Mixing In PT: Bounce vs. Record to audio track

[recall]

Hi Guys,

First of all I'd like to introduce myself as this is my first post.
I am from West Wales UK. I mainly record and produce welsh artists and specifically welsh language music, as well as producing my own music and doing music for welsh TV.

For my first post I have a question mainly directed at Charles (thanks for Milar!) but I'd like to hear anyone else's opinions.

There have been many posts on other forums such as the DUC and Gearslutz regarding the "printing" of the final mix in PT.

Some claim that bouncing to disk reduces the resolution of the output file and causes errors in automation playback.

I can somehow see the logic in this especially on more comples maxed out sessions.

The work around seems to be to bus all tracks to free aux input and set aside a stereo track to record from the aux.

I have done this and the results are good. I try and do this from now on, but do not always have the free bus and audio track since I am working in PTLE (32 voices/auxes)

Basically I was wondering if you had any thoughts on the possible downsides of bouncing or if anyone has done more thorough A/B tests between both methods.

Thanks ...and I salute one and all!:very happy:

[EricTheKid]

I have found a great article for this subject that really puts things in perspective (and dispells some myths). Give me a bit, and I will post again soon with the link! :Thumbsup:

Talk to ya soon,

ETK

[EricTheKid]

Basically I was wondering if you had any thoughts on the possible downsides of bouncing or if anyone has done more thorough A/B tests between both methods.

Alright...after surfing the web, I'm back (and boy am I exhausted...)!! Yeah yeah...pretty bad...I know. :Roll eyes: Anyway, here's the link: Bounce-to-Disk vs. Recording a Mix to an Audio Track (http://www.blog.protoolstraining.com/?p=1)

"Over the past decade, the debate has raged on which sounds better, Bounce-to-Disk, Bounce-to-Disk using busses as bounce source, or Recording the Mix to an Audio Track.

I’m going to add my 2 cents to this debate and hopefully find some resolution by testing each method. I’ll share with you what I’ve found to be the pros and cons of each approach.

So let’s get started!

What we are going to do is record a mix using all three methods listed above, import them back into the original session, invert the phase of the mixes, and then compare..."

Article is by "James" of ProMedia Training (http://www.protoolstraining.com)


Hope this re-confirms things for you, Recall.

Peace,

ETK

[PSN Big Al]

Excellent research there Eric. Thank you! (And a bump to your
rep for good measure. :) )

[EricTheKid]

Excellent research there Eric. Thank you! (And a bump to your
rep for good measure. :) )

Muchas gracias, Senior Al. :Thumbsup: Enjoy...

[recall]

Nice one Eric,

My ears were not deceiving me then:D

Anyway I have found recording to a spare track to be useful for another not so apparent reason.

The bounce handler in PTLE often crashes for no apparent reason often very close to the end. Trying all the digi fixes and playback settings helps somewhat but there is a definite bug there.

Recording eliminates this, which saves time.

All in all great work:Wink:

[seagate]

Wonder if this is the same in Logic...

:Confused:

[pounce]

i've been recording to an additonal track for some time for a variety of reasons. in DP, bounce can be faster than real time which seems cool, but i integrate a mixer and a lot of real outboard gear in the mix even though i'm using a DAW. so i have to ultimately just record a stereo track being output from the mixer with everything nicely summed. been working great.

[Baddo]

Why would I bounce to disk if I don't trust the computer? I mean, don't you check your bounce's after the fact anyway?

[CaptainHook]

I'm slightly confused by this..

You would think (and by that, i mean I think) that when
you choose the bounce to disk option, more resources are being
used to carry out the automation, summing of the mix, etc
INSTEAD of keeping the graphics updated etc..

Although the link is interesting, he doesn't explain how he
concluded that recording into an audio track was more
accurate.. he just states "it is". Did he just hear that the
timing slipped on the bounce to disk?
If so, fair enough.. But he doesn't say this is how he knows.
Is it because phase-reversed there was more audio to be
heard in than the other way?

Not trying to argue for the sake of it..

