keep your files 24 bit!

ernest@303.nu · Post by **ernest@303.nu** » Wed Apr 03, 2002 4:14 pm

For some this tip will be obvious but others might benefit from it.

If you're used to 16-bit files you might at first not notice the difference.... until you do

Especially post-processing (dynamics, eq, enhancer, stereo adjustment etc) will sound much more natural an detailed if source and destination are kept in the 24-bit domain, even when the final master wil be resampled to 16 bit.

djody · Post by **djody** » Thu Apr 04, 2002 12:23 am

I'm with you.. !! It's a small differents at first but after a will you'll hear the benefit.

Laterz

Sunshine · Post by **Sunshine** » Thu Apr 04, 2002 1:20 am

Yes, of course!!!

Regards,
Sunshine

subhuman · Post by **subhuman** » Thu Apr 04, 2002 9:27 am

related: and avoid Sample Rate Conversions (ie: 44.1khz to 48khz, etc) if at all possible!

Neutron · Post by **Neutron** » Fri Apr 05, 2002 12:25 pm

especially now sound forge supports 24 bits!

BIT01 · Post by **BIT01** » Fri Apr 05, 2002 3:02 pm

Agree complitely!! it would also be nice to have 88.2 Khz as freq available for Pulsar, since this will introduce less dithering error and give some theorical bit of resolution for nyquist theorem...
Hope in new Fusion OS

:]

DJATWORK · Post by **DJATWORK** » Fri Jul 26, 2002 3:40 am

Im not sure about Nyquist would agree with you... 88.2 Khz gives you "a bit of resolution"?

May be you can explain that a little better...

dougal · Post by **dougal** » Wed Oct 09, 2002 1:43 am

88.2 divides by 2, a whole, round number, into 44.1 (the final sampling rate), thereby minimizing rounded off calculations when dithering, say from 48 kHz to 44.1 kHz. Mastering legend Bob Ludwig is emphatic on this point. So I record at 44.1. I wish we had 88.2 kHz, because my Apogee convertor does!

DJATWORK · Post by **DJATWORK** » Wed Oct 09, 2002 12:48 pm

I think that you are confused between "sample rate" and "bith depth"

The dither is aded to minimize the distortion listened in the less significative bits of a digital signal.
IT is added in the D/A conversion and in a bith depth conversion (I.E. from 24bits to 16 bits).

When sampling or resampling a digital signal a "antialias" filter is aplied to avoid alias frequencies.

fransje · Post by **fransje** » Tue Oct 22, 2002 4:15 am

I am just beginning to work with soundforge and recording. You guys mean save the files as 24bit after recoding ?

Also, I am recoding via creamware play/rec devices. Is that the best way too go ?

And I am using a asio2-24bit source device.

Tips/advice more then welcome

ernest@303.nu · Post by **ernest@303.nu** » Tue Oct 22, 2002 5:19 pm

If you want better sound quality and diskspace allows for it, then you should use at least 24-bit audio drivers and save wav-files at least at 24-bit!

The advantage of 24-bit over 16-bit is even more obvious when several steps of effects processing are applied to the sound files (which will quite often be the case)

Nestor · Post by **Nestor** » Wed Nov 06, 2002 1:26 pm

I have used 24 bits for my last CD and it’s sound better, there is no doubt about it. It’s much more demanding for the system, but it worth the effort.

As you said, the more you go processing, the more you use EQ and FXs in general, the more you can notice the difference by the end, when doing your final mix.

Those like me needing to bounce all the time due to lack of DSP power, will benefit most of all, continuous bouncing kills headroom and all the details of your music.

To go 24 bits it is almost as if you were able to record at once all the needed instruments of your project, it’s the closest thing you can do. Now that I have done it, I can tell for sure 24 bits is way better than 16, it’s not an unnoticeable little something, but an important implementation for your final work.

A close friend of mine, who is a composer as well, said to me something very intelligent we all should remember: “THE COMPOSITION PROCES FINISHES WHEN THE MIX IS FINISHED”. His philosophy is that mixing is still composing; I have to agree completely with this.

Nestor · Post by **Nestor** » Wed Nov 06, 2002 1:27 pm

I have used 24 bits for my last CD and it’s sound better, there is no doubt about it. It’s much more demanding for the system, but it worth the effort.

As you said, the more you go processing, the more you use EQ and FXs in general, the more you can notice the difference by the end, when doing your final mix.

Those like me needing to bounce all the time due to lack of DSP power, will benefit most of all, continuous bouncing kills headroom and all the details of your music.

