what this process is about. But I do have some questions. Why should we lower resolution when we bounce to a stereo track in mastering (or mixing)? What if we leave it as it is or lower it as less as possible? Are there playback devices that do not work on higher resolutions or are all of them different? What is the criteria to choose the right bit depth? Thanks a lot man. Cheers
Wonderful video, thank you very much. I am left with one question, though: The harmonics, which you bring in at 13:54, I can see very well, but hardly hear. The noise, which you bring in at 14:08, I can hear very well. So, my question is - why dither, when the quantification error is much more silent and hardly as disturbing as the dithering noise, which is being used to mask the error?
You have to remember that he is using a single sine-wave tone and vastly exaggerated quantization. In a real scenario you'll need a lot less noise.
Pretty complicated topic, but I think, I finally got it.
Is this maybe also partially an explanation as to why older mp3s, which had like 96, perhaps 192 kbps data rate max seemed to sound a little louder than they would do nowadays with 320 kbps? I think, I had such a feeling in the past, when I was shopping mp3s or just listening to all my files on my pc, where you could compare songs with various data rates to one another... I just could never explain it...
kbps isn't resolution, its data rate
@@debrucey Yeah, fine... doesn't help with my train of thought though...
Finally, I know how dither does the job!
Nice! Thank you for the info.
Great video! I have a quick question - the nature of harmonics is to get down in amplitude as the frequency goes up, so why would we like to add more noise there? It seems more rational to add the noise at the mid frequencies so we can mask them better no?
Guy Shilo i'm no expert on the topic and the harmonics indeed reduce in amplitude as you go up the spectrum, but human hearing is a different thing, there's the fletcher-munson curve, which kind of "allows" you to boost the higher frequency noise to maximize the masking effect and it still won't be audible unlike in the upper mids. so my guess would be, you do it because "you can"
Very helpful, thank you!
Brilliant! Thankyou sir! Xx
So basically i need to do this after i am done with my mix? Should i then put the entire mix to a lower bitrate and add Some noise?
Thank you! Subscribed.
Amazing video 🙏🏽
I still don't get the meaning of dithering. I thought dithering gives depth to your mix, or am I wrong? Any help, would be much appreciated!
Grate video very informative!
Nice video! Love it!
I have some question that just curios to me.
What's the point of adding noise at 15-20khz as you not gonna hear it anyway, Isn't just belling out some middle enough for noise shaping? Thanks!
Why wouldn't you be able to hear it?
Great videos!
Thanks,,Great video!
Very helpful. Thanks!
Nice vid! Learned alot! But is this dithering still relavant these days? Because everything has a super high quality standard now right?
very good Demonstration
Thanks!
Thank you!