[old tech]

Just wondering if someone could provide a rebuttal to this article from a studio.

https://theproaudiofiles.com/dither/

I'm always keen to learn more about dither, ie how it eliminates quantasion errors and self dithering in analog devices (ie random noises introduced by errors in mechanical/transducer playback).

On another forum, a member there strongly argues that dither results in a loss of clarity, in addition to higher noise. While most of his posts do not make sense he does raise an interesting question. If the trade off with dither is just increased noise (ie tape hiss noise) then why wouldn't dithered 10 bits sound as distortion free as 24bits, just with more noise?

 

[71 dB]

I'm always keen to learn more about dither, ie how it eliminates quantasion errors and self dithering in analog devices (ie random noises introduced by errors in mechanical/transducer playback).

We can't avoid noise, but in digital audio we can choose what kind of noise we have. We can have quantization noise (no dither), reduced quantization noise (weak dithering) or no quantization noise at all (full dithering). The idea of dither is to have less harmful type of noise. The big problem with quantization noise is that it correlates with the signal causing granulation. Dither noise breaks this correlation: We lose the correlation due to the added noise, but that's what we want. Weak dither just reduces correlation while adding just a little bit of dither noise in the signal. Some of the granulation remains, but it's perhaps not as harmful to sound quality. Increasing dither noise level removes the correlation completely, but at the cost of more dither noise added. The result is (noise free) signal + dither noise.

Now, here is the punch line: Breaking the correlation between signal and noise makes it possible to store distortion-free signal levels below the least significant bit! This is among the counter-intuitive stuff that goes with digital audio and requires better understand of digital audio beyond the basic understanding. We can't increase signal to noise ratio. On the contrary using dither is about increasing noise level, but in audio technical signal to noise ratio doesn't tell everything. The spectral density of noise dictates, how harmful the noise is. Dithering is about being smart with the noise, make it as little as a problem as possible. Yes, we perhaps have dither noise, but the sound of piano attenuates below -96 dBFS levels (16 bit audio) undistorted into the noise and finally becames masked by the noise. That is, if you crank up the volume so much that you are able to hear signals this quiet. Dither noise shaping tries to optimize the spectral density of noise so that it's as inaudible as possible.

Analog sound with (enough) noise is self-dithered and will experience "clean" analog-to-digital transform, but we must live with the original noise level of course.

On another forum, a member there strongly argues that dither results in a loss of clarity, in addition to higher noise. While most of his posts do not make sense he does raise an interesting question. If the trade off with dither is just increased noise (ie tape hiss noise) then why wouldn't dithered 10 bits sound as distortion free as 24bits, just with more noise?

Dither is about adding noise to the signal, but if we for example use shaped dither, instead of having distorted signal with correlating quantization noise, we might have distortion-free signal with inaudible dither noise thanks to smart dither noise spectrum having most of the noise energy at frequencies our hearing is insensitive at.

Loss of clarity? Clarity defined as what? I wouldn't call use of dither properly loss of clarity.

In fact if we truncate and dither 24 bit audio to 10 bits, it IS as distortion free as the original but with "huge" noise.

 

[bigshot]

Dither is fine, but with reasonably well recorded music, I've found that there's not much of an audible difference between properly dithered bounce downs to 16/44.1 and undithered ones. In a lot of cases, I think the difference is probably totally inaudible. It's operating on a very small scale in the grand scheme of things.

 

[gregorio]

Just wondering if someone could provide a rebuttal to this article from a studio.

There are a couple of points which are incorrect but they're not directly related to dither and there are a couple of places where I'd quibble a little with the wording but that's about it. There's nothing substantial to rebut. A couple of points worth baring in mind, the article is talking about mastering and therefore really re-dithering and secondly, the author has normalised the dither examples by 60dB. 60dB equates to about 10bits, so the dither noise you're hearing in the examples is akin to what you would hear if you reduced the bit depth to 6bits.

On another forum, a member there strongly argues that dither results in a loss of clarity, in addition to higher noise. While most of his posts do not make sense he does raise an interesting question. If the trade off with dither is just increased noise (ie tape hiss noise) then why wouldn't dithered 10 bits sound as distortion free as 24bits, just with more noise?

