to dither or not

[dmaxj]

I have been recording for a few months with PTLE. I am starting to dig it now... I always record my sessions at 24 bit/44.1 khz with no problems.

Question: does dithering improve the sound quality of PT recordings???

If so, is there supposed to be an audible improvement, or is dithering just 'one of those steps that you are supposed to take to make better recordings?'

Thanks

[evilchopper]

What a timely post, I've had more dither problems last month than in the previous year. Instead of giving a bunch of opinions not rooted in fact or practice like most of these posts seem to do, I'll propose an exercise which will enlighten you to the mystery of dither and when it's best used and not used.
Simply mix 2 seperate tracks of silence, one with dither and one without. Now add about 80 db of gain to each signal. One track is kind of quiet with random artifacts in the background, the other track is a track of hiss(dither) which hides the artifacts you hear in the other track.
The basic rules of physics concerning wave theory still apply: specifically the additive nature of signal amplitude. (i.e. dither can add up, so if you use it before the mastering stage, it will end up at a not so optimal level and sometimes exceed the noise floor.)
I'm growing increasingly weary of clients (of my mastering lab,) bringing in their work in a digital format which they recorded at home with no idea or training, and then they wonder why there is a light hiss after mastering when they recorded it on digital. I've recently adopted the policy to refuse the use of broadband noise reduction(the epitomy of unprofessionalism in professional audio) to counter poor recording techniques (the improper use of dither.)
I know of no succesful way to remove dither transparently, but I can always add dither, even to a final recording that has been mastered, it's never too late to add it.
Also, regular dither is ok, but when mastering, we prefer UV22 dither which is now available in the digital domain thru the latest version of Cubase 32 VST. Once you compare them critically, you will see why it was only previously available in the high end Apogee converters at a great expense.

Chop

[dmaxj]

evilchopper,

I follow what you are saying, however, will there be an audible improvement in the sound quality??

If not, then what's the significance of dithering???

Thanks

[j old school]

Somebody posted this web page a while back. It really helped me understand dither a lot better:

dither explained

This is sort of a costly solution, but if you have the resources, may be a better solution than using dither processes. You would need a high end D/A and a high end A/D. You can send the high resolution (24/96, or whatever) digital audio through your D/A, and send the output through a very clean analog path into your highend A/D which would create your 16/44.1 digital audio file, no dithering or bit truncating or sample rate conversion needed. If your converters are anything but the best, you are probably better off not using this technique and staying within the digital domain using a good dithering plugin for bit reduction.

As evilchopper explained, stay within the higher resolution for as long as you can. Only reduce resolution when you have to, as the very last step in the mastering process. If someone else is doing the mastering, give them the 24 bit file, and leave it to the pro to create the 16/44.1 file.

[AE]

Originally posted by j old school:
This is sort of a costly solution, but if you have the resources, may be a better solution than using dither processes. You would need a high end D/A and a high end A/D. You can send the high resolution (24/96, or whatever) digital audio through your D/A, and send the output through a very clean analog path into your highend A/D which would create your 16/44.1 digital audio file, no dithering or bit truncating or sample rate conversion needed.

<font size="2" face="Verdana, Arial">Except that the A/D convertor dithers as part of the quantization process. So, you actually can't avoid dithering. But, really, it's just not a big deal -- it does improve your mixes.

By the way, contrary to popular opinion, dither does
not hide or mask quantization distortion -- it actually prevents it from happening in the first place.

[Calvin]

Yes read j old school's link carefully. It will explain almost everything that you need to know. Dither in itself is not a bad thing at all. In fact it is recomended that you use it almost all the time. The only thing is knowing what dither to use. Do not use the Pow-R dither until you are to the mastering stage.