Tell me about dithering

Hi Guys!  Brand new to the forum here, but I have been doing the home studio thing on and off for quite a long time. Ive been a cakewalk user since back in the day, but got away from the recording thing for a while.  Thankfully I'm back to it and just recently bought Sonar Platinum and have some general questions that I haven't quite been able to find answers to yet, and I hope you guys can help.
 
Concerning dithering...  I understand how and why to use,  but here is where my confusion lies.   I'm trying to think of the best way to form this question, and really its the dithering plugins that confuse me more than anything.  Let me use an example
 
Lets assume I've been working on a project with everything recorded at 24 bit, 44.1.  I mixdown still at 24/44.1. No mastering has been done yet, no plugins at all on my master buss at this point, and no dithering applied because I'm not doing any kind of conversion.  I've simply just created a stereo mix down of my project in prep for mastering.
 
I then take that 24 bit, 44.1 stereo mixdown and import it into a brand new session where I can master it.  My project would simply have that one stereo track(that is of course still 24 bit/44.1), going to the master buss where I apply all of my mastering plugins.  When it is time to export, the software asks me what sample rate I want to use and if I want to dither.  I select 16 bit, 44.1 and at that point I add dithering through the DAW when exporting.
 
The paragraph above is how I normally approached things, but there are all these dithering plugins out there nowadays that seem to be highly recommended.  Unfortunately they seem to have complicated the issue for me lol.  My question is...  how do they work?  Does the plugin itself actually convert the audio to 16 bit while in use and applying dithering?  I'm having a hard time wrapping my head around what the plugin is actually doing if the source is still 24 bit and it doesnt get converted until exporting out of the DAW....  unless of course the plugin itself converts to 16 bit in realtime.  Does that make sense what I'm trying to ask?  It just would seem to me that when you ask the DAW to mixdown the file, it is going to apply all the plugins in use in the master buss(including dithering if you use a dithering plugin)......but the audio isnt converted yet is it?  I wouldnt think it would convert until it was being exported/saved, and that means the dithering applied before the conversion and not during the conversion.
 
Thanks for dealing with my long winded question!
 

I appreciate the response, and yes this does help!  After reading your article you linked and then reading some instructions on some of the plugins you included in the article...  I think I had the order of events mixed up.
 
I was assuming that the process of dithering was occurring at the exact same time that the bit dept was being reduced.  Sort of like.....as things are being truncated and a rounding error pops up....the software notes the rounding error and says 'add noise here' basically.
 
Reading more though it sounds like they just add general noise into the bottom 8 bits before the conversion, and then once the bottom 8 bits are truncated, the noise thats present helps mask the artifacts.
 
Thats why I was having a hard time wrapping my head around it cause I was thinking to myself....  how can a plugin correct something that hasnt happened yet.  If that makes sense

Some plugins do dither. Not dithering on its own per se but includes it as part of the plugin. One is PSP Xenon which is  a fantastic limiter.  The dithering section is very last in the process and can be switched in or not. The limiter can also be switched out and just the dithering enabled. But it costs $250 for the privilege. I would not go this route unless you are after an excellent limiter. 
 
I use it because it essentially better than most limiters around and then I don't have to tell the DAW to dither at all, Just leave it alone. And let Xenon do it instead.
 
 

Waves limiters do it too.

Anderton
First, I wouldn't worry too much about it unless you're doing classical music or acoustic music with a lot of dynamics. Second, when you export, what you exported from doesn't change. Dithering is typically added as part of the conversion to 16 bits.

 
Good Day all,
 
So, not to worry about dithering something twice? or dithering at all? It seems your article says dont do it twice.
 
So I was wondering same as OP, when using OZONE 7 as a plugin, as last in signal chain, with its dithering option enabled, should I then not apply dithering in Sonar export options? and/or is one way better than the other?
Actually what may not be clear in Ozone is if this option it is just for auditioning how it will sound with dithering applied, or if it will actually be applying it to export, since when exporting out of Ozone it will also ask again whether to apply dithering, so not clear if already using dithering options enabled within Ozone, should I then be declining dithering in export options? (I guess Im reaching out to Ozone users in this case?)
 
Cheers   

Anderton
You may find this article I wrote helpful in de-mystifying what dithering is, and how it works. Unfortunately the link to the audio example is broken, I'll fix it as soon as I find the original files.

Excellent article!
 