To go 24 bits it is almost as if you were able to record at once all the needed instruments of your project, it’s the closest thing you can do. Now that I have done it, I can tell for sure 24 bits is way better than 16, it’s not an unnoticeable little something, but an important implementation for your final work.

A close friend of mine, who is a composer as well, said to me something very intelligent we all should remember: “THE COMPOSITION PROCES FINISHES WHEN THE MIX IS FINISHED”. His philosophy is that mixing is still composing; I have to agree completely with this.

Omb · Post by **Omb** » Mon Nov 11, 2002 6:21 pm

a litle thought;s .
should i make all the mix doen of the track 24bit or ?
what happend if i have worked all the way with 24bit get the track ready and in the end i need it agai with 16bit to burn it to cd and send it to label's 'isn't converting
it to 16bit can harm the clarity of the sound?

omb

Post by **hubird** » Mon Nov 11, 2002 7:37 pm

I think you certainly can finish the job in 16 bit, I will explain what my policy is.

The last step in every (?) mastering proces is the limiter, as it prevents to crossborder the zero db level.
(Actually I always limit at -0.3 db, seems to be safer for some cd player)

I use the socalled +L1 UltMax limiter from Waves, a professional soft limiter which covers the quantising (what we are talking about) and the dithering proces also.

What I wanne say here is that there are factory presets to choose, and one of them is called '16Bit FinalMaster highest res.'.
(Other presets are meant for pre-mastering and other funktioning).

So, after your music processing has done with 24 bit , including the other mastering tools *before* the limiter (like EQ, enhancing and compressing), the mixdown is supposed to get done in 16 Bit.

The profit of 24 bit processing, as stated elsewhere here, lays in the intensive and multiple types of processings you apply to all the files of the song (specially the mic recorded ones), like eq, effects, bouncing etc.
So burning down the mix to 16 bit doesn't mean at all cutting away all the quality profits of 24 bits

You could choose to master in 24Bit, but than another program, f.i. a sample editer like Peak, has do the job.
So I better trust the pro Waves limiter which suggest to do it in 16 bit.

There is one reason however you can decide to master in 24Bit.
That is when your file will be given to another person who will do some extra mastering edits like equalising or compr/limiting, say when you give your song to a record label for a release.

But than strictly we're not talking about final mastering by you, coz in this case you should not use the dithering funktion of your limiter (if a limiter at all).
Dithering must be the very last algorhytm (calculation) you apply to your mix, you shouldn't do it more than once!

This is what I have learned from written articles and specially from 'pro' friends in music bussiness, and I do it this way for years now.

Correct me if I'm wrong or forgetting something!

_________________
Let There Be Music!

<font size=-1>[ This Message was edited by: hubird on 2002-11-11 19:46 ]</font>

Omb · Post by **Omb** » Mon Nov 11, 2002 10:02 pm

thank's hubird.
just for understanding:
after i have done my final mix and mix it down to one file 24bit and than open sample editer like soundforge and there i convert it to 16bit with limiter like the L1 that you mentioned , am i rihgt?

another tought , is it o.k to record the final mix (to one 24bit file) in real time ,i meam,
open vst and sound forge ,open a wave destiation and conect it to the MixL and MixR
of the pulsar mixer preess record in sound forge and play in cubase .

cheers
Omb

marcuspocus · Post by **marcuspocus** » Tue Nov 12, 2002 9:45 am

That what i do usualy when i'm in a hurry...With wavelab though.

Post by **hubird** » Tue Nov 12, 2002 9:01 pm

On 2002-11-11 22:02, Omb wrote:
after i have done my final mix and mix it down to one file 24bit and than open sample editer like soundforge and there i convert it to 16bit with limiter like the L1 that you mentioned , am i rihgt?

yes, you can safely do it that way

Whatever sequence in programs you like to follow you could say there are 3 headrules:
1. stay in 24 bit as long as you can
2. the last effect in mastering is limiting
3. the last step should be dithering (as part of limiting).

I think it doesn't really matter if you go down from 24 to 16 bit during limiting or afterwards in a sample editor.
Choose the tool you think it's the best option, I would say a qualified limiter which is designed for the job is safest.

One thing at this point is, mostly there are small pieces of silence at start and end of the file, that need to be cut out of your mastered and dithered file.
You can safely do this 'afterwards', in your sample editor.

At this point the very last thing I often do is applying a smooth fade-out at the very end of the file, during the end of the natural fade you have, to mask some possible noise showing up when the sound dies away (due to the exiter for instance).
Strictly this is against my own rules, but I never hear anything wrong, and the fade itself is its own 'protection'

To your second question: as Marcus says: yes, no problem, but for me also just when in a hurry