Let's go a whole lot more extreme than 10 bits and let's say just 1 bit. If what the "member" is stating were true, the amount of distortion, loss of clarity and "more noise" would be truly horrendous with just 1 bit? Ever heard an SACD? Surely you couldn't have missed all that horrendous distortion, loss of clarity and ridiculous amount of noise, SACD is just 1 bit? Err no, you missed it because it's not there. There is no audible distortion or loss of clarity even with just 1 bit (SACD). There is though a ridiculous amount of (dither) noise but you can't hear it because it's been "shaped" (redistributed to a freq band starting at around 24kHz)!

G

 

[RRod]

Yeah the -60dB thing is really screw*ed-up, obvious hacking. Ask them what happens when they suddenly play the track with the original gain.

On another forum, a member there strongly argues that dither results in a loss of clarity, in addition to higher noise. While most of his posts do not make sense he does raise an interesting question. If the trade off with dither is just increased noise (ie tape hiss noise) then why wouldn't dithered 10 bits sound as distortion free as 24bits, just with more noise?

Because noise-shaping entails increasing the total energy in certain bands, and at some point you have to shape SO much that you start exceeding the amplitude capabilities of some part of the delivery chain. SACD avoids this by having a huge range of frequencies into which to shape and then lowpassing them off.

 

[castleofargh]

all the soundtracks on the page started playing at once after I told noscript to let it go. ^_^

I personally pay no mind to people claiming that stuff in the least significant bits are ruining anything in any significant way. having to apply the dither to stupidly high levels just to show it ruins the sound is evidence enough that at the usual levels it's not. I take a 10bit dither example as playing the result of PCM track minus MP3 track at really loud levels and go "see!!!! MP3 really messes up the sound!!!".
it's nice if you're curious, but also irrelevant for real life use.

 

[old tech]

Thanks for the replies.

There certainly is a lot of misunderstanding about digital audio in the community, particularly by the layperson. Many talk about quantisation error being digital's Achilles heel (putting aside all the other nonsense regarding audibility of jitter, pre/post ringing etc) but do not appreciate the role of dither, or if quantisation would be audible at all at 16bits.

 

[bigshot]

That's a bingo. Knowing the basics of how things work is a great way to avoid bologna pumped out by audiophools and snake oil salesmen. It doesn't take an Einstein to know that the stuff they are talking about doesn't add up to a hill of beans.

 

[71 dB]

There certainly is a lot of misunderstanding about digital audio in the community, particularly by the layperson. Many talk about quantisation error being digital's Achilles heel (putting aside all the other nonsense regarding audibility of jitter, pre/post ringing etc) but do not appreciate the role of dither, or if quantisation would be audible at all at 16bits.

If you can't admit that you prefer analog sound for it's distortions (i.e. technical inferiority), you have to invent Achilles' heels for digital audio to ease cognitive dissonance. A lot of people misundertand digital audio so badly because the common staircase representations (and to make things worse for visual reasons the amount of steps is small, ~ 4 bit or so) of digital signals are often fundamentally wrong and even if they were correct, what we see isn't what we hear. Visually bad looking things (such as pre/post ringing) might be totally ok for our ears. Digital audio makes one feel he/she understands it when there's a lot to learn and understand. I myself have held some wrong beliefs for years in the past (maybe even today!?) so I know understanding digital audio properly takes some effort.

 

[old tech]

If you can't admit that you prefer analog sound for it's distortions (i.e. technical inferiority), you have to invent Achilles' heels for digital audio to ease cognitive dissonance. A lot of people misundertand digital audio so badly because the common staircase representations (and to make things worse for visual reasons the amount of steps is small, ~ 4 bit or so) of digital signals are often fundamentally wrong and even if they were correct, what we see isn't what we hear. Visually bad looking things (such as pre/post ringing) might be totally ok for our ears. Digital audio makes one feel he/she understands it when there's a lot to learn and understand. I myself have held some wrong beliefs for years in the past (maybe even today!?) so I know understanding digital audio properly takes some effort.

Yes, digiphobia is alive and well amongst certain groups that call themselves audiophiles.

 

[71 dB]

That's a bingo. Knowing the basics of how things work is a great way to avoid bologna pumped out by audiophools and snake oil salesmen.

