24 bit mastering worth for 16 bit session in PT?

[Vynka]

Hello, I've looked for a precise answer in the archive topics , but couldn't get it so here I go with my question. Thanks for reading !

I have songs recorded in 16 bit sessions on a TDM Mix (+ an old PCI Farm) to be mixed, mastered and burnt on my own CDR.
I chose 16 bit format at the beginning because I thought maybe it wasn't worth the extra disk space to record in 24 bit because I was using two old 882/16 converters.
Now, even if I'm wondering if that was the best choice I have to deal with it the best way I can.
I plan to first mix them by using the bounce method and then to import each bounced mix in another session for mastering purposes (using L1, level balance the songs,etc..)

I'm wondering what choice is the best:

1) bounce in 16 bit and open the new mastering session in 16 bit too, or

2) save the 16 bit sessions in 24 bit format make all the mix and plug in treatment in these 24 bit sessions, mix (bounce)them in 24 bit stereo files, create my mastering session in 24 bit and finally bounce every song in 16 bit ready to be burnt on CDR, or

3) keep my sessions in 16 bit, do the mix and plug in treatments in it, bounce them in 24 bit and do the mastering stuff in 24 bit just before ending in bouncing the whole thing in 16 bit CD format.

Thanks for your advices and excuse my approximate english (I'm french).

Vynka. (a very happy reader of this Bulletin board) http://www.digidesign.com/ubb/images/icons/smile.gif

[This message has been edited by Vynka (edited October 29, 2000).]

[Arno Peeters]

I'm no wizzard at this, but it seems to me that if you want to apply effects, dynamics-processing etc., convert tehm to 24 bits: you would have more headroom and clarity (stereo-balance, high transients).
Be sure to dither back to 16 bits, preferably with L1 last insert on the master.



------------------
Arno Peeters
www.tapetv.nl (http://www.tapetv.nl)

[5down1up]

as far as i know and hear ...
theres no need bringing 16 bit files onto 24 bit.the only reason i know is if u wanna use sample cds ( audio ) in 24 bit sessions.
converting increases the dynamic levels downwards.so if u got an 24 bit file the is nothing under the 16 bit dynamic level.and it will not come back if you convert.
set the internal solution to the highest ( in plugins like the L1 ) thats all ability u got left.good sounding , bye

[Vynka]

Thank you very much for all your answers

<BLOCKQUOTE><font size="1" face="Arial,Helvetica,sans-serif">quote:<HR>Originally posted by 5down1up:

converting increases the dynamic levels downwards.so if u got an 24 bit file the is nothing under the 16 bit dynamic level.<HR></BLOCKQUOTE>

Do you mean that when I would finally convert to the CD 16 bit I would loose the 24 bit treatment details and find back the tiny details I primarly had on the very first 16 bit files or loose both ?
Don't you think that the TDM plug ins processing in 24 bit would benefit out of it like Arno seems to suggest ? (all the tracks are acoustic recordings made from the 882/16)

Finally, if it isn't worth doing so, why do mastering studios convert to 24 bit everything they are about to master (even 16 bit digital mix) before processing the mastering when they have to end in the 16 bit format of the CD anyway ?
Is it because the processing is better on their systems http://www.digidesign.com/ubb/images/icons/frown.gif or just because those systems have only high resolutions ?

Thanks again for the answers and your encouragements ! http://www.digidesign.com/ubb/images/icons/smile.gif

[Oli P]

Keep the audio files as 16 bit, but use the 24 bit optimized mixer, and make the bounces in 24 bit.

It sounds a lot better to sum many 16 bit tracks to 24 bit stereo, and then dither it to 16 bit at the final mastering.

