The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Good dither practices, what are yours? Dynamics Plugins
Old 2nd February 2017
  #241
Motown legend
 
Bob Olhsson's Avatar
 

Masking depends on both the particular frequency spectrum involved and the holes from damage to an individual's hearing that reduces masking at certain frequencies. (We all begin to acquire such damage from the moment we are born.)

It is not linear, and in fact, noise can actually preserve low level information. "You can't hear anything that far down" is a common but ignorant excuse for lazy engineering.
Old 2nd February 2017
  #242
Lives for gear
 
Deckdaddy's Avatar
I will try it on my mext mix, and do one pass without dither on the outputs. Bummer that I don't get the Airwindows to work, the controls/UI doesn't show up, and bummer that the Kazrog wasn't great for stereo. Feels a bit "heavy" to drop my Ozone 7 on every channel just for the dither, don't know how it affect latency/CPU.
Old 2nd February 2017
  #243
Motown legend
 
Bob Olhsson's Avatar
 

Airwindows need no controls so there aren't any!
Old 2nd February 2017
  #244
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Deckdaddy View Post
I will try it on my mext mix, and do one pass without dither on the outputs. Bummer that I don't get the Airwindows to work, the controls/UI doesn't show up, and bummer that the Kazrog wasn't great for stereo. Feels a bit "heavy" to drop my Ozone 7 on every channel just for the dither, don't know how it affect latency/CPU.
Yeah, the Airwindows dithers were working. You should never have to open the window, there's nothing in it: it's blank. All of the dithers including the just-released 'Not Just Another Dither/CD' are the opposite of 'heavy' and incur next to no resources at all: there is no GUI and in fact no UI, you just put them in and they work (at 24 bit).

Exception: 'NotJustAnotherCD' works at 16 bit, and is otherwise identical to NotJustAnotherDither. The other ones I've done lately are all 24 bit.

(I'll be putting together some videos today: I have the release of those 'boutique' high performance ones, and I also did a refinement of PaulDither for Bob O it's called DoublePaul and all it does is extend the 'FIR filter' of DoublePaul slightly, so it continues to be perfect TPDF dither but skews just a teeny bit more towards noise in the highs, and quieting the noise in the mids. Looking forward to your reaction, Bob!)
Old 2nd February 2017
  #245
Lives for gear
 
Deckdaddy's Avatar
Perfect! I saw it added something on the meter but thought it should be at least an option of bit, I wondered if it autodetected the internal DAW resolution. Will def try it then, should be fast and easy to just assign it.
Old 2nd February 2017
  #246
Lives for gear
 
Deckdaddy's Avatar
So then I have to adjust track send volume before the dither in Reaper and let the output gain be 0 right?
Old 3rd February 2017
  #247
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Deckdaddy View Post
So then I have to adjust track send volume before the dither in Reaper and let the output gain be 0 right?
Sure, it goes straight to the output. Unless you're making one of my 24 bit control-less ones do 16 (or 8) bit. You do that with the BitShiftGain plugin: cut volume before the dither, increase after. This also works with competing/alternate dithers, at least those that truncate as part of the algorithm, so you can listen to all sorts of dithers 'up close' as it were.
Old 3rd February 2017
  #248
Gear Maniac
 
Yuri Korzunov's Avatar
 

Quote:
Originally Posted by Audiop View Post
If we can not measure those I think it makes sense to say that the distortion is gone (not only masked or burried in the noise).
Dither looks like masking.

One man recommended me don't say about dithering as masking for avoiding misleading.

So I preffer say as "smooth distortions". Like painter's finger smear pencil at paper. Before was 2 lines at white paper. After: only grey blur.
Old 3rd February 2017
  #249
Motown legend
 
Bob Olhsson's Avatar
 

It de-correlates the gating so that it sounds like hiss. Correlation with the signal is the difference between noise and distortion.
Old 5th February 2017
  #250
Lives for gear
 
Deckdaddy's Avatar
So today i used Airwindows Tape Dither on my sends to the DAC channels, made a mix on the console, recorded the stereo mix both with and without the dither on the console feeds. I think it actually took off a digital edge and made it more organic, more real, less computer, and the edge on the snare drum got smoother and less crunchy. I think it's crazy that it does this when the stuff is down really low at 24 bits, but it's ridding some harsh digital distortion so why not. I also think I heard a similar more full/whole/organic/smoother sound from adding a dither to the monitoring DAC feed after the Sonarworks correction plugin. I mailed Sonarworks and Ref3 do not have a dither.