But it seems weird that during normal playback, when
more resources are being used, this can generate a better
stereo mix of your session.
Maybe the graphics etc are unrelated, but i assume there is
a reason that during bounce to disk, they disable updating
graphics.

Maybe i'm looking at it all wrong.
Does anyone else have more conclusive evidence?
I have no issue if recording to a separate track IS the better
option, it's just not adding up in my head.

I guess i'm gonna have to try it and "hear" for myself.

[Skwaidu]

I've been bouncing to disk close to 10 years with no problems I can remember of... Occasionally I might record to a track if I feel like it though. In my setup I can do both but BTD is usually more convenient.

[recall]

I have no proof apart from my experience. I am mixing an album at the moment. On the pop/rock songs with lots of automation and full track count btd sounds pinched in the hi-mids. I can't describe it as anything other than that. I can't say I have noticed automation errors, but what I can say is when i record to a new track i get back what I hear from the speakers, which really to me is what mixing should be about.

Granted I have used btd with succes on smaller ensemble recordings with no problems.

As such I would say that the key determinent in succes with BTD is the complexity of your recordings.

[archtop]

The link of the


"Bounce to disc vs recording to a new track"


2 tests ?


and it's conclusive ?




all he's proved to me is


he's a dipstick.

[PRobb]

A sidenote- if you are mixing by recording to new channels ITB, if you want to stop and punch, doing it in destructive mode will create a single, continuous audio file.

[Charles Dye]

Some claim that bouncing to disk reduces the resolution of the output file and causes errors in automation playback.

BTD – Reduced Resolution?

This is completely false in my experience. I've tested this many times + found it 2 never be true. I've found 0 difference in the audio resolution between the mix playingback live + a bounced version of that mix.

Furthermore, I've read way 2 many threads on this topic 2 count + I've never once read proof that disuaded me from thinking any different. I'm open to hear it, but I just haven't heard it yet. Please post links if you know of a thread that convinced you this is true. I'd love 2 read it.

Basically, I think this is an internet legend. Many people repeat this as true, but no one ever seems to show that it is ACTUALLY true.

BTD – Automation Errors?

OTOH, depending on the power of your CPU + the density of your automation, this sometimes seems to be true in my experience. During the bounce to disk process, when there is a lot of automation on multiple tracks @ the same time (which is played back by the cpu, not the PT hardware), it may sometimes not playback 100% accurately.

This doesn't always happen, but I have experienced it on slower/older CPUs. If u've found this on ur system u may want to try printing to a stereo track thru an aux buss (which apparently u r).

The work around seems to be to bus all tracks to free aux input and set aside a stereo track to record from the aux.

I have done this and the results are good. I try and do this from now on, but do not always have the free bus and audio track since I am working in PTLE (32 voices/auxes)

Printing 2 Stereo Track

I bounced to disk for a long time with no problems. I switched to printing thru an aux buss to a stereo track, not because of any problems with BTD, but because it allows me to quickly go in + print a tweak to the bridge (for example) w/o rebouncing/printing the entire mix. Big time saver.

I hope this helps.

(+ you're welcome for MiLaR, recall!)

[Charles Dye]

Great link there ETK. :Thumbsup: James is a real smart guy + I know him well.

As I understand the PT mixer software, everything he said makes sense with the technical side of the story. IOW, there should be no difference between those three. It's just that sometimes the computer doesn't playback the automation perfectly. Which also happens sometimes with larger consoles.

[recall]

Great link there ETK. :Thumbsup: James is a real smart guy + I know him well.

As I understand the PT mixer software, everything he said makes sense with the technical side of the story. IOW, there should be no difference between those three. It's just that sometimes the computer doesn't playback the automation perfectly. Which is also happens sometimes with larger consoles.

I'm sure the cpu thing is the key Charles, I preume when working on a tdm/hd rig this issue is less of a problem as most of the plug processing is handled by the dsp cards allowing the mac's cpu to mainly process the automation.

However it is my experience that on LE and with the foibles of LE'S handling of multi processor engines, the BTD does cause errors in the processing.

Basically if I have a spare bus and track i will from now on print rather than bounce. I've wasted way too much time watching the bounce handler crash and adjusting my prefs to try to overcome the problem.

