To bounce or not

[Dog_Chao_Chao]

Hi. I´ve been hearing and reading some different opinions about to bounce or not to bounce your sessions. I still can´t find a major agreement about this. Some people i work with say bouncing is just fine while others say that there are better ways to do it. For example, a guy from digi here where i live says he prefers to print (rec) the mix to a stereo track. What are the risks of doing so?
What is your opinion about this and what different ways are out there to get your mix ready for a cd burning? What is the best way to do it? Any tricks?
Thanks

[max cooper]

I guess this is the standard reply, but what differences can you hear?

And what's his procedure for getting the stereo track to a disk?

I've heard people say that there are better ways than bouncing, but I've never gotten around to trying other ways. Just curious.

I'd guess that the first thing to do is to bounce the mix and then import it back into the session, then compare it to the stereo track (making sure the levels match, of course.) This way the bounce and the stereo mixdown will be played back through the same chain (right?)

[Dog_Chao_Chao]

Thanks for your reply max. About the procedure the guy explained me for getting the stereo track to a disk its a very simple one: he just opens a stereo track in the session ( so he just uses 30 tracks max. with pltle) and records everything there. Then makes an export and sticks that file into nero or some other burning soft. About the differences, i didn´t make a test like you were refering. I´ll try it out. But, maybe it´s just my impression, i can notice some deterioration after the bounce, specially in the very high and very low frequencies. this can also be a impression caused by the fact that i get to hear the bounce in a lot of other different places ( car, house stereo, computer, etc.)...Still i´m curious to know about what people think of this here..
Thanks

[Dragonfly]

I was aksing myself that question myself a few months back. I record all to a stereo track typically, and on occasion I will make mix splits in the same way. If I do have to, or just plain decide to, bounce to disk, then I make sure I do it in split mono. Stereo Interleaved causes an unnecessary convsersion. Whenever I send files off to be mastered, they go as split mono, 88.2 / 24 bit. I try to avoid conversions wherever possible until the very end, and I do hear a difference in quality.

Dante

[Dog_Chao_Chao]

Thank you for your help here. Dragonfly, i really didn´t know about the stereo Interleaved / split mono diference. I will try that out too. By the way, do you notice a big diference betewen the 88.2 and the 44.1 sessions? I am having doubts about wich sample rate do record ( 96, 88.2 or 44.1)
Thanks

[Dragonfly]

Originally Posted by Dog_Chao_Chao

do you notice a big diference betewen the 88.2 and the 44.1 sessions?

I do when everything in the chain is the best it can be. Unless I am using higher end mic pres, nicer microphones, I'm in a decent room, and the artist or sound source is good, than 88.2 isn't going to make a world of difference. But if everything is running cleanly than yes, the difference is substantial.

In regards to 48 and 96k, that is once again another conversion which needs to be performed by the computer. The math to convert a 96k file to 44.1 is complicated, even for a computer. A simple chop in half from 88.2 to 44.1 sounds much better than the latter, and I've tried this as well with my own tests. And if I can't do 88.2 due to system compatibility, I work in 44.1

Dante

[theblue1]

For example, a guy from digi here where i live says he prefers to print (rec) the mix to a stereo track. What are the risks of doing so?

I can understand the rationale for mixing out of the box using a multichannel output to send subgroups or even single tracks into an analog board. The notion there is you'll be using as much dynamic bandwidth as possible for the small numbers of key instruments on separate channels as you send out your submixes and then use the mixer to sum the tracks.

[BTW, for you digital generation types: 'summing bus' is an old, pre-digital term. It's just that before 'summing bus hysteria' broke out a couple years ago on the Digidesign User Conference a couple years ago, most musicians probably never knew that that was where all the channels in their (analog) mixers got added together.]


But... if the the choice is between bouncing a project ITB down to a stereo mix or outputting the same mix out a stereo pair of analog outs and back in to two new tracks as I think the engineer you mention above apparently is... if that was true, what would it suggest? And what would it buy you?

In both cases, if I understand correctly, whatever negative effects one posits from the process of software summing is there. And if you then run the summed stereo signal out through A/D and then back in through A/D, you're just adding another layer of aliasing distortion in order to, what? Avoid saving the already summed stereo to disk?

What could that possibly buy?

Really.

[superburtm]

On the split mono files subject funnily enough my mastering guy requests stereo interleaved files. Hmmmn..

[max cooper]

Originally Posted by theblue1

In both cases, if I understand correctly, whatever negative effects one posits from the process of software summing is there. And if you then run the summed stereo signal out through A/D and then back in through A/D, you're just adding another layer of aliasing distortion in order to, what? Avoid saving the already summed stereo to disk?

