The No.1 Website for Pro Audio
Dithering 101
Old 17th December 2007
  #1
Gear Maniac
 
BushmasterM4's Avatar
 

Dithering 101

Question 1. When I finish mixing a tune I export it as a stereo track in its native form (24/44.1) that it was recorded in. Now since its the same (24/44.1) then dithering isnt required, correct ?

Question 2. After I do that, I reload the stereo mix and use a few plugins to finalize my mix (my weak version of mastering). Now when Im done doing that I export that as a 16/44.1 tune. That way its cd compatable. Now I use the dithering thats an option with my recording software. I use Tracktion 3. My question is some of the plug ins I use also incorporate dithering. Should all be enabled ? Or should I use just one instance of dithering ?

I figure this is the best place to ask this question. I a relative newbee still learning the ropes. Thanks
Old 17th December 2007
  #2
Lives for gear
 
peeder's Avatar
 

Just use one dithering process, and place it immediately before (or during, as the case might be) your truncation from 24 bits to 16 bits.

So if your DAW is dithering when it truncates don't use a plug to dither, unless you can turn off the DAW dithering. If you can turn off the DAW dithering, then use the plug dither as the very last plugin that touches the audio before it is truncated.

I generally like noise-shaped dithers such as Pow-R type 3 or izotope MBIT+.

To clarify for below, you only need to dither when truncating from a higher bit depth to a lower one. Do not dither at any other time, it's just adding noise. Don't do any further processing on the file at 16bit if you can help it; do all your processing at 24 bit prior to dithering.
Old 17th December 2007
  #3
Gear Maniac
 
BushmasterM4's Avatar
 

Quote:
Originally Posted by peeder View Post
Just use one dithering process, and place it immediately before (or during, as the case might be) your truncation from 24 bits to 16 bits.

So if your DAW is dithering when it truncates don't use a plug to dither, unless you can turn off the DAW dithering. If you can turn off the DAW dithering, then use the plug dither as the very last plugin that touches the audio before it is truncated.

I generally like noise-shaped dithers such as Pow-R type 3 or izotope MBIT+.
Thanks for the reply. I can actually shut off any (plugin or DAW). As far as first question. As long as I stay in the same format (bits,sampling) then I dont have to dither ?
Old 17th December 2007
  #4
Gear Addict
 
hackenslash's Avatar
 

Correct. You only need to dither when reducing the bit depth, and then it should be done only once, as the very last thing you do to the audio.
Old 18th December 2007
  #5
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
Most signal processing expands bit depth which then needs to be reduced to 24 bits in order to be written as a file. Truncating generates 10 dB. of extra distortion or dithering adds 3 dB. of noise.

It's your choice...
Old 18th December 2007
  #6
Lives for gear
 
Ben B's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
Most signal processing expands bit depth which then needs to be reduced to 24 bits in order to be written as a file. Truncating generates 10 dB. of extra distortion or dithering adds 3 dB. of noise.

It's your choice...
Bob,

Just curious where you got those figures (particularly the 10 dB part). Dither noise should be within one quantization step of the LSB. If 16 bits yields approximately 96 dB S/E ratio, a single quantization step should be approximately .001 dB, no? Please correct me if I'm mistaken here.

Ben B
Old 19th December 2007
  #7
Gear Maniac
 
BushmasterM4's Avatar
 

Quote:
Originally Posted by hackenslash View Post
Correct. You only need to dither when reducing the bit depth, and then it should be done only once, as the very last thing you do to the audio.
Thanks for the help. Ive actually been double doing it at times
Old 19th December 2007
  #8
Gear Guru
 
Ethan Winer's Avatar
 

Quote:
Originally Posted by BushmasterM4 View Post
My question is some of the plug ins I use also incorporate dithering. Should all be enabled ? Or should I use just one instance of dithering ?
This came up recently at Lynn Fuston's forum and I believe I conclusively proved that dither doesn't matter at all. It's a long thread (it began about something else many pages earlier) so you can jump in HERE and go to Post #182. Or just read my summary where you can also download the test files:

Does Dither Really Matter?