Its also far easier to print alternate mixes to playlists on the mix track (easier for me anyway:) )

[weedywet]

I know it's almost a religious argument, but for ME, I find recording back in sounds better than BTD.

plus while the mix is printing I can still look ahead and change something if I am in a twiddling mood.

[Skwaidu]

As such I would say that the key determinent in succes with BTD is the complexity of your recordings.

I haven't found this to be true. I've bounced sessions with 200+ tracks and nearly as much voices with heavy automation without problems...

[Skwaidu]

I bounced to disk for a long time with no problems. I switched to printing thru an aux buss to a stereo track, not because of any problems with BTD, but because it allows me to quickly go in + print a tweak to the bridge (for example) w/o rebouncing/printing the entire mix. Big time saver.


Actually, I do this fixing of just the affected place with BTD too... I always import the file to the session when BTD:ing(for quicker recalls and comparisons) so I can just select the bridge or whatever, bounce, import, edit and consolidate.

[Charles Dye]

Yep, so did I, but it's like 3 extra steps.

I lose like a whole 45 seconds per tweak... OMG!!



:shudderdoodleft: :shudderdoodright: :shudderdoodleft: :shudderdoodright:

[recall]

I haven't found this to be true. I've bounced sessions with 200+ tracks and nearly as much voices with heavy automation without problems...

I guess you must be using HD or TDM, in which case this issue probably doesn't raise its head.

[Brendo]

Skwai also uses LE at home, I believe.

[Kenny Gioia]

As one of the early complainers of this at the DUC, I'd like to add my 2¢

I've never heard an overall sound quality issue with BTD.

But I have heard automation glitching and definitely heard Autotune not tracking the same in Auto Mode during the bounce.

No question.

My solution early on was to Print to a second computer.

I'm still doing this now.

1. It doesn't require recording onto 2 new tracks.
2. And I can Dither during the bounce instead of having to export and Dither afterwards.
3. I like having a separate stereo computer for burning CDs and DVDs while still running PT sessions.

The disadvantage is I can't just bounce the bridge as Charles is doing.

The bottom line for me is…

Why would I want to print the mix in a different way than I was listening to it while I was mixing?

It makes no sense.

[Kenny Gioia]

Why would I bounce to disk if I don't trust the computer? I mean, don't you check your bounce's after the fact anyway?

Yes. But you have to trust that some of the moves you've made are still there.

An ear can be easily fooled with minute changes.

[Skwaidu]

Skwai also uses LE at home, I believe.

Not much... Not enough to say anything about it, really.

[Charles Dye]

Why would I want to print the mix in a different way than I was listening to it while I was mixing?

It makes no sense.

I shoulda been clear. This is for client tweaks @ a later date.

[EricTheKid]

Alright, back again. And as if I hadn't caused enough trouble already, I'm going to instigate some more...

Why not just ask the source? Doesn't digidesign have some sort of stance on this?? Now, I know some of you guys probably have inside tracks to digi (and the guys who actually program the software). Surely they can't be as divided as us guys on the internet forums... I'm guessing that they know what BTD is actually doing to our precious audio.

I know, I know... we should just trust our ears, right? But if BTD has inherent weaknesses in the programming...they will eventually rise to the surface. And I think we've seen scenarios where BTD does not handle our audio how we expect it to (ie - examples listed above... auto tune, slow CPUs, LE systems, etc).

Perhaps, we're fretting about nothing. Perhaps, it's just peace of mind some of us seek. Kenny, I know you mentioned the DUC. Charles, I'm willing to bet you've talked to a programmer or two in your career. I'm just curious what role (if any) info from digidesign plays in answering this question...

[Stick]

I've recently started printing to audio tracks mostly because I've starting switching from Logic to doing all my MIDI production in PT. Well, I guess my "old" G5 won't quite keep up, because if I BTD, BFD or Ivory or whatever chokes or doesn't play at all.

That said, I've always used BTD. I can't hear the difference.

[slabrock]

Thank you, everybody, for an excellent thread.