These are just initial impressions, but I'll keep on trying this out as I liked the results. I agree with Airwindows' Chris that tape hiss is a more relaxing noise than the straight high passed ones.
Old 5th February 2017
  #251
Motown legend
 
Bob Olhsson's Avatar
 

When you don't dither, the bottom bits make the sound of a chattering noise gate. Dither makes the sound of that cutoff sound like hiss.
Old 5th February 2017
  #252
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Deckdaddy View Post
These are just initial impressions, but I'll keep on trying this out as I liked the results. I agree with Airwindows' Chris that tape hiss is a more relaxing noise than the straight high passed ones.
It's important to note that it's still TPDF dither, much like regular TPDF, PaulDither and so on. Turns out you can get many different sounds out of simple TPDF if you wish: glad you liked that one
Old 6th February 2017
  #253
I have a hybrid system and the way I mix is, all of my Pro Tools channels are routed to a dead internal bus, then I use sends to route the audio out to the various analog channels I have for processing and summing, then everything gets brought back into two channels in pro tools for further mix bus work, then done. So I should have a 24bit dither plug at the end of every pro tools channel right, since all audio is routed out to analog channels? does it matter what level the fader is at? I thought I saw someone say it has to be at unity gain?

One time I tried to just put dither plugins on the master tracks to all my analog outputs but the noise was wayyy too loud, and I'm assuming that's because dither is a fixed amplitude?

assuming I'm supposed to put dither on every track since it all goes out to analog, then I guess the next step is dither on the monitor path? is that not the same thing as the last plugin on the mixbus? And for the I'm doing now, I never bounce to 16 bit, just directly to mp3 with the pro tools bounce to disk function. would I use a 24 bit dither for that final dither plug? 16? do I bounce to 44/16 with 16 bit dither then bounce to mp3 after that with no dither since no changes are being made?
Old 6th February 2017
  #254
Motown legend
 
Bob Olhsson's Avatar
 

If the noise was way too loud, something was broken!
Old 6th February 2017
  #255
Quote:
Originally Posted by Bob Olhsson View Post
If the noise was way too loud, something was broken!
I don't think so. I think it was a gain staging thing. Let's say I had put a dither plugin directly on a track that I had lowered by 12 db on the fader, the dither would then also be 12db lower. However if I put the dither directly on the master track of the analog output I sent the original track to, then the dither never gets lowered by 12db, however it does get the 12 db boost somewhere down the line when I try and make up that gain. Do that for several tracks whose final.fader before going out to analog was minus 10 or 12 db, and you're basically getting a lot of extra db of dither. I need to place them on the original track so they follow the same gain stage, instead of directly on the master tracks of the analog channels themselves, if that makes sense. But what about the rest of the questions? I'm still a bit confused.
Old 6th February 2017
  #256
Airwindows
 
chrisj's Avatar
No, pretty sure they should be on the master tracks, which are probably 24 bit if that's what your converters use.

If the noise was way too loud, maybe you're using an awful lot of compression and/or distortion! If you're just setting levels and EQ in the analog domain it should be really hard to bring up the noise on the channels unless your real levels are incredibly low (way lower than 12 db)

Dither goes just before the converter and it's a fixed level, but it ought to be really, really quiet.
Old 6th February 2017
  #257
Quote:
Originally Posted by chrisj View Post
No, pretty sure they should be on the master tracks, which are probably 24 bit if that's what your converters use.

If the noise was way too loud, maybe you're using an awful lot of compression and/or distortion! If you're just setting levels and EQ in the analog domain it should be really hard to bring up the noise on the channels unless your real levels are incredibly low (way lower than 12 db)

Dither goes just before the converter and it's a fixed level, but it ought to be really, really quiet.
I think I found out why there was so much noise last time. I think I had it on 16 bit and not 24, which was much louder. I just downloaded your dither plug and put it on the analog outputs and am going to see if I can hear the difference.

What about bouncing from 24 bit to mp3.. how should i be dithering on the master bus when doing that? 24 but dither? 16? Should I bounce to 16/44 first with dither then convert to mp3 with no dither?
Old 6th February 2017
  #258
Lives for gear
 
KevWind's Avatar
Quote:
Originally Posted by chrisj View Post
No, pretty sure they should be on the master tracks, which are probably 24 bit if that's what your converters use.

