digi dither preferences

[Riccardo Corbari]

...I still cannot understand how to choose the dither algorhythm for BTD...

[love666]

just insert the dither plug of your choice as the last plug on the master fader
make sure it's set for 24-16 and on the noise-shaping option of your choice if applicable
all the preferences, etc., checking, unchecking doesn't relate to this

[Riccardo Corbari]

just insert the dither plug of your choice as the last plug on the master fader
make sure it's set for 24-16 and on the noise-shaping option of your choice if applicable
all the preferences, etc., checking, unchecking doesn't relate to this

Do you mean that the resulting file will be directly truncated to 16 bit?

So the procedure should be insert the dither plugin in the track and dither to 16 bit, while bouncing I must specify only 44.1 Khz and not dithering to 16 bit?

[PTUser NYC]

What kind of decimation are you hearing? How do you change your mixes to overcome it?

When I enable dither on the master fader, I can easily hear the difference between the 24 bit file, and the 16 bit (of resolution) version.

The mix is not quite as sharp, and when I change the settings on the dither, particularly the noise shaping options, it changes the character of the mix.

I find that if I am hearing the effect early on, and I make my Eq and level decisions with it engaged, that rather than the mix changing when I add the dither, I am mixing to it, and so it sounds "right" with the dither engaged.

When I'm choosing dither, and shaping, I close my eyes, and very quickly click the buttont to go through the options 10 or 15 times. At that point, eyes closed, I don't know which setting is being used, and I use my ears. Then I open my eyes to see what I've selected. Then I repeat the process and if I get the same answer (I do about 95% of the time) then I know its right.

[Mt.Everest]

Does anyone know if Peak or Toast does the truncation with no dither? I have many times "saved as..." in peak from a 24bit file to a 16bit file, hoping that it was just truncating the file. And also, when dragging a 24bit file into Toast to burn to a CD, it automatically converts it to a 16bit file to be ready for CD audio. I wonder if Toast is adding dither or, hopefully, truncating..

Also, I am fully aware or how important dither is theoretically, but I really never try to 'hear' different dithers or noise shapings. I mean, I really dont understand how one can hear stuff that is meant to be happening so low into your noisefloor at a reasonable level. When Im playing back a mix that is averaging at -5dBfs on the master fader after all master bus processing, how am i supposed to hear what is happening at -93dB (or whatever the correct number is)??!

Can someone tell me what Im missing. I will be the first to admit if someone played my mix back for me, and blindly switched between 3 different dither plugs, I would surely have no idea which was which. Hell, I dont even think Id be able to tell if there was a dither plugin even ON the fader at all.

So before telling me my ears arent tuned or golden or whatever, what am I listening for? seriously.. Am I listening for how the mix sounds at pauses or when I can hear a reverb tail clearly, like on a full band rest, or when the mix is fading? I mean really, I need to be guided for what i am listening for. Any tests I can do that will clearly expose how IDR is different than Pow-r different than UV22? How about even a test that would expose how using no dither at all would sound? I HAVE heard the effects of this with BLANTENT examples using a 4 bit word with and without dither so I know what quantization noise sounds like, but damned if I can hear it under normal circumstances.

thanks to all contributors of this post
--it has been quite interesting and brain frying!
MT

edit:
upon reading some of those digital audio papers, I guess what Im missing is that its not the actual dither one is listening for when choosing a dither plug, its its affect on the audio. Still, I dont think I can tell a difference cuz Im not sure what to be listening for.

[bassmac]

I tried A/Bing different dithers, but I just don’t hear it either - at least not enough to worry about it. That’s why I just record and export - using the internal dither.

I even tried the Crane Song dither CD so I could turn the dither way up, but once my distorted guitars come in, it all becomes blurry to me.

I figure it must be more apparent in “undistorted” music, or my ears just suck.

[Seeee]

I need clarification on this point NYCPTUser brought up.

I too record back into pro tools. i mix on a console, etc.

rather than doing a BTD of the 2 track I recorded back in, I have been using the "export function"

is this dithered?

I guess i'll start using the "import dual mono" feature of Peak and dig those dual monos out of the "audio files" folder.

what a pain in the ass if it dithers when you export.