I understand this will probably ruffle some feathers, but I believe it to be the truth. However, I would love nothing more than for someone to prove me wrong! So go for it guys. heh

--Ethan
Old 19th December 2007
  #9
So Ethan, Let me get this correct.
Let's say I take a 1kHz sine wave, apply a 1 min fadeout to black and truncate. Are you suggesting that I don't her zipper noise?
If I take the same scenario and add 1/2 LSB of dither(choose your flavor), doesn't the zipper noise go away?
Now, let's do the same test with music, Don't I get the same outcomes?
I think that a good fundamental knowledge of the digital theory and it's real world application will show the correct way which to deal with bitrate reduction. Are you suggestion that your single example is going to refute the work or people like Stan Lipschitz and Tom Stockham?
This is one of the most egregious examples of psuedo-science I've seen recently.

-mark
Old 19th December 2007
  #10
Ethan

You are usually so scientific?
I don't know how you have refuted
dither's relevance?
how could one "prove you wrong"
because if you or I fail or pass a blind test
it doesn't mean that a process like
dither is always unimportant or not based on
real factors at play?

I mean how does it sound is a totally valid question
yet that is very relative and not so static.
I mean devil's advocate why not dither
if it could be of benefit ?
Old 20th December 2007
  #11
Lives for gear
 
Cellotron's Avatar
 

Verified Member
As Mark mentioned above all you have to do is a few minutes of listening (at cranked levels in a quiet room) to 24bit files whose audio is disappearing slowly into the noise floor (i.e. fades or reverb tails) first at full resolution, then trucated to 16bit, and then with various dither algorithms added when requantizing to 16bit - to very easily hear that the fades and reverb tails are smooth at 24bit, fitzing and sputtering when truncated at 16bit, and at a good compromise of slightly added noise but much smoother than truncated when dithered to 16bits.

SO - does it ultimately make a difference in the sound quality? To many listeners such as Ethan the answer is no. HOWEVER - it takes very little effort to add dither prior to requantization - AND the end result is indeed smoother sounding fades and tails. To me sputtering and fitzing as sound disappears into the noise floor is possibly a (very) subtle clue to the mind that something is recorded and not "in the room with you" - so if it can be avoided with minimum effort - which it indeed can be - then I'm all for avoiding it.

After lots of personal testing - I can state that for myself while the importance of dither choice has been way over hyped by many in the mastering world - to my ear truncation to 16bit simply does not give you anything of benefit whatsoever in comparison to the use of a good dither algorithm does.

Best regards,
Steve Berson
Old 20th December 2007
  #12
Lives for gear
 
Ben B's Avatar
 

Personally, what I find difficult to comprehend is how dither can be of much (if any) benefit when the music is loud. I too have heard the "fitzing and sputtering" (great descriptives, by the way) of quantization noise at the very end of fades, but the rest of the time -- when the music is loud -- does it have any benefit? It's a rare circumstance in recorded popular music that a reverb tail decaying to -96 dB would be fully "exposed" (and not modulated upwards by other simultaneous mix elements at higher levels) for dithering to have its intended benefit in any location other than the very end of a song. While I do see the benefit it has at the ends of fades, I'd love to hear a clear and scientifically valid description of how a word-length-reduced recording benefits from dither when the music is loud.

Anyone have the answer to this?
Old 20th December 2007
  #13
Mastering
 

Quote:
Originally Posted by Ben B View Post
Personally, what I find difficult to comprehend is how dither can be of much (if any) benefit when the music is loud. I too have heard the "fitzing and sputtering" (great descriptives, by the way) of quantization noise at the very end of fades, but the rest of the time -- when the music is loud -- does it have any benefit? It's a rare circumstance in recorded popular music that a reverb tail decaying to -96 dB would be fully "exposed" (and not modulated upwards by other simultaneous mix elements at higher levels) for dithering to have its intended benefit in any location other than the very end of a song. While I do see the benefit it has at the ends of fades, I'd love to hear a clear and scientifically valid description of how a word-length-reduced recording benefits from dither when the music is loud.

