Bounce to disc vs record to track?

Please post your Mixing Question here for Pro Tools, Logic Audio and GigaStudio. Zumbido is our resident expert.

Post Reply
Carmichael
Posts: 36
Joined: Wed Mar 02, 2005 3:05 pm

Bounce to disc vs record to track?

Post by Carmichael »

There is support for recording the mix to a track instead of using the DAWs bounce feature (at least for protools). This can either be routed through an internal bus, or externally using a patch from the D/A output to the A/D channel inputs.

Here is a strong endorsement for internal routing - http://www.audiomidi.com/classroom/prot ... ner_22.cfm.

This raises some questions:
1. Why does it matter, especially if routing internally (routing externally, you could argue that the D/A and A/D hardware impart some curve smoothing or other processing). But internal bussing?
2. Where can you dither if you want to downsample the bit depth (if bussing externally, I presume you couldn't dither at all during this bounce, but would dither as a later step).
3. if just mastering a stereo file (one track of source material instead of 40 tracks), is it still important to record to track or is bounce to disk just as good?

OK - a separate issue:

Does it make sense to apply as much mastering as possible within the actual mix project (on the master fader) since some truncation will occur when you first bounce your mix. For example, a high pass filter, analog tape simulator, pultec eq, or even multi-band compressor (I know that using harbal before a multiband compressor helps the compressor work better, but) - is there some argument to applying these before the first bounce of the mix (ie is protools really 48 bit internal bus and how much truncation occurs when bouncing to 24 bits)? Harbal will always be critical, especially for analysis and developing an eq filter profile, but maybe it makes sense to do as much on the first pass as possible?

I am trying to decipher this article which I suspect contains much of the answer.: http://akmedia.digidesign.com/support/d ... _26688.pdf

Thanks!
Last edited by Carmichael on Wed Jun 13, 2007 12:00 am, edited 6 times in total.
Carmichael
Posts: 36
Joined: Wed Mar 02, 2005 3:05 pm

Post by Carmichael »

Ok. This thread addresses the first half of my post:
http://duc.digidesign.com/showflat.php? ... =7&fpart=1

It says:
1. Most people are routing internally rather than externally.
2. Most people think that Record to disc (RTD) is sometimes better than bounce to disc (BTD) but nobody knows quite why, except it may be likely to happen when there is a lot of automation (perhaps overloading the system) causing protools to shortcut some processes (ie autotuner) and to loosely follow automation envelopes instead of in detaill.
3. RTD in protools has the advantage that you can look around and edit the mix as it is bouncing, and you can fix a problem by punching in instead of restarting the entire bounce.
4. If your project is at 44khz, and want to dither, you could do that here. Dither is not a good option if you are >44khz b/c you should cut the frequency first - you would take your 24/88 product and import into another program or project file at 44 khz then dither and export at 16 bit (hopefully in a program that bounces faster than real time).
5. If just mastering a stereo file, BTD is probably ok, since you are not overloading the system and are not using a bunch of CPU critical tasks (such as autotuner).

As for the second half of my post:

My current thinking is that for protools HD, there is no advantage to applying mastering plug-in effects to the master fader of the mix project file, because the inserts bus for plug-ins is at 24 bit and the signal will be truncated (or dithered if using the optional "dithering mix engine") to 24bits prior to the plug-in input. So there is no advantage to applying plugins prior to saving the file to disc. For DAWS at 32 bit float, it also doesn't matter b/c you can save the file as 32 bit float.

Furthermore, this is a lot of discussion on DUC about the handicaps of protools HD and how it in fact appears inferior to native DAWS (such as protools LE or nuendo) b/c the plug in bus on the native DAWS is at 32 bit float (or even 64 bit), whereas TDM is 24 bit fixed. RTAS on HD helps bridge the gap some since RTAS plugs can pass the signal off to another RTAS plug at 32 bit float, but incompletely because the first RTAS in the chain receives only 24bits and the last one in the chain outputs only 24 bits. At least with PR 7, you can now use RTAS on Aux and Master faders, not just audio tracks. Anyway, mastering may be slightly superior in a native DAW than in HD. Of course, bouncing faster than real time is another huge advantage of a native DAW. But in summary, there is absolutely no loss in resolution when you bounce your mix out in order to go to harbal.

http://duc.digidesign.com/showflat.php? ... art=3&vc=1

Thanks
Last edited by Carmichael on Tue Jun 26, 2007 11:17 pm, edited 2 times in total.
zumbido
Posts: 123
Joined: Mon Oct 18, 2004 11:52 am
Location: Los Angeles
Contact:

Post by zumbido »

I've heard this debate about BTD v. RTD for years.

I for one do NOT hear a difference.

I've tried the 'phase cancellation' test countless times and ALWAYS get dead silence as the result.

Having said that... I use BOTH methods depending on the mix.

If I am satisfied that a one-time-through with no further tweaking works, then I use BTD. And, I'll leave for a cup o' coffee.

If I want to make manual changes during a mix, then I use RTD - I'll punch in/out several times during the mix. I very rarely use fader automation.

Furthermore, I do not hear differences AFTER consolidating either, as some claim that they do. Probably the same ones that make the claim of hearing a difference between BTD and RTD. Plus, I can never get these 'difference believers' to try the 'phase cancellation' test. I wonder why?

By the way, I always consolidate the 'MIX' track if I've done any punch ins and outs.

I use PT 5, Mix 24 on a Mac running OS 9.
Post Reply