This is just conjecture, since I haven't had the opportunity to A/B analog vs digital summing as much as I would like to, but what if it turns out that digital summing is actually too 'good' (or maybe the word is 'accurate'?) which is why we don't 'like' it as much? Maybe we're 'losing' something in the summing amp that makes the result more palatable? Again, I don't know, but is it possible?

[Bob Olhsson]

I've had no problems at all with split mono tracks.

Interleaving in Samplitude actually sounded a hair better than interleaving in Pro Tools the one time I tested it to make sure Samplitude wasn't broken. (I have no idea why this was and haven't considered it worth pursuing since all I wanted to know was that Samplitude works ok.)

digital, ughhhhh!

[Dog_Chao_Chao]

Thank you all for your answers. As refered in the begining of this thread it doesn´t seam to be a general agreement about this issue. that is just what i have been experiencing with the people i work or have contact. I guess it isn´t so relevant if we bounce stereo interleaved or split mono, if we print it in a stereo track or bounce to disk or whatever. Anyway i´m thinking of getting a 2 track studer or ampex to do this, or even a 8 track to play with!
Thanks again.

[theblue1]

Originally Posted by max cooper

This is just conjecture, since I haven't had the opportunity to A/B analog vs digital summing as much as I would like to, but what if it turns out that digital summing is actually too 'good' (or maybe the word is 'accurate'?) which is why we don't 'like' it as much? Maybe we're 'losing' something in the summing amp that makes the result more palatable? Again, I don't know, but is it possible?

Absolutely.

I was just reading a sales pitch on Mercenary for the FATSO. Whoever wrote the copy (Fletcher, himself? Empircal?) bent over backwards to make it clear that he felt that what people like about analog storage media (specifically tape) and analog gear in general was often its lack of linearity and "flattering" disortion.

[mixerguy]

Originally Posted by Dog_Chao_Chao

Hi. I´ve been hearing and reading some different opinions about to bounce or not to bounce your sessions. I still can´t find a major agreement about this. Some people i work with say bouncing is just fine while others say that there are better ways to do it. For example, a guy from digi here where i live says he prefers to print (rec) the mix to a stereo track. What are the risks of doing so?
What is your opinion about this and what different ways are out there to get your mix ready for a cd burning? What is the best way to do it? Any tricks?
Thanks

I've done tests.... in OS10 PT 6.xx bounce to disc DID sound different than bussing the mix back and re-recording it into the session..... very slight difference, but it did sound different.

Unless it was the crack.

Do your own tests - maybe I am nuts.......

[Lagerfeldt]

1) There's no audible difference in split files and interleaved, only the file size. That's easily proved by doing a null test, flipping the phase on one stereo set. It's a misunderstanding that going from interleaved to split is a conversion that changes the sound, it is not. Unlike sample rate conversion, that actually changes the sound.

2) The old ProTools engine had some serious sonic problems when bouncing, so that's why some preferred to go out and in (including myself). Also the old Logic engine had a similar problem. But not today, at least in Logic Pro which I use now. Whether there's a difference today in ProTools, I guess there certainly could be one, but I haven't tested that since I only use ProTools a few times a month now.

3) Also there's no difference in realtime bouncing and offline bouncing in terms of sound quality. Also this can be proved doing a null test.

If you do a null test beware that you cannot use audio instruments with LFO's (unless they are perfectly re-triggered) and you also cannot use any other kind of arbitrary device or process as this would cause a discrepancy.

There's really no need for guess work as to whether something sounds "a little different", etc. It can easily be proved or disproved using the null test.

[mitchell]

Originally Posted by Lagerfeldt

1) There's no audible difference in split files and interleaved, only the file size. That's easily proved by doing a null test, flipping the phase on one stereo set. It's a misunderstanding that going from interleaved to split is a conversion that changes the sound, it is not. Unlike sample rate conversion, that actually changes the sound.

2) The old ProTools engine had some serious sonic problems when bouncing, so that's why some preferred to go out and in (including myself). Also the old Logic engine had a similar problem. But not today, at least in Logic Pro which I use now. Whether there's a difference today in ProTools, I guess there certainly could be one, but I haven't tested that since I only use ProTools a few times a month now.

3) Also there's no difference in realtime bouncing and offline bouncing in terms of sound quality. Also this can be proved doing a null test.

If you do a null test beware that you cannot use audio instruments with LFO's (unless they are perfectly re-triggered) and you also cannot use any other kind of arbitrary device or process as this would cause a discrepancy.

There's really no need for guess work as to whether something sounds "a little different", etc. It can easily be proved or disproved using the null test.

Lager- thanks for the useful info... I've read lots of your posts and you appear to be the only guy in the free world who knows more about Logic than I do!!!

mitchell

[Lagerfeldt]

Thanks!