Anyone have the answer to this?
Some people can hear the difference and care about the sound quality of dithered product. The differences certainly can be undeniably measured on a distortion analyzer, and qualified psychoacousticians such as Bob Stuart and Jim Johnston would never deny that there is ample psychoacoustic evidence that human beings can detect sonic changes due to dithering or not dithering 24 bit down to 16. I could get into the discussions of the hearing thresholds, Fletcher-Munson and the rest, but the bottom line is that there is enough physical and strong but anecdotal evidence that dithering to 16 bit is very important.

However, the sound of one (undithered) truncation to 16 bit is not going to bite a newbie in the face. Around here, in this studio, however, we are very sensitive to those issues, and we can hear (blind tests) the shrinkage, the losses in depth and space that often occur when the 24-bit signal is reduced to 16 by different methods.

At this point the science is in, it's quite strong. Doubting that dither is important is about as futile as (and quite comparable to) claiming that global warming is not occurring.

As for the loud passages effect, obviously it is more subtle than at low levels, but I claim that I can hear the effects of different dither shapes even for loud rock and roll. Bob Ludwig has also spoken that he hears the effect even with loud music. The theory is obviously weaker there as you'd think that the loud passages would mask it, but there is a lot of space, even in dense rock music, and we think to hear the tails and the ambience better during that space. Ultimately I don't have to prove it to my peers in order for this to be important to me... and the number of luddites who deny it are simply small nuisances.

Researcher Bob Stuart (Meridian) takes this attitude, and I paraphrase, though using quotation marks. "We have not publicly released double blind tests proving that noise-shaped dither has an audible effect, but we have enough theory and confidence and enough of our users have shown it so that we do not have to go through the expense of trying to prove it. If it's important to you, then you should consider that it is audible."

You should consider that avoiding the distortion, soundstage shrinkage and "coldness" of truncation is the evil you are trying to fight even if you cannot yet hear in your environment the benefits of the dither.

As for the other post that people prefer to record in 16 bit because of the sound, I suggest that you consider the long run, which is that multiple generations of expanding to 24 (32 and up) and then back down to 16 take their toll. So you may prefer 16 (for reaasons of euphonic coloration, perhaps) for your mix, but it's a very fragile situation. 16 may be fine for an END product, but it doesn't make as good a source. ALL OTHER THINGS BEING EQUAL I probably can get a better-sounding result from your 24 bit mix (and also higher sample rate) than your 1644. Again, I can only speak from personal experience and that of mastering engineers whose ears I trust, but I can say that universally, the wider the pipeline coming in, the less the loss at the end of a long chain. I call this the "Source Quality Rule", which holds in so many disciplines, not just audio.
Old 20th December 2007
  #14
Lives for gear
 
Ben B's Avatar
 

Quote:
Originally Posted by bob katz View Post
Some people can hear the difference and care about the sound quality of dithered product. The differences certainly can be undeniably measured on a distortion analyzer, and qualified psychoacousticians such as Bob Stuart and Jim Johnston would never deny that there is ample psychoacoustic evidence that human beings can detect sonic changes due to dithering or not dithering 24 bit down to 16. I could get into the discussions of the hearing thresholds, Fletcher-Munson and the rest, but the bottom line is that there is enough physical and strong but anecdotal evidence that dithering to 16 bit is very important.

However, the sound of one (undithered) truncation to 16 bit is not going to bite a newbie in the face. Around here, in this studio, however, we are very sensitive to those issues, and we can hear (blind tests) the shrinkage, the losses in depth and space that often occur when the 24-bit signal is reduced to 16 by different methods.