If the noise was way too loud, maybe you're using an awful lot of compression and/or distortion! If you're just setting levels and EQ in the analog domain it should be really hard to bring up the noise on the channels unless your real levels are incredibly low (way lower than 12 db)

Dither goes just before the converter and it's a fixed level, but it ought to be really, really quiet.
Chris I see on your web sight AU and VST so no Pro tools AAX ?
Old 6th February 2017
  #259
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Obe1 View Post
I think I found out why there was so much noise last time. I think I had it on 16 bit and not 24, which was much louder. I just downloaded your dither plug and put it on the analog outputs and am going to see if I can hear the difference.

What about bouncing from 24 bit to mp3.. how should i be dithering on the master bus when doing that? 24 but dither? 16? Should I bounce to 16/44 first with dither then convert to mp3 with no dither?
If you make a 24 bit file, use 24 bit dither. Personally I'd make the mp3 from the 24 bit file, why intentionally make it from the CD quality file? That only causes there to be more noise, which is harder to compress. Go with the 24 bit (or directly out of the DAW, if it'll let you).

And yes, no AAX or TDM. AU and VST are accessible to me
Old 6th February 2017
  #260
Quote:
Originally Posted by chrisj View Post
If you make a 24 bit file, use 24 bit dither. Personally I'd make the mp3 from the 24 bit file, why intentionally make it from the CD quality file? That only causes there to be more noise, which is harder to compress. Go with the 24 bit (or directly out of the DAW, if it'll let you).

And yes, no AAX or TDM. AU and VST are accessible to me
I usually just check the mp3 option when bouncing to disk, so in that case 24bit dither at the end of the master bus chain?
Old 6th February 2017
  #261
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Obe1 View Post
I usually just check the mp3 option when bouncing to disk, so in that case 24bit dither at the end of the master bus chain?
If the program makes a 24 bit temp file. I remember Logic doing that, but it's not a rule. See if the process generates a WAV or AIFF file along the way: if so, then yes.
Old 6th February 2017
  #262
Quote:
Originally Posted by chrisj View Post
If the program makes a 24 bit temp file. I remember Logic doing that, but it's not a rule. See if the process generates a WAV or AIFF file along the way: if so, then yes.
OK I'll look into that, I'm on pro tools 12. Also, I heard the master tracks in pro tools are 48 bit.. still ok to put 24bit dither on that kind of track? last question I swear!
Old 6th February 2017
  #263
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Obe1 View Post
OK I'll look into that, I'm on pro tools 12. Also, I heard the master tracks in pro tools are 48 bit.. still ok to put 24bit dither on that kind of track? last question I swear!
You mean, what gets written to disk as a temp file? If so, don't bother, either way it's glorious overkill and you wouldn't be using a 24 bit dither on it anyway
Old 6th February 2017
  #264
Quote:
Originally Posted by chrisj View Post
You mean, what gets written to disk as a temp file? If so, don't bother, either way it's glorious overkill and you wouldn't be using a 24 bit dither on it anyway
No, aside from when the mix is printed, the master faders that I put the dither on, the ones that go out to the analog channel, are of a different architecture than the aux and audio tracks in pro tools.. I've read they are 48 bit. I wasn't sure if placing a 24 bit dither plugin on a 48 bit master fader would have a different result (possibly a negative one) than if it was on an aux or audio track. I suspect not, but I really dont understand this stuff in depth.
Old 6th February 2017
  #265
Airwindows
 
chrisj's Avatar
Quote:
Originally Posted by Obe1 View Post
No, aside from when the mix is printed, the master faders that I put the dither on, the ones that go out to the analog channel, are of a different architecture than the aux and audio tracks in pro tools.. I've read they are 48 bit. I wasn't sure if placing a 24 bit dither plugin on a 48 bit master fader would have a different result (possibly a negative one) than if it was on an aux or audio track. I suspect not, but I really dont understand this stuff in depth.
That part doesn't matter. Dither is just audio (or interaction with audio). Most of the audio tracks I put dither on are 32 bit floating point, and Cubase offers 64 bit floating point. The only part that matters is (a) it's big enough for the dither to even be present, and (b) what the output file is going to be.

It can be a 48 bit master fader all day long, and good for it (my internal plugin busses are 64 or 80 bits) but your DAC is still 24 bit and the files are probably going to be 24 bit or 16 bit.
Old 6th February 2017
  #266
Quote:
Originally Posted by chrisj View Post
That part doesn't matter. Dither is just audio (or interaction with audio). Most of the audio tracks I put dither on are 32 bit floating point, and Cubase offers 64 bit floating point. The only part that matters is (a) it's big enough for the dither to even be present, and (b) what the output file is going to be.

It can be a 48 bit master fader all day long, and good for it (my internal plugin busses are 64 or 80 bits) but your DAC is still 24 bit and the files are probably going to be 24 bit or 16 bit.
Awesome. Thank you for clarifying all this. I now have a better understanding of dither.
Old 9th February 2017
  #267
So after looking around on the forums and looking through my Apollo Quad manual, I realized the Apollo can do real time sample rate conversion. This got me to thinking maybe it would be worthwhile for me to set my Burl B2 ADC at a higher sample rate to capture the mix bus output of my SSL X-desk, then when it goes back into the Apollo it will bring it back to 44.1 and I won't have to bog my CPU down by running the whole session at the higher sample rate. The only downside is that I don't get to use the Burl as the master clock, but if the benefits outweigh the cons I'm thinking why not?

My question is, should I then put a dither plug at the beginning of the chain when it comes back into Pro tools from the burl after the live SRC? Or does the internal bit processing of Pro Tools make that unnecessary?
Old 9th February 2017
  #268
Lives for gear
 

Quote:
Originally Posted by Obe1 View Post
So after looking around on the forums and looking through my Apollo Quad manual, I realized the Apollo can do real time sample rate conversion. This got me to thinking maybe it would be worthwhile for me to set my Burl B2 ADC at a higher sample rate to capture the mix bus output of my SSL X-desk, then when it goes back into the Apollo it will bring it back to 44.1 and I won't have to bog my CPU down by running the whole session at the higher sample rate. The only downside is that I don't get to use the Burl as the master clock, but if the benefits outweigh the cons I'm thinking why not?

My question is, should I then put a dither plug at the beginning of the chain when it comes back into Pro tools from the burl after the live SRC? Or does the internal bit processing of Pro Tools make that unnecessary?
Sample rate conversion is not truncation or bit-depth change. You are confusing the two. Dither is for bit-depth change.

(also, I really don't get what you are trying to accomplish because what you are describing sounds like a worst end result and not a better end result)
Old 9th February 2017
  #269
Quote:
Originally Posted by pentagon View Post
Sample rate conversion is not truncation or bit-depth change. You are confusing the two. Dither is for bit-depth change.

(also, I really don't get what you are trying to accomplish because what you are describing sounds like a worst end result and not a better end result)
I'm not completely sure to be honest. I thought I was using the same logic behind why people record at a higher sample rate.. but maybe that is only relevant if the session stays at that sample rate? I assumed capturing the analog 2bus mix coming out of the summing mixer at a higher sample rate would somehow make for a more accurate wordlength calculation when converted back into digital into pro tools. But then again I guess if the higher sample rate is a multiple of 44.1 (my daw sample rate) then it really won't matter since the calculation to digital will end up being the same.

I really don't know, was just trying to make the most of the gear I have. I did bounce a version with the Burl at 192 on internal clock, and the Apollo on internal as well with live SRC on and I can at least say the difference was not worse in my opinion.. it's probably just my imagination, but it seems like the midrange is more present and the depth seems more sophisticated, but that very well could, and probably is, the result of the clock change.
Old 9th February 2017
  #270
Lives for gear
 

If you are going to continue processing/mixing the signal at a higher sample rate until the final stereo file which ends up at 44.1, then there's a benefit at doing higher sample rates.
If you are going to capture with the Burl at 192k and then immediately downsample to 44.1 before continuing to process/mix, you are not doing a good thing. Better to let the Burl capture at 44.1 than have a "live" SRC to 44.1. Much to the confusion of many, many people, SRC done today is not a simple division if things happen to be multiples. There is oversampling and filters applied (and resulting pre-ringing and cut-off artifacts.)
And all of this has nothing to do with dither unless for some reason you are not working at 24 bit at this point (and then possibly doing SRC at 32 bit float which also doesn't need dither since you are going up not down.)

Last edited by pentagon; 9th February 2017 at 06:29 AM..
Topic:
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