Dithering and 48k

[nz]

Hello,

A quick "newbie" dithering question for everyone...

If I do a session at 24 bit/44.1 and eventually have to dither down to 16 bit for CD anyway, is there an advantage to recording 24 bit/48k from the onset? With there be some audible difference with beginning with a better sample rate?

Thanks in advance.

[Frank S]

Neil (or is it Neal?)

There may be a slight advantage in sound but there is a major hitch to working at 24/48. You have to get the 24/48 back to 16/44.1 if you plan on putting it on a CD. Most people argue against sample rate conversion from 48 to 44.1 if you do it via software - digi's included. If you happen to have some amazing AD and DA converters, you could take an analog out of your 24/48 and reconvert it by recording it via the amazing AD converters as a 44.1 when you are done mastering it. If you don't have very good converters - STAY AWAY from any software SRC - its not that good.

Now dithering really only applies to reducing the word length, and not sample rates per se - so it only applies when going from 24 to 16 bits - regardless of the sample rate.

If I am working on something that I will burn to CD - I keep it at 24/44.1 and dither to 16 bit when getting ready to make the CD during the 'mastering process' - its easy and it sounds great. If its going to a mastering house (or somewhere with much better converters than I have ) - I might do a 24/48 - dither to 16/48 - and digitally record a 'master' to DAT. Then we can do the DA and AD sample rate conversion thing.

Does this help?

[This message has been edited by Frank S (edited May 03, 2001).]

[nz]

Frank,

That's perfect! Exactly what I was asking. Thanks so much for your very informative answer.

BTW, it is Neil.

nz

[ThomCat]

If you have the opportunity to record at 88.2 that might be advantageous, and converting to 44.1 is as simple as dividing by two, so even software conversion is probably free of artifacts, but 48 to 44.1 is much tougher math to do without adding rounding errors, and I agree with Frank that it's better to stay away from it as any small advantage of 48 over 44.1 would be compromised by the conversion.

[Jim S]

What about getting 24 bit/96 Khz to 16 bit/44.1Khz CD?

Also, I think this is true but: all tracks in an audio project must be recorded at the same bit & sampling rate, correct? You can't record some at 48 and others at 96 into PT LE for the same song, right or wrong?

[guitates]

The Nyquest Theory suggests that one difference in analog and digital is the frequency response of the harmonic overtones available.

With analog tape, we can realize up to 200k in frequencies.

With digital, we are limited to 1/2 of the sample rate.

So, if you record at 96k you can realize and "hear" up to 48k of frequencies, which would only be 22k when recording at 44.1. 48k gives you 24k---big deal!!

In the Digital Domain, this is part of the necessity of using plug-ins properly to simulate the analog "warmth",
and I for 1, say that it is entirely concievable to achieve analog "warmth" and even surpass the quality of analog with the availability of quality plug-ins when recording at
44.1................Frank's comments are right on .....

[Jim S]

What about 88.2?

[guitates]

88.2 would be Ideal!!

Who offers this???

I want one!!!!!!!!!!!!

[Jim S]

RME ADI-8 DS has 88.2. But uon further thought: what's the use of it ProTools is proprietary 48 khz and not 88.2 or 96?

[guitates]

Just kiddin bout the 88.2.

If you can't do it in 44.1---forget about it!!

44.1 works GREAT!!!!!!!