I'm however somewhat confused. I run a PT at studio and a old Cubase SX (and i don't want to update, thank you, i already updated from VST32 :lol: and the Cubase4 doesn't seem to be any better, quite the contrary, than its predecessors) at home. In studio i always print, because we have so many hardware reverbs and stuff, that we want to use in real time. And because i was told a long time ago, that printing sounds better in PT, anyway. So i've never bounced on PT.

At home i, however, bounce (or Mix to a stereo track as the program says) if i don't have to use any outboard gear (i still haven't figured out a satisfactory way of introducing sends and returns in a bounce project). Because at home i'm always working on the limits of the computer and the cpu anyway, up to the fact that there's some audible snaps in the audio. And if i'd print, i'd have them with the audio.

But when bouncing, i can allow the computer any time it wants to work on the mix and the audio, and the more complicated the mix gets, the more it takes to do the bounce. So... did i understand correctly, that the computer is generally doing a worse job when bouncing than playing back?

I find it hard to believe. Like seagate earlier in this thread, i'm wondering if this is just a PT-specific problem?

Curious,

Slabrock

[NickH]

I have personally tested this myself on both HD & LE and I got the results to null completely every time no matter how hard I pushed the systems.

I even deliberately tried drawing simply silly amounts of automation & loading daft amounts of heavy plug in… Every single time on ether the HD or LE the files nulled…

If you are at all concerned it’s an extremely simple thing for you to test on your own system & put your own mind at rest….

I was lead to believe that there was a bug for a short time in PT V5 that caused automation inaccuracies during BTD but it was fixed before V6 came along… maybe this is where the rumour started…

I would love ANYONE that can run this test and get the files to not null on PT 6 or later to post the files up here…

Cheers

N

[Kenny Gioia]

Alright, back again. And as if I hadn't caused enough trouble already, I'm going to instigate some more...

Why not just ask the source? Doesn't digidesign have some sort of stance on this?? Now, I know some of you guys probably have inside tracks to digi (and the guys who actually program the software). Surely they can't be as divided as us guys on the internet forums... I'm guessing that they know what BTD is actually doing to our precious audio.

I know, I know... we should just trust our ears, right? But if BTD has inherent weaknesses in the programming...they will eventually rise to the surface. And I think we've seen scenarios where BTD does not handle our audio how we expect it to (ie - examples listed above... auto tune, slow CPUs, LE systems, etc).

Perhaps, we're fretting about nothing. Perhaps, it's just peace of mind some of us seek. Kenny, I know you mentioned the DUC. Charles, I'm willing to bet you've talked to a programmer or two in your career. I'm just curious what role (if any) info from digidesign plays in answering this question...

So you want Digidesign to say that BTD is flawed?

Good luck with that!!!!!

[EricTheKid]

So you want Digidesign to say that BTD is flawed?

Good luck with that!!!!!

Perhaps it is just another rumor, but that is exactly what I heard.

Now, I don't think we could honestly get them to say that it is flawed. But I believe what they would say is that they refuse to change it (to re-design/re-program it). Because it's been part of the core of the Pro Tools program for so long, it would require too large of an overhaul to PT.

I think what Digidesign expects is that we live with the quirks and adjust our workflows (IF we see fit). Again, this might be all for naught. I, personally, see some quirks with it... BUT, without finding any conclusive sonic flaws, I'll also admit that BTD does seem to stand on its own.

[Kenny Gioia]

The people that complain have found a workaround.

The others are happy.

Digidesign has nothing to gain and plenty to lose.

The happiness of customers that are quite happy.

[robmacki]

My solution early on was to Print to a second computer.

I'm still doing this now.



I'll chime in FWIW. Before PT 5.1 (I think it was v5.1)I was able to bounce via Ethernet to a second computer. It was very convenient. Keep in mind, this was around the time a 9gig drive was $1k and CPU speeds were at a snails pace. After PT was unable to perform that task I'd BTD to separate drives. For some time now, I've gone to RTD. I have some jobs I BTD and some RTD within the session.
There are a few main reasons I like RTD over BTD.
RTD:
The session is still online and you can make final adjustments.
The file waveform appears immediately so you can monitor visually as well.
The file remains in the Master Session (thanks Charles for the session guidelines)
I can spot things that may have been missed in BTD visually with a final visual check of the waveforms.