[5down1up]

never heard about that a mastering studio converts 16 bit files to 24 bit files ???
why should they ??? the only reason to convert is if u dont stay digital.means you r using analog outboard to master , too.
like i said , you´ll get nothing back if u convert 16 b to 24 b.
internal resolution and bit file size got nothing to do with each other.
any opinions ... i am ready for my lesson ...

hope that helps you out

[Studio Dweller]

By all means, when you do your mix, bounce to disk at 24 bits! Even though your individual tracks are 16 bits, even the simplest DSP functions (ie: changing gain, panning) result in more bits. Always leave your dither preferences set to 24 bits, even if you're using the 16 bit mixer - it has more than 16 bits of internal resolution. If you can use the 24 bit mixer for everything, it has more headroom than the 16 bit mixer (16 bit mixer=18db headroom, 24 bit mixer=30db headroom).

If you were to bounce down to 16 bits in your mixing session, then you would put a dithering plug on the master fader, but I prefer to bounce at 24 and then start a new session to do the bit reduction.

So, the bottom line is, regardless of which mixer you use, bounce to disk at 24 bits with your dither prefs set to 24 bits. If you can afford the DSP, use the 24 bit mixer for your mix and definitely use it for your mastering session.

Cheers,
-Larry

------------------
Larry Bentley
Cellar Dweller Productions www.cellardweller.com (http://www.cellardweller.com)

[This message has been edited by Studio Dweller (edited November 03, 2000).]

[cknight]

<BLOCKQUOTE><font size="1" face="Arial,Helvetica,sans-serif">quote:<HR>Originally posted by 5down1up:
...you´ll get nothing back if u convert 16 b to 24 b. internal resolution and bit file size got nothing to do with each other. any opinions ... i am ready for my lesson ...
<HR></BLOCKQUOTE>

You are correct that converting 16-bit file to 24-bits doesn't add to the sound quality of the file. But you can't have 16-bit sound files in a 24-bit session. (At least not in PT.) You will get better sounding results if you mix in a 24-bit session. So the advantage in converting 16 to 24-bit is that it allows you to mix in a 24-bit session. Then at the very end you dither your mix down to 16-bit.

-and then you upoad your mix to an mp3 site and everyone hears your fabulous 24-bit dithered to 16-bit compressed to mp3 mix in beautiful mp3 glory. http://www.digidesign.com/ubb/images/icons/wink.gif

Also, I would think that a mastering house would only convert a 16-bit file to 24-bit file if they are using an all digital system that doesn't allow you to use 16-bit files in a 24-bit session. But there would be no reason to convert 16-bit to 24-bit if the are mastering analog. All you would need is the highest quality D/A converters, then master accordingly and use the highest quality D/A converters when making your digital master. Am I right? Or is it time for my lesson? http://www.digidesign.com/ubb/images/icons/smile.gif

[This message has been edited by cknight (edited November 03, 2000).]

[tld]

I'm no expert in this area, but this is what makes sense to me: Although bringing 16 bit tracks into a 24 bit session only adds empty bits, it still makes sense as long as you are going to do ANY kind of mixing...even level changes, let alone something like L1. The mixing calculations are simply going to be more accurate in a 24 bit session. If this weren't true, anyone mixing for CD would do just as well recording at 16 bit...which certainly isn't the case.

Tom

[Vynka]

Hi everybody, thanks for your answers ! http://www.digidesign.com/ubb/images/icons/smile.gif

Although I'm a little bit confused sometimes reading your suggestions I feel like you all agree at least on bouncing to 24bit and it gives a partial conclusion to my question.
What I 'm still wondering a little bit now is whether I should use a 24 or a 16 bit session in order to bounce to 24bit.
Is there a difference in the mixer and plug ins processing between a 24 bit and a 16 bit session regardless of the original files resolution, knowing that I would use the 24 bit optimized mixer in both cases (accordingly to the PT manual)

Thanks for all of you, it's good to feel not lonesome on this side of the atlantic http://www.digidesign.com/ubb/images/icons/smile.gif

[nipple]

For the best possible sound quality you can get by starting with 16 bit files the thing to do is convert to all the audio files and the session file to 24bit (save copy in). The sound benefits you will get from this are not subtle because although your 16 bit files won't sound any better, the additional headroom by mixing at 24 bit will make a huge difference.