At this point the science is in, it's quite strong. Doubting that dither is important is about as futile as (and quite comparable to) claiming that global warming is not occurring.

As for the loud passages effect, obviously it is more subtle than at low levels, but I claim that I can hear the effects of different dither shapes even for loud rock and roll. Bob Ludwig has also spoken that he hears the effect even with loud music. The theory is obviously weaker there as you'd think that the loud passages would mask it, but there is a lot of space, even in dense rock music, and we think to hear the tails and the ambience better during that space. Ultimately I don't have to prove it to my peers in order for this to be important to me... and the number of luddites who deny it are simply small nuisances.

Researcher Bob Stuart (Meridian) takes this attitude, and I paraphrase, though using quotation marks. "We have not publicly released double blind tests proving that noise-shaped dither has an audible effect, but we have enough theory and confidence and enough of our users have shown it so that we do not have to go through the expense of trying to prove it. If it's important to you, then you should consider that it is audible."

You should consider that avoiding the distortion, soundstage shrinkage and "coldness" of truncation is the evil you are trying to fight even if you cannot yet hear in your environment the benefits of the dither.

As for the other post that people prefer to record in 16 bit because of the sound, I suggest that you consider the long run, which is that multiple generations of expanding to 24 (32 and up) and then back down to 16 take their toll. So you may prefer 16 (for reaasons of euphonic coloration, perhaps) for your mix, but it's a very fragile situation. 16 may be fine for an END product, but it doesn't make as good a source. ALL OTHER THINGS BEING EQUAL I probably can get a better-sounding result from your 24 bit mix (and also higher sample rate) than your 1644. Again, I can only speak from personal experience and that of mastering engineers whose ears I trust, but I can say that universally, the wider the pipeline coming in, the less the loss at the end of a long chain. I call this the "Source Quality Rule", which holds in so many disciplines, not just audio.

No need to get into hearing thresholds or Fletcher-Munson. I've just never come across any concrete information (as opposed to subjective or anecdotal opinions) on the effect of dithering during loud passages. If you, or anyone else are aware of any good data regarding this online, I'd very much like to read it.
Old 20th December 2007
  #15
Lives for gear
 
peeder's Avatar
 

Well this was interesting so I ran a quick test using Audio Ease Make a Test Tone, Audiofile Sample Manager, and Izotope RX for visualization (these combined cost me $300 purchased direct...beat that for what they all give me).

I created a custom sweep from -145dbFS to 0dbFS at 1KHz, 44.1 sample rate, 24 bits deep, with Make a Test Tone. I imported it into sample manager and reduced bit depth (nothing else) using both "No Dither" option and "MBIT+" options. MBIT+ is izotope's noise-shaped dither, it is a bit simpler looking filter than pow-R type 3 but similar in general shape. I then imported into izotope RX to visualize on the great (!) spectrogram and spectrum analyzer that it sports. I have attached the screenshots below.

What I found was visible distortion patterns up to about -40dbFS on this simple sinewave sweep. So we can probably take it for granted that dither matters at least up to -40dbFS RMS. The other thing I found was that even at levels near 0dbFS, the noise floor was about 10db lower within the most pscyhoacoustically sensitive regions when using the noise-shaped dither vs. none at all. I think that means that accuracy for those frequencies should be greater with noise-shaped dither even at full scale (you might not be able to hear the noise over the loud signal, but won't that noise randomly shift the exact values of the samples slightly?).

In any case I would imagine there are plenty of moments in real-world program material that are below -40dbFS and therefore dither will matter if you want to avoid those (somewhat pretty looking, probably very ugly sounding) inharmonic distortion patterns you can see plotted in the No Dither version.

