I have looked at other threads, and I cannot seem to find an answer to my question. To be completely honest, dithering is something I do not understand whatsoever.

From what I have read, POW-3 dithering seems to be the best option for exporting, especially since I am going from 24bit-48Hz to 16bit-44100Hz for CD/MP3 purposes (please correct me if POW-3 is not the best setting).

What is confusing me is the dither option when you are bouncing all your audio to a stereo track for mastering. If you are not dithering down the audio to a different bit depth or Hz value, why would the dither option be there?

In either case, what is the best option to choose there? Triangular is picked by default

Hi Ambient,
I'm not sure if it's right or wrong, but I export my audio to a Master 2-track file UNDITHERED, in my case 24-bit due to hardware limitations.  I then use iZotope Ozone5 to master that track and utilize the dithering facility in it to master my CD track or mp3...to 16-bit/44.1Hz.

Dithering is quite nebulous to me also, but the iZotope Ozone manual on "Dithering" is available for downloading, and I can't imagine having more to say about the subject than what you will find in that manual.

The Sonar Power! book by Scott Garrigus also has some good advice on the subject.

Good luck!

Dither is unnecessary for internal bounces, since you're going from 32-bit to 32-bit.

Whether or not it's necessary when exporting to a 24-bit file has been the subject of some debate. Personally, I do not feel that it is necessary unless you're going to 16 bits from a higher bit depth, and even then the algorithm you choose is pretty much irrelevant.

BTW, I think we should start a campaign to stamp out the phrase "dither down". I see it used everywhere, and find it annoying. There is no such thing as "dithering down", since adding dither noise and reducing the word length are different things entirely. (And I'm not picking on you, Ambient. You're upfront about not understanding dither, and therefore get a pass. It's when I hear the phrase from experienced audio professionals who should know better that I get bugged.)

Ambient: If you are going to just do a quick project mix out of Sonar to burn to CD or whatever, pow-r 3 is what I use. Why? Because the help files gave the best description for that option. LOL! Supposedly some people can actually hear the difference each dithering option gives you. I never really tried it to see for myself, but when exporting out of Sonar for quick ideas or whatever, that's the one I use.

Now if I'm going to master something, what Tony told you is how I go about it also. The reason being, you should master your tunes in the bit and sample rate they were recorded in and dither/convert sample rate last. So you'd want to come out of Sonar like that as well with no dithering options enabled. Once you master the song, then you do that stuff.

Dithering basics: in a nutshell without getting deep into the science end of it, when you reduce a bit rate without dither, the numbers don't jive and it gives you errors. Most people will not hear these errors but some claim they can. The errors can come by way of sound degradation or even distortion....so they say. I've never heard it when I've chosen NOT to dither something for quick ideas. When you dither, it supposedly adds in a little random distortion but it de-correlates it and rounds out the errors and any distortion will be inaudible.

Another way to put it in simpler terms: You have a 24 bit file. You change the bit rate to 16 without dithering....numerically speaking, (if you choose to look deeper into the science) it will be wrong math wise. Dithering corrects the math when you reduce a bit rate.

Do we need to dither? I'll be brutally honest here. Personally, the jury is still out for me. I hear no difference with these supposed errors. Do I dither? Yeah, but only because everyone else does and I don't hear it degrading my audio...so I do it too even if I can't hear what it's actually doing. I have pretty good ears and get hired for jobs based on my exceptional hearing. However, I cannot tell the difference between a file 16 bit dithered and a 16 bit undithered wave file. If someone CAN, I would love for them to show us what and how to listen for it.

Hope this helps a bit but that's how I've always understood dithering.

-Danny

bitflipper


Dither is unnecessary for internal bounces, since you're going from 32-bit to 32-bit.

Whether or not it's necessary when exporting to a 24-bit file has been the subject of some debate. Personally, I do not feel that it is necessary unless you're going to 16 bits from a higher bit depth, and even then the algorithm you choose is pretty much irrelevant.

BTW, I think we should start a campaign to stamp out the phrase "dither down". I see it used everywhere, and find it annoying. There is no such thing as "dithering down", since adding dither noise and reducing the word length are different things entirely. (And I'm not picking on you, Ambient. You're upfront about not understanding dither, and therefore get a pass. It's when I hear the phrase from experienced audio professionals who should know better that I get bugged.)

P.S. and make sure you guys listen to this man up here ^ as he is extrememly exceptional in knowing about this stuff. :)
 
Bit, just curious...when people say "dither down" my thoughts are, since they are reducing the bit rate down and dithering, wouldn't "dither down" be acceptable? Meaning...dithering to a lower bit rate, so "dither down" for short? That's how I've always looked at it. I just want to be sure though because the last thing I want to do is say "dither down" and annoy you. :)
 
-Danny

If you are not dithering down the audio to a different bit depth or Hz value, why would the dither option be there?

To confuse you:)


The bounce feature could and is useable for your master if you're mastering in Sonar.

Triangular is suppose to be better than Rectangular or so the Help states.

Thank you for all the info! Can anyone clear up the language in the help file? I am loosely quoting what I can remember, but under disadvantages for POW-3, it says that it is not recommened for "succesive bounces e.g. bouncing, freeze". What does that mean?

If you apply dither more than once to a given clip, the noise is naturally cumulative and after a few repeated bounces with added dither noise it will eventually become audible. Another reason not to use dither for internal bounces. Some dither algorithms are more tolerant than others because they shift the noise up to less-audible frequency ranges, so I'm guessing that's what they're alluding to.

Forget about POW-r3 versus POW-r2 versus MBIT+ versus UV22 versus triangular versus Spongebob Squarepants. If you can hear the difference between them in a pop/rock song you're frickin' Superman. Just pick one and go with it.


Danny: go ahead and use "dither down". I will not think less of you, my friend. However, given your reputation here (and on other forums) as an educator and patron saint of noobs, you might want to aim for semantic accuracy when possible, for their sake.

bit
Danny: go ahead and use "dither down". I will not think less of you, my friend. However, given your reputation here (and on other forums) as an educator and patron saint of noobs, you might want to aim for semantic accuracy when possible, for their sake.

Point taken....no more dither down if I've said it (which I don't think I do, but I'll be paying attention to my habits lol) as what you say makes total sense looking at it that way. It's nice to know you would have given me a free "non-annoyance" pass though, so I thank you for that. :)

-Danny