[This message has been edited by nipple (edited November 04, 2000).]

[5down1up]

if they master analog , they convert to 24 bit.why ??? cause analog mastering will cause the levels and the dynamics.so the analog process will be rerecorded to a 24 bit file.thats why.
changing the internal mixer solution effects all processes being made with the mixer ( volume , panning , automation ).maybe sometimes that needfull for a mastering-session.but to master a "simple song" its not that important.the only plugin which can change its internal solution ( so far as i know )is the waves L1.setting that plugin to 20 bit dont mean , that youll get a 20 bit file out of an 16 bit file being processed.its just helpfull if u wanna have high levels on your music without getting digital distortion ( especially on cymbals ) .
just let your ears decide which way to go . not just computers and levelmeters and analyzers etc.its just helpfull to visualize but it cant display "GOOD SOUND , GOOD SONG".
Jimi Hendrix had 4 tracks and its still rocking :-) without knowing " was that 21 or maybe 22 bits ??? "
this looks all to a marketing thing to me.
or way got great fx stuff like eventide , ams , lexicon etc. some of the "world best sounding gear" and just got 16 bit or less converters ??? why does a real marshall sound better than ever amp simulation that has been created?
why do high end processors like studer just got 21 bit converters?

keep the discussion rolling , all the best to you all

[Vynka]

Thanks for your very interesting views on the question ! http://www.digidesign.com/ubb/images/icons/smile.gif
You are right about Jimmy Hendrix; I used to say exactly the same thing (except that I was using the Sgt Peppers Beatles album example http://www.digidesign.com/ubb/images/icons/wink.gif) to people asking for more tracks than my analog studio could provide at the time they were asking.
But as usual nowadays my work has to be done quickly and I won't have time to make comparisons tests very much http://www.digidesign.com/ubb/images/icons/frown.gif. I also have the feeling that the differences are very clear only after a certain time of multiple auditionings and I would like to start with the best chances to have very few regrets later.

Speaking about very good gear there is an option (for one or two special tracks) I would like to have your opinions about:
I was wondering what would happen to the bounced 24 bit files if during the mastering session I digitally send-return them in my Lexicon L300 (which is 16 bit) in order to use the L300 converters (that I presume to be better than my digi 882/16) all this in order to use inserted analog tube devices.(It's possible to use the analog Lexicon IN/OUTs as analog insertion points in the digital chain), I tried that another time with 16 bit files and I felt it was worth the 2 extra convertions.

To resume the question:
What happen to a 24 bit file when sent to a 16 bit D/A converter; is it troncated?
Maybe I should have asked my question in a new topic, too late, sorry http://www.digidesign.com/ubb/images/icons/blush.gif
nice to read you all ! http://www.digidesign.com/ubb/images/icons/smile.gif

------------------
Denis :)

[Vynka]

Hi me !
In case someone would read the last question, I've partially found an answer in the topic 'Connecting effects digitally only' in this same section.
And on Lexicon web site :
Question: What happens if you send a 24 bit AES/EBU word into the M300?

Answer: The M300 will process AES/EBU signals of up to 20 bits in length. A 24 bit word sent to the M300 will be truncated down to 20 bits. For optimum audio performance it is suggested that the 24 bit signal be dithered down to 20 bits prior to sending it the M300.

That's all folks !
Bye http://www.digidesign.com/ubb/images/icons/smile.gif

------------------
Denis http://www.digidesign.com/ubb/images/icons/smile.gif

[This message has been edited by Vynka (edited November 07, 2000).]

[This message has been edited by Vynka (edited November 07, 2000).]

[This message has been edited by Vynka (edited November 07, 2000).]