Other than that, my experience has not seen (uuuuh heard) a sonic difference, however RTD is a better QC workflow IMVHO.

Kenny how are you bouncing to another computer now?

[Charles Dye]

You're welcome, rm. (Thanks for sharing your R+BTD tips... good stuff.)

I agree... my pref for RTD is 100% about workflow.

[Charles Dye]

But if BTD has inherent weaknesses in the programming...they will eventually rise to the surface. And I think we've seen scenarios where BTD does not handle our audio how we expect it to (ie - examples listed above... auto tune, slow CPUs, LE systems, etc).

Perhaps, we're fretting about nothing. Perhaps, it's just peace of mind some of us seek. Kenny, I know you mentioned the DUC. Charles, I'm willing to bet you've talked to a programmer or two in your career. I'm just curious what role (if any) info from digidesign plays in answering this question...

Seriously, I haven't had a reason to. We've pretty clearly listed the times where it may sometimes not playback your automation exactly the same. Other than that, there is no problem with BTD.

There is no problem with BTD.

There is no problem with BTD.

There is no problem with BTD.

There is no problem with BTD.

There is no problem with BTD.

There is no problem with BTD.

(I know I'll never catch up with how many times the opposite is said, but just thought I'd get a head start. :) )




So, to be real clear... I have NEVER, EVER seen, read or heard ANY proof that BTD is flawed.

In my experience, this is a pure web fallacy.

And I'm no digi-defender. I'm just a DAW user, wanting to help other users. And repeating rumors doesn't help.

So you want Digidesign to say that BTD is flawed?

Good luck with that!!!!!

Perhaps it is just another rumor, but that is exactly what I heard.

What I do see A LOT of is posts like yours (no dis intended, ETK) where someone will say, they "heard" that there is a problem with BTD.

Quite seriously, if ur gonna say there is a problem with a product, please back it up with something, anything. But please don't say you "heard it" + leave it @ that.

Luv ya bro, no insult meant.

[recall]

think I opened a can of worms:D :Roll eyes:

[Kenny Gioia]

Seriously, I haven't had a reason to. We've pretty clearly listed the times where it may sometimes not playback your automation exactly the same. Other than that, there is no problem with BTD.



I don't understand.

If there is ever a difference IMHO then it is flawed.

I have zero time to spend listening more carefully as it's and after it's been bounced.

There is a problem with BTD.

There is not problem with BTD.

There may be a problem with BTD.

There is a problem with BTD.

There is not problem with BTD.

There may be a problem with BTD.

Your turn.

:Yawn: :Yawn:

[Kenny Gioia]

I'll chime in FWIW. Before PT 5.1 (I think it was v5.1)I was able to bounce via Ethernet to a second computer. It was very convenient.

Please explain. I'm very curious.


Kenny how are you bouncing to another computer now?

Spdif out of output 1-2 into a Tascam US224 interface on an iMac using SoundStudio software. (I hate peak - Still miss SDII)

http://images.google.com/images?q=tbn:H8jhAbC55pGr4M:http://namm.harmony-central.com/WNAMM02/Content/Tascam/PR/US-224-large.jpg

http://images.google.com/images?q=tbn:MbB1Dk5D7qLnXM:http://us.gizmodo.com/gadgets/images/imac.jpg

[Charles Dye]

The problem that has been hotly debated previously was that the audio was degraded. NOT that the automation didn't playback accurately.

I agree that if/when automation doesn't playback accurately, that THAT is a problem. But THAT is not the problem that every1 has been talking about for friggin years. I know you said you complained about it on the DUC, but almost all of the complaints I've seen are about sonic degradation.

The stuff I heard people say was that somehow during the actual bounce to disk process the "sound" was altered. IOW, it sounded differently AFTER you played it back, than it did while you were bouncing. And as far as I know THAT is simply not true. But seems to still be repeated quite often.

So, excuse my meladrama, you're right the automation thing IS a problem. No disagreement there.