Note also how the dithered version preserves the sinewave well below the -96db range of 16-bit PCM audio. The undithered version just suddenly shows up with noise, harmonics, etc. I'm not sure all truncation algorithms will work like the one in Audiofile Sample Manager however, you can repeat these tests with the sweep file attached.
Attached Thumbnails
Dithering 101-mbit-spectrogram.jpg   Dithering 101-no-dither-spectrogram.jpg  
Attached Files
File Type: aif -145db to 0dbFS 1khz sweep.aif (1.54 MB, 273 views)
Old 20th December 2007
  #16
Lives for gear
 
peeder's Avatar
 

And oh by the way here's what the 24-bit original file looks like.

If anyone ever wants to see what the advantages of working at 24 bit are, here's an illustration.

(And I have noted what looks like a bit of DC offset down there at the bottom on these custom sweeps....not sure if this is Make a Test Tone or RX doing it... Plus you'd think they'd be able to crop at a zero crossing, which I haven't done here manually.)
Attached Thumbnails
Dithering 101-24bit-original-sweep.jpg  
Old 20th December 2007
  #17
Lives for gear
 
Coyoteous's Avatar
 

I've wrestled with dither for years - most of the time, at normal volumes, I can't hear the difference. But, I accepted it as a practice. It's just one more thing that may make things sound better (or not worse) - maybe not on it's own, though. If you can do 10 or 100 things that improve or preserve audio quality, the cumulative effect could/can be dramatic. I liken it to using good cables. Also, subtle timbral differences (upper harmonics) of loud sources played at loud volumes may be very, very low and still matter (- 90 though? that may be a stretch!). OTOH, dither practice could just be superstition - if so, it's one that doesn't really hurt anything.
Old 20th December 2007
  #18
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
The main thing to understand is that any argument claiming you can't hear truncation at a certain level is an argument that you most certainly can't hear properly implemented dither. (Improperly implemented dither has unfortunately been common in software.) Meanwhile the science and math involved supports its necessity 100%.

Truncation distortion increases 6 dB. per successive truncation because it is correlated with the signal while dither increases 3 dB. per generation because it is random. To me that reason alone is a good reason to dither because every piece of audio a professional touches is going to see a lot of digital signal processing and truncation before it finally reaches the listener's ear.

Myself, I'm not interested in spending hours determining if I or somebody else can hear the effect of dither on every single piece of audio I handle. You can turn the bias off on an analog recorder too but why add unnecessary distortion when there is simply no need to? I have occasionally asked people to re-bounce their mixes with 24 bit dither because they sounded a little "edgy." There has never been a single case where I was wrong about the mix not having been dithered. It can't possibly hurt and might help.
Old 20th December 2007
  #19
Lives for gear
 
Ben B's Avatar
 

Quote:
Originally Posted by Coyoteous View Post
Also, subtle timbral differences (upper harmonics) of loud sources played at loud volumes may be very, very low and still matter
True, but if they (the subtle tonal differences) are occurring at the same time as something louder is also present, the audio voltage present at the moment that sample is taken would be much higher, and not prone to the same degree of quantization error as if it were residing near the LSB all by itself. To the equipment or software, a sample is a sample is a sample.

I'm not trying to be stubborn here, nor do I wish to invalidate anyone's opinion or experience. I just feel that this is an area that could be better understood by most of us, myself included. Let's keep this discussion going.

I was thinking of trying a test with a loud music signal. I would take a part of a typical 24 bit unmastered rock mix, and bounce it to disk twice (converting to 16 bits), once with dither applied and once without. During these bounces, the latency of the dither plug-in would be compensated for.

I would then import both bounces into a 16 bit session, and invert the polarity of one of them. I would then sum the signals, and look at any noise or distortion products on a spectrum analyzer. Under these conditions, I'd expect to recover the original dither noise, plus some harmonics that were generated by the truncated, un-dithered bounce.

I'll probably do this test several times with other types of signals as well (speech, pink noise, tones, etc.), to see if any broad conclusions can be drawn.