[5down1up]

depends if the 300s cutting off??? basically if u connect digital there should be no loss.only time you lose signal is for example :" you run a 24 bit file digital into a dat and the dat converts d/a.then the dat d/as depend on the bit length you´ll get."but if u stay digital means 24 bits out into the dat and back in a 24 bit d/a converter you´ll get 24 bits.only thing 16 bit dats "cant record 24 bit length" in fact of the tape has to be speeded up.but there is no problem with routing em digital in and out.means you´ll send a 24 bit file to your 300 ( theory ) and your 300 just got 16 or 20 bit converters you will get 20 bits at the analog output. but if you output the signal back to PT ( digital ) .there should be no losing bits at all ?!?!
means - the 300 dont even has to convert . it just processes the incoming signal but theres no conversion from the original file.
imagine you got tons of digital gear.would be whack if you got a chain and all gear´s got different converters.convertes just mean "conversion" . either from digital to analog or from analog to digital.digital to digital means no conversion at all.
any opinions ??? let em hear ... good luck

[Extreme Mixing]

I'm no scientist, but here goes...

I was just reading in another topic, that making a 6 db gain change creates a decimal point followed by about 30 or so digits. That means that you are way better off to mix in the 24 bit mixer. Converting the individual 16 bit files to 24 bit files just adds 8 zeros somewhere in the bit depth (not necessarily at the bottom?), and takes 1/3 more disk space to store the file.

The 24 bit mixer calculates your gain changes, pans, and plug-in resolutions in much finer detail than the 16 bit one. That should make a big difference in the sound quality of your mix. When you convert your mixes to 16 bits, make sure you dither so that you get 18 or 20 bit resolution. If you don't, those 8 least significant bits just get cut at the 16th bit, and that's not good.

My understanding of bit depth is that the dynamic range stays the exactly the same. Clipping is clipping and silence is silence. You don't get more headroom. You get smaller gaps between the bits and the ability to better describe lower level events. Higher resolution. The dynamic range is increased downward only. You get 24 digits to describe the same event, whether it is silence or clipping. That's why throwing those least significant bits out is bad news. A 16 bit recording would describe the event to the best of its ability. Hard truncating of a 24 bit event to 16 bits effectively looses one third of the data--information that would have been described by the 16 bit converter. That can't be good.

I too am willing to be educated if my thinking is flawed.

Steve

[Digi9000000]

You're right Extreme.

24 bits gives more room for error, literally. The reason that things "sound worse" at lower bit depths is because of rounding errors in the math. The more room you have for calculation, the better things sound, theoretically.

So even though you started with 16-bit files, you will get some serious benefit from copying the session to 24-bit. Your Plug-Ins will sound better, you will have more headroom, and your mix will shine just a little more.

Then in the end, bounce to a 24-bit file, no dither necessary. Then make a new "master" session if you are mastering in Pro Tools, and dither to 16-bit...

Always dither if you're going to less than 24 bits, because the Pro Tools master mixer is ALWAYS 24-bit, regardless of your session bit depth and mixer plug-in.

[Vynka]

Thanks to all of you http://www.digidesign.com/ubb/images/icons/smile.gif

Accordingly to the majority of these kind answers ,I will open the 16 bit files in 24bit new sessions, bounce them to 24bit, do the mastering in a 24 bit session and in the end dither and bounce to the 16bit files ready to be burnt.

I still wonder a little bit if it's really worth to start by opening the 16bit files in 24bit new sessions. The PT manual seems to outline the importance to use the 24bit optimized mixer(rather than the 16bit optimized mixer) whenever 24 or 16bit files are used so I have the feeling that the really difference of quality between 16bit or 24bit processing lies there and not in the session itself. So as soon as I use the 24 bit optimized mixer it sounds like there will already be the maximum processing quality with it whatever session (24 or 16) I open my files in.
Would someone from Digidesign like to bring light on this please?
Once again sorry for my aproximate english !
Thanks again !

------------------
Denis :)

[Oli P]

Right on! Use the 24 bit optimized mixer. Then all the processing is in 24 bit, but your audio files still in 16 bit.

Then make the bounce in 24 bit, and dither at the final stage.