We live in a world where quite a lot of people are flat earthers. How do you expect people to have even basic understanding about audio when they struggle with the shape of the planet they walk on? People are ignorant with pride, and that makes the life of snake oil salesman and other charlatans easy. If you are a con man you want people to stay ignorant. It's good for you.

 

[71 dB]

Yes, digiphobia is alive and well amongst certain groups that call themselves audiophiles.

I don't mind people liking the sound of vinyl. What we like is not that rational. Vinyl means familiarity one can get fond of: All records have the same additional distortions to create the familiar sound one likes. Even I think there is something cool about the distortions of vinyl and I call myself an advocate of neutral/transparent digital audio. That's why I consider vinyl sound a sonic effect we could use in music production. If metal music or underground techno sounds "better" with vinyl distortions then why not add those distortions to the master and have then in digital formats too? Vinyl recorded on a CD sounds that same vinyl, because CD is so transparent.

 

[old tech]

I don't mind people liking the sound of vinyl. What we like is not that rational. Vinyl means familiarity one can get fond of: All records have the same additional distortions to create the familiar sound one likes. Even I think there is something cool about the distortions of vinyl and I call myself an advocate of neutral/transparent digital audio. That's why I consider vinyl sound a sonic effect we could use in music production. If metal music or underground techno sounds "better" with vinyl distortions then why not add those distortions to the master and have then in digital formats too? Vinyl recorded on a CD sounds that same vinyl, because CD is so transparent.

Yes I know, I have been digitising LPs for some 20 years now, 16/44 for the most part and 24/96 now as I stream music these days rather than play CDs. Once I got the hang of it all my digitised LPs sounded identical to the record being played on the donor turntable, whether it be 16/44 or 24/96. I agree that some music does sound better on vinyl (hence the digitisation) either because more love went into the mastering or the nature of the music suits the way vinyl thickens the sound below 200Hz, eg AC/DC. But when it comes to transparency or the ultimate fidelity of well produced material it is not a contest. The figures do not lie.

 

[71 dB]

Yes I know, I have been digitising LPs for some 20 years now, 16/44 for the most part and 24/96 now as I stream music these days rather than play CDs. Once I got the hang of it all my digitised LPs sounded identical to the record being played on the donor turntable, whether it be 16/44 or 24/96. I agree that some music does sound better on vinyl (hence the digitisation) either because more love went into the mastering or the nature of the music suits the way vinyl thickens the sound below 200Hz, eg AC/DC. But when it comes to transparency or the ultimate fidelity of well produced material it is not a contest. The figures do not lie.

Why do you need 24/96 for streaming? The dynamic range of vinyls is 60 dB ( = 10 bits ) at best so even 16 bits is overkill for digitizing vinyls. Some vinyls do contain ultrasonic information, but that's not for humans. That's for dogs and bats. So, 16/44.1 or 16/48 is all you need for vinyls.

Maybe more love goes to vinyls mastering, but I believe it's actually about the lack of love being hidden by the distortions of vinyl that is the reason more often than not. Bad mastering can sound harsh and cold, but some distortions soften these up and add warmth that should have been in the mastering itself. Especially electronic music can benefit from the additional "effects" given by vinyl. Classical music on the other hand often sounds awesome on transparent digital formats because the sounds of the acoustic instruments are so rich, the spatial information is "real" etc.

 

[old tech]

Why do you need 24/96 for streaming? The dynamic range of vinyls is 60 dB ( = 10 bits ) at best so even 16 bits is overkill for digitizing vinyls. Some vinyls do contain ultrasonic information, but that's not for humans. That's for dogs and bats. So, 16/44.1 or 16/48 is all you need for vinyls.

Maybe more love goes to vinyls mastering, but I believe it's actually about the lack of love being hidden by the distortions of vinyl that is the reason more often than not. Bad mastering can sound harsh and cold, but some distortions soften these up and add warmth that should have been in the mastering itself. Especially electronic music can benefit from the additional "effects" given by vinyl. Classical music on the other hand often sounds awesome on transparent digital formats because the sounds of the acoustic instruments are so rich, the spatial information is "real" etc.

No other reason than that is because I record at 24/96 and I have heaps of storage and excellent signal strength for streaming.

I can't hear a difference between 16/44 and 24/96 and there hasn't been any convincing controlled tests over the past 30 years that demonstrate otherwise, so I don't disagree.