Any other ideas?
Old 20th December 2007
  #20
Lives for gear
 
Ben B's Avatar
 

Something else just occurred to me... If it would be possible to recover these noise and distortion products, could they themselves be polarity reversed, and recombined with the first dithered 16 bit bouce? In theory, doing this would remove (cancel) the noise and harmonic distortion generated by the process, but the benefit of the dither would be retained (because the file is already 16 bit, and was bounced with dither applied). This combination could be re-bounced, and kept at 16 bit resolution.

I'm absolutely sure there has to be a flaw in this thinking somewhere.

Ben B
Old 20th December 2007
  #21
Lives for gear
 
Ben B's Avatar
 

Never mind, I just realized the flaw, which is that the dithered version already lacks the aforementioned distortion. Combining the polarity-reversed noise and distortion products would indeed remove the dither noise, but reintroduce the distortion products.

Forgive me, I'm not fully caffientated yet.

-Ben B
Old 20th December 2007
  #22
Motown legend
 
Bob Olhsson's Avatar
 

Verified Member
You can potentially use pseudo-random noise and mix it in inverted after quantization. It is part of the Pacific Microsonics patent.
Old 20th December 2007
  #23
This may just be me, but I I have the sneaking suspicion that most of the people that are saying dither doesn't matter weren't working in the early days of digital when Sony PCM-1600 and 1610 masters were being made on DAE-1100's without dither.
I'll never forget the eye-opening experience of listening to F1 recordings we made as back-ups to the 1610 that sounded MUCH better than the master. The difference was Dither.
All the best,
Mark
Old 20th December 2007
  #24
Gear Addict
 
Nishmaster's Avatar
 

I think what people fail to understand concerning dither is that at any given point in your program material, there is a certain probability that individual samples may fall at a very low amplitude LSB area regardless of how loud the material is. When truncating instead of dithering, you are effectively bashing those samples to digital zero, thus generating wideband distortion that gets worse as you approach the Nyquist limit and your probability of low amplitude signals increases and number of samples per a given frequency decreases (i.e. sampling the rising/falling edge of a waveform near the zero crossing).

This is that "edge" you hear, and it is completely audible and measurable. Intrinsically, louder program material has a lower probability of low amplitude samples, so it may not be as audible as quieter sources. However, as the RMS of the signal is decreased, the probability of low amplitude samples is increased until it is nearly 100%, resulting in zippering/sputtering/fuzzing/etc. Dithering combats this of course by randomizing slightly the amplitude. The point at which zippering occur and where the signal is buried in dithering noise may be close to the same, but the audible artifacts of white noise are not as objectionable as the distortion products caused by truncation.
Old 20th December 2007
  #25
Gear Guru
 
Ethan Winer's Avatar
 

Heh, I knew this would ruffle feathers. And that's good! If we don't actually test stuff scientifically we'll never know what matters and what doesn't. I do agree with the comment that dither is so easy to do that it's silly not to. But that's not the same as saying it's critically important, or even audible, which is the conventional wisdom.

Quote:
Originally Posted by mpdonahue View Post
Let's say I take a 1kHz sine wave, apply a 1 min fadeout to black and truncate. Are you suggesting that I don't her zipper noise?
Do you mean at normal playback levels, without cranking the volume way up just to hear it?

Quote:
If I take the same scenario and add 1/2 LSB of dither(choose your flavor), doesn't the zipper noise go away?
Try it and report here what you find.

Quote:
Are you suggestion that your single example is going to refute the work or people like Stan Lipschitz and Tom Stockham?
Well, it's not just my one example. And I agree that dither reduces artifacts. What I question is whether those artifacts are actually audible.

Quote:
Originally Posted by mahasandi View Post
how could one "prove you wrong" because if you or I fail or pass a blind test it doesn't mean that a process like dither is always unimportant
What else would it mean if you, and me, and everyone else, cannot pick out dithered versus non-dithered in a blind test using normal (not contrived or overly soft) program material?

Quote:
why not dither if it could be of benefit ?
Absolutely! I'm not arguing against dither. I'm merely arguing that it's not as important as most people believe it is.