I've always thought of dithering simply as a way to make sure that loss of bits doesn't get too obvious. It's akin to if you have a number with two decimal points and you only can use integers: do you truncate (throw decimals away), round to nearest integer, or randomize the rounding and truncating (dithering). I have to say I'm not sure I've ever noticed any difference between the various algos (not that I've especially paid attention)
 

John
Ozone dithers. 
 
https://www.izotope.com/content/izotope/en/search.html?q=dithering+guide  Here is Izotope's Dithering Guide. 

Thanks I should have checked that out, but will now, lol
 
Cheers

Thanks for the insight everyone.
 
I thought of a good way of explaining my thought process(read confusion lol).  I build and maintain a lot of databases for a living.  So...error handling is part of the process in coding this stuff out.  You are basically adding in logic that says 'here is what to do if/when this error happens', and of course.....the errors you might try to account for may or may not happen.
 
I was rationalizing the process of dithering in the same fashion.  I just assumed dithering happened at the exact same time the conversion was going on.  As in "here is where a rounding error happened here.  Apply dithering here to hide that error".
 
So thats why I couldnt understand a plugin application, cause I'm thinking to myself...  how can a plugin 'dither' rounding errors, when the audio isnt yet converted and hasnt yet experienced rounding errors.  Im learning that the dithering/noise is in fact applied first before conversion though and the process doesnt work the way I thought it did.
 
To the plugin questions.  I dont have intentions to go out and buy something specifically for the purpose of dithering.  But I have a couple plugins that do offer it as part of the deal.  Like the waves limiters.  I have a couple of those, and they allow for dithering, and I hear that the IDR dithering process in them is 'supposed' to be pretty good.
 
I wanted to understand it better before I tried using the limiter to apply the dithering vs just letting the DAW do it on export
 
 

John
Ozone dithers. 
 

Thanks again John, there's an example for setting up Ozone/Sonar dither options when Ozone is used within Sonar, (and lots of dithering theory). However, much like not asking for directions until Im lost, (LOL), Ill have to go back into Ozone 7 manual to search for when exporting out of Ozone, (i.e., not using Sonar to master) whether to decline dithering in Ozone export options, if enabled within actual Ozone application signal chain, as I didn't find this specified or 100% clear in this document. At this point Im thinking I should decline dithering in Ozone export options, if Ive enabled it within Ozone application, in assuming it would be similar to selecting "none" in Sonar export/dither options, (unless you have the quick answer?). Albeit, I would like to think that Ozone application should not allow dither to be selected in export options, if it is already enabled within application. (i guess that would be better directed to Ozone developers).
 
Cheers    

What we sometimes call "dithering" is actually a combination of several things:
 
1. dithering - essentially this is just adding noise at a particular level before reducing bit depth. The level and type of noise required to eliminate the distortion caused by quantization (bit reduction) and convert it to noise instead of distortion is very particular - too little noise and you still get some distortion or noise modulation (noise that varies in level with the signal, which is more noticeable than just plain noise); too much noise and you're adding additional noise with no benefit. So if we add noise in a plugin, we don't want to change the signal level after that point in the chain.
 
2. bit reduction - this is the actual reduction in bit depth (quantization). If we do this in a plugin, we don't want any changes of any kind after we do it. IOW it has to be the absolute last thing in the chain - no mixing, EQ, level changes, sample rate conversion or anything else. What happens with a plugin that does this is it outputs a signal where the lowest bits are all empty (zero), so if you truncate those bits when exporting you don't lose anything.
 
3. noise shaping - this happens at the same time as bit reduction and effectively moves the noise/error from quantization (including dither) to frequencies where we are less able to hear it. Generally the difference between advanced "dithering" algorithms with fancy technical names is really noise shaping and not the dithering part, and different noise shaping algorithms are effectively just different EQ curves moving most of the noise to high frequencies with everyone claiming that their EQ curve is better than everyone else's. As noise shaping is done along with bit reduction, the same rules again apply - absolutely no changes of any kind can occur after it if we do it in a plugin.

Hi all,
 
Excellent thread. I enjoyed reading everyone's input, and especially Craig's 'Dithering Demystified' article.
 
I just wanted to point out, mainly because I didn't see it mentioned in any of the discussions of where dithering happens (audio export and various plug-in related considerations), that there is also a dithering option in the Preferences dialog under "Audio - Playback and Recording". The reference guide explains that this is for rendering audio in processes such as bouncing, freezing, or applying effects. It looks like it might also apply for importing audio into a track.
 
Thanks for all the great info!
 
Chris