[Mt.Everest]

Heres another question relating to double dithering:

When I record my mixes inside of protools via bussing to a new audio track, I usually have the L2 as the last plugin in my chain before it gets bussed to my new track for recording. I set IDR to 24bit. When this 24bit signal gets bused to my new track, is it needlessly getting dithered again by the PT TDM mixer when its on its way from my AUX master to my record track? Or is it neccesary for that to happen?


Also:

What are you guys using these exported 16-bit files for? when i record my mixes in protools via bussing and not using BTD, I give these 24bit data files to mastering. Then I also export the 24bit dual mono files in protools as stereo 16bit files just to burn cds for clients to take home or for the label people. Does it really matter if those cds are double dithered?

MT

[bassmac]

When exporting 24bit into 16bit, dither and noise shaping is automatically applied. Exporting 24 to 24, nothing is applied.

I also use the L1 on my master aux, but just the limiter - the one without dither/noise shaping. (since it gets done on export)

For pro mastering, I re-record it (with more headroom!), and send the 24bit L/R files. I figure if I can’t even hear dither, I definitely should stay far away from mastering... haha

[PTUser NYC]

I need clarification on this point NYCPTUser brought up...
I have been using the "export function" is this dithered?

Yeah, my understanding is that Pro Tools does that. I understand why too. It makes sense in most applications to dither the outgoing file. I just like to choose (and hear) the dither before the truncation, while I'm mixing, so for me, exporting adds a double dose of uneccesary dither.

I think the issue of dither is poorly implemented in Pro Tools in a couple ways. Clearly, its it used throughout the mixer in thoughtful ways. Pro Tools also avoids situations where word reductions would go un-dithered. The export example is a good case. In all but the special case that we're talking about, you want PT to dither that file. So Digi is being resposible with dither.

My first criticism is that there ought to be a way to disable the export dither. I doubt there are any significant technological reasons why this would be difficualt to do, and my guess is that since its not a TDM or real time process, that it wouldn't require much time to address either. That they didn't consider it when they developed the system, in a time when we were all (except maybe a few) much more ignorant about dither is understandable. Its time to address it now, though.

My other criticism is the way inserts on the Master fader are implemented. Master Fader Plug Ins are Post Fader. This is so that any DSP done by the fader (moves, fades etc) will happen before the dither is applied. This is completely right. The prblem is that all five inserts are prefader, and only one of them, the last one, needs to be. In the meanwhile plug ins (especially compressors) will behave differently as the input level to them changes whenever the master fader moves. Do we really want our records compression to fall apart in the fade? I don't get it.

There are workarounds, but none are quite as good as making only the 5th insert on a Master Fader "post fader". I understand that this is probably a bigger change in code than the first fix I suggested, and might ultimately cost some small extra chip DSP usage too, but it really should be addressed.

What are you guys using these exported 16-bit files for?

Advertising, for one. Song Demos are another example. Anytime that there will not be a Mastering Session

i think dithering is similar to oxygen-free digital cables. someone will claim they make a difference, but most of us know better.

I mean, I really dont understand how one can hear stuff that is meant to be happening so low into your noisefloor at a reasonable level. When Im playing back a mix that is averaging at -5dBfs on the master fader after all master bus processing, how am i supposed to hear what is happening at -93dB

One guy being rude, another being completely nice, but both are asking the same question.

I have good ears, but I think the best reason that I hear the dither so well is that I use Genelec S30C monitors. They have a ribbon tweeter, and sound great. I can really heara lot of detail.

Also, I don't hear the dither per se, but what the effect on the mix is. Often, it sounds like a bit of high end EQ, and the focus around certain things becomes more detailed. I hear it best in the detail of the room ambience around drums and guitars for instance. I don't mean solo'ed either. In fact its harder to hear on individual tracks. Its in the clash of the mix that I hear the most pronounced difference. It can bring the rosin out of bowed strings, or change the throatyness of the snare. Its subtle, but it does have an effect that I can repeatedly hear. 24 bit files have more air to them. 16 bit files can sound great, but hearing the dither while you mix, and adjusting the noise shaping to complement the situation can really help keep the mix detail in a truncated file.