Quote:
Originally Posted by Cellotron View Post
As Mark mentioned above all you have to do is a few minutes of listening (at cranked levels in a quiet room)
Right, "cranked levels" being key.

--Ethan
Old 20th December 2007
  #26
Gear Addict
 

Quote:
Originally Posted by Ethan Winer View Post
Do you mean at normal playback levels, without cranking the volume way up just to hear it?

What I question is whether those artifacts are actually audible.
If they're audible at loud levels, they are audible, right?


Quote:
Originally Posted by Ethan Winer View Post
Absolutely! I'm not arguing against dither. I'm merely arguing that it's not as important as most people believe it is.
No, that's not what you were arguing. Here's your claim earlier in the thread:

Quote:
Originally Posted by Ethan Winer View Post
I conclusively proved that dither doesn't matter at all.
Big difference, claiming to prove conclusively that it doesn't matter at all, and arguing that it's not as important as most believe.

Old 21st December 2007
  #27
Lives for gear
 
peeder's Avatar
 

Quote:
Originally Posted by Bob Olhsson View Post
You can potentially use pseudo-random noise and mix it in inverted after quantization. It is part of the Pacific Microsonics patent.
What's the patent number and when does it expire? This sounds like an interesting idea.

I wonder if it could be done one-way, or whether it would need to be applied at the decoder, like Dolby.

Oh and btw Ethan I bet I could write code to look at the distribution of your LSB and find where your dither was applied and not. Of course you can't hear an LSB with all that jangling going on.
Old 21st December 2007
  #28
Lives for gear
 
Ben B's Avatar
 

Quote:
Originally Posted by Nishmaster View Post
I think what people fail to understand concerning dither is that at any given point in your program material, there is a certain probability that individual samples may fall at a very low amplitude LSB area regardless of how loud the material is. When truncating instead of dithering, you are effectively bashing those samples to digital zero, thus generating wideband distortion that gets worse as you approach the Nyquist limit and your probability of low amplitude signals increases and number of samples per a given frequency decreases (i.e. sampling the rising/falling edge of a waveform near the zero crossing).

This is that "edge" you hear, and it is completely audible and measurable. Intrinsically, louder program material has a lower probability of low amplitude samples, so it may not be as audible as quieter sources. However, as the RMS of the signal is decreased, the probability of low amplitude samples is increased until it is nearly 100%, resulting in zippering/sputtering/fuzzing/etc. Dithering combats this of course by randomizing slightly the amplitude. The point at which zippering occur and where the signal is buried in dithering noise may be close to the same, but the audible artifacts of white noise are not as objectionable as the distortion products caused by truncation.
Ah, now that's a very good explanation! Indeed, no matter how complex the wave, there have to be zero crossings, and any samples occurring at or near those zero crossings are susceptible to increased quantization error and thus the accompanying distortion. That actually makes a great deal of sense -- thanks!
Old 21st December 2007
  #29
Lives for gear
 
Watersound's Avatar
 

Ok sorry if this is a dumb question: When I complete a mix in Digital Performer at 24/44 I then bounce it down to 16/44 to burn to CD (when not having it professionally mastered). Would DP not be dithering automatically? Do I need to buy a separate dithering program? If so what do you guys recommend? Thanks!
Old 21st December 2007
  #30
Gear Guru
 
Ethan Winer's Avatar
 

Quote:
Originally Posted by twwalsh View Post
If they're audible at loud levels, they are audible, right?
It depends on how you define loud. If you play a track with the volume as loud as you can reasonably stand, and cannot hear when dither is used or not, then dither doesn't matter. If you have to turn up the volume even more than that to tell, then Yes it's audible in that one contrived situation but still doesn't matter in practice.

Rather than argue, I urge you to download my test file linked earlier. Its low recorded level is as favorable to dither as I could make it, so let's see if you can identify which places use dither. Then we'll both know for sure. heh

Deal? And not just you, but everyone else here who believes dither matters.

--Ethan
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump