When to Dither?.. please help

  • Thread starter Thread starter Enzo66
  • Start date Start date
E

Enzo66

New member
Yes I've done a search - NO Im none the wiser?

OK - here it is: I always record at 24bit ~41 /

NOW; I have one session in ProTools that has parts from different studios - I have one instrumental track at 16 bit ~41 / I have one sample track at 16 bit ~48 / and all my parts are at 24 bit 41...

Which parts should I be dithering where? /The overall session spec that I want to work in, is 24bit ~41..

--
SO.... One guy said I should be opening up each sound file in its own session at the same spec as the file in question (so he means - open up the 16bit 48/ file in a session with a spec of 16bit 48)/ then bounce that file, dithered at the same spec as my target.. (always 24bit ~41)/ so when I open up the new 'dithered' file, it will sound better matched with what I'M doing (which is always 24bit ~41 as I said earlier) - BUT this would mean bloody files dithered all over the show, and then dithered once again at master output / bounce! That can't be right or good.. HELP?!

Surely you don't have to dither EVERY time you change a file from one spec to another, do you?
 
Dithering is important when you are reducing the bit depth of the file. For example, when you want a stereo mix recorded in a higher bit depth to be converted to the CD standard of 44.1KHz/16 bit. I keep everything where it was recorded until I get to that point, and then convert (and add dither) once.

Dithering simply avoids having the lowest bit toggle back and forth between 0 and 1. The process raises that lowest bit up slightly, improving noise and hopefully making your recording sound **awesome**.
 
Thanks man! SO; if I'm loading a 16bit audio file into my 24bit session.... do I need to dither? (Your're saying 'no' right?) - and also if Im taking another 48k file into my 41k session... do I need to dither??

I see now that its mainly important when you are DOWNgrading a bit file from 24 to 16.. I got that but the above questions still have me flumoxed!
 
in Pro Tools you must ALWAYS dither when you're going to 16 bit or lower.
Even if you may have recorded at a bit depth of 16, Pro Tools LE/TDM utilizes 24 bit audio I/O paths as well as 32 floating point processing and mixing for LE (TDM uses 48 bit).

create a master fader for your session (a good habit to get into anyway) and just add the Dither plugin on the last insert in a typical session.

you cannot have a 48kHz and a 44.1kHz file in one pro tools session...nor can you have a 16 bit and a 24 bit audio file in the session...Pro Tools will convert them to what your session is setup at.
As far as I know, Pro Tools will add dither automatically if you are bringing in a 24 bit audio file into a 16 bit session. I know it does for sure when exporting regions as files.

Really all you need to do is just work at the highest sample rate that you're importing into PT, and the highest bit depth that you're importing in. In other words...just work at 48kHz, 24 bit...then dither down when you bounce. That way nothing is lost, it's all just been converted.
 
Enzo66 said:
Thanks man! SO; if I'm loading a 16bit audio file into my 24bit session.... do I need to dither? (Your're saying 'no' right?) - and also if Im taking another 48k file into my 41k session... do I need to dither??

I see now that its mainly important when you are DOWNgrading a bit file from 24 to 16.. I got that but the above questions still have me flumoxed!

I truely don't think you spent much time in search! There has been countless threads on this BBS about dithering! If you didn't spend that much time, just admit it. I would actually be the first guy to defend your laziness! ;) (really, I scold the people that try to get down on people for not using the search function of this site.....some people feel comfortable asking in the forum, and do, and others are a little more shy or something and use the search....)

You have to UNDERSTAND what dithering is used for. Basically, dithering is used when you REDUCE the bit depth of audio. So, if you were going to try to import a 24 bit file into a 16 bit session, you would want to dither the 24 bit file. I could go on for a bit about WHY you would want to do this, but just take my word for it that it is a good thing, and worthy of your time to do.

So, in your scenario, you want to import a 16 bit file to a 24 bit session. Now, I am not an entrapped ProTools user who is at the mercy of ill-thought out software from them, but in just about every OTHER application around, you may import files of differing bit depths to a session. If ProTools cannot do this rather simply task, just download something like GoldWave, open the file in it, do a Save As, and in the Save Sound As window that opens, under the Attributes: box, select "PCM signed 24 bit, stereo (or mono if it is a mono file), then Save. Vola, you instantly made your 16 bit file a 24 bit file, so now confused ProTools can use it. :)

As to sampling rate differences, again, MANY other programs allow you to import in files that are at a different sample rate than the session is. It would appear that ProTools will not automatically convert the sample rate for you. Such a simple thing to offer, but yet another basic feature that PT doesn't offer, but I regress! :D Again, you could use GoldWave (by the way, I recommend Gold Wave because the demo version is FULL FUNCTIONING with only a restriction on how many things you can do per time you open the program.....it also happens to be a very good editor for audio!). Open the file in GoldWave, go to the Effect menu, and select Resample (it is at the bottom), in the box that opens, select 44100 from the drop down box on the right, then hit OK. You will see a little progress bar indicator while it does it's magic. When it is done, do a Save As! Vola, it is now at the proper sample rate for ProTools to not get confused! ;)

In Sonar, sample rate conversions are automatic if you import and file into the project. Same with bit depth.

If you are a Mac user, enslaved to that horrid platform and the constant expensive upgrades (funny that ProTools ONLY worked on Mac for a long time! You got double screwed! :D), all bets are off with GoldWave, as it is a PC only application. I wouldn't even know what cool program you could use for these simple functions that ProTools should do automatically, but there is possibly something out there you could buy, or you could just do the slow and confusing way of opening then saving in ProTools to get everything so that PT doesn't cripple itself when you try to import a file from another bit depth or sample rate. If you are a PC user, cool! Use GoldWave. But also think about ditching PT and getting an audio application that doesn't have so MANY restrictions, and is probably far more efficient to work in.

In case you haven't noticed from my rather subtle (;)) comments, I am NOT a ProTools fan at all. It constantly amazes me that people will buy that junk! Most of PT's reputation comes from their fully functional software, NOT LE, which is what almost everybody in the low end market is using. One of my major gripes about PT is that unless you spend a LOT of extra money, it is not very compatible with anything other than ProTools. Try to export OMF. ;) Try working with VST's!!!

Also, PT does a LOT of internal bit depth junk in the plugin chain that is bullshit! EACH plugin in a PT session is 24 bit in, 24 bit out! So, the more plugins you have on a single track, the noisier it gets! As I recall, the Aux busses are 24 bit only too, so, if you are compounding bit depth problems by sending a track, Post Fader, on a track with a lot of plugins on it to an Aux Send, that Aux Send audio get's even MORE noise added! That is simply shoddy code! ProTools boasts 48 bit internal resolution, but never fully explains WHERE, and they certainly don't call attention to the places in their application where they don't even retain 32 bit float! :( Too many fuzzy math things going on, and this explains why ProTools session usually sound "funny" to me. They almost always have a signature sound, and I don't like that sound. To each their own I suppose.

Anyway, good luck.
 
It would appear that ProTools will not automatically convert the sample rate for you. Such a simple thing to offer, but yet another basic feature that PT doesn't offer, but I regress! :D
wrong
if you had bothered to read my above post, you would see that even such a "confused software" such as pro tools will do this.

I wouldn't even know what cool program you could use for these simple functions that ProTools should do automatically, but there is possibly something out there you could buy, or you could just do the slow and confusing way of opening then saving in ProTools to get everything so that PT doesn't cripple itself when you try to import a file from another bit depth or sample rate. If you are a PC user, cool! Use GoldWave. But also think about ditching PT and getting an audio application that doesn't have so MANY restrictions, and is probably far more efficient to work in.

again, wrong. see above.


To each their own I suppose.

wow, the only intelligent thing you've said
aren't you getting tired of complaining about software that you don't even use?
You know, I think I'm gonna go start a "I Hate Nuendo" thread. Just because I don't like it. I don't care if other people do, no point in trying to help people use their software...just tell them to buy something else! Brilliant!
 
HAH; well I never said PT didnt convert automatically - it does, but I just got lost in this particular session with so many different file types.. they all run perfectly but this guy planted a seed of doubt by saying I should have dithered each file specifically and I was like WHAT!!!?

As far as PT - well I've used every DAW and setup and I would only ever work in Pro Tools or Nuendo.. they're all great but it's the feeling you get with them. Pro Tools LE 7 Im using is spot on so far, and actually the trademark PT sound we hear - I find to be the best of all the programs.... You're right about Mac's though... and you're right I was lazy!!! AGH!!
 
bennychico11 said:
You know, I think I'm gonna go start a "I Hate Nuendo" thread. Just because I don't like it. I don't care if other people do, no point in trying to help people use their software...just tell them to buy something else! Brilliant!

Please do! I hate Nuendo too! And Cubase!
 
As to the rest, my bad.

But, you can't dispute the 24 bit I/O to each plugin instance, as well as the Aux Send routes, because those are true.
 
Ford Van said:
As to the rest, my bad.

But, you can't dispute the 24 bit I/O to each plugin instance, as well as the Aux Send routes, because those are true.

I really can't speak about that stuff because the whole math part of DAWs is a little hazey with me. Plus, I haven't found much info on host processing and how it all works (including with regards to LE).
But Digi did release a white paper on their TDM cards: http://akwww.digidesign.com/support/docs/WhitePaper_48BitMixer.pdf which I think only explains how the cards work. And from looking at that the entire mixer is 48 bit throughout, so that includes Aux tracks....including the pathway between timeslots. I don't know how everything is affected when you intorduce an RTAS plugin into the equation.

I think there comes a time when you have to throw all the science and math out the window and just use the DAW for what it is there for.
 
Back
Top