[Kenny Gioia]

Excuse my perfectionism here, but what difference does it really make how or in what way it's flawed?

Sonically or Automation being missing or suddenly running in reverse.

The real question is whether BTD should be used at all.

I don't care whether the guitar sounds bad because of intonation problems or because the tone is muddy or the pickups are too close to the strings.

The bottom line is…

I'm not using that Guitar!!!!!!!

The problem that has been hotly debated previously was that the audio was degraded. NOT that the automation didn't playback accurately.

I agree that if/when automation doesn't playback accurately, that THAT is a problem. But THAT is not the problem that every1 has been talking about for friggin years. I know you said you complained about it on the DUC, but almost all of the complaints I've seen are about sonic degradation.

The stuff I heard people say was that somehow during the actual bounce to disk process the "sound" was altered. IOW, it sounded differently AFTER you played it back, than it did while you were bouncing. And as far as I know THAT is simply not true. But seems to still be repeated quite often.

So, excuse my meladrama, you're right the automation thing IS a problem. No disagreement there.

[Charles Dye]

Well I guess it doesn't technically. But my point is I used it for years + I never heard something playback differently than I expected it to.

From my POV, the noise about this seems way out of proportion with the problem. Cuz it ain't never been a problem for me. If it had, I wooda stopped using BTD a long ass time ago.

:spinningBTDpeacesymbol: :spinningBTDpeacesymbol: :spinningBTDpeacesymbol:

2 B clear...

[MacGregor]

Excuse my perfectionism here, but what difference does it really make how or in what way it's flawed?

Sonically or Automation being missing or suddenly running in reverse.


If automation is missing (aka CPU overload) you can do something,
if audio is messed up you're lost.

Mac

[EricTheKid]

Okay, so when Digidesign was questioned about the resolution of their 48-bit mixer....they released a white paper (dispelling the myths about digital summing and other false information).

Then, everyone said the clocking on the 192 I/O sucked and wasn't as good as other third party clocking devices...so they released a white paper (explaining how the 192 boasts lower digital jitter than any other device in it's class and basically said the criticism was BS).

NOW, here we are with this little debate...which seems like it's been around for some time, according to many of you. Lots of disagreement/opinions/arguments by digital audio guys about whether BTD can be used. Essentially, we seem to be asking - Can BTD be trusted? And Digidesign says... [cricket] [cricket] ...nothing

Nothing?? C'mon. This is IMHO an important step in the process. It deserves an explanation! Digi should either explain to us why we're crazy for debating this or fix it. As loyal users, don't we deserve that??

Maybe I (helped) raise this beast of an argument from the dead in this forum, but I sure as hell didn't program that BS functionality. Charles, I know you disagree. I understand your points, and I trust your ear...fully! But I don't buy into this.

As perfect as Pro Tools appears on the surface (and I love the program for the same reason many of you do)...BTD is not perfect. Any honest Digi programmer would have to admit that. ...For now, that's my statement....the proof? ....I'm working on it! :Twisted:

Much Respect to all,

ETK

P.S. Kenny, I think it's in our nature to be pefectionists. Comes with the territory (if you ask me).

P.P.S. I still luv PT

[DaveC]

I thought BTD was ropey in earlier versions of PT (4 or 5 maybe, can't remember) From 6.something on it seems to work fine for me, for the last few years. Stereo bounce sounds the same as the mix, and I haven't noticed any automation irregularities.

I never use autotune as a live plugin, always ausiosuite and print the tunings, so if that plug in has some irregularities, that could be an issue, if that's the way you work.

Excuse my perfectionism here, but
I don't care whether the guitar sounds bad because of intonation problems or because the tone is muddy or the pickups are too close to the strings.

The bottom line is…

I'm not using that Guitar!!!!!!!
... yeah - if it sounds bad. But don't throw away a guitar with bad intonation/pickups etc - it might sound great for that bridge which only needs notes at the fifth fret!

If BTD works perfectly when the CPU isn't maxed and when you don't use autotune, then that is a valid and solid way of working. Everything has its limits and its optimum, surely it is a case of what works for you.

[Kenny Gioia]

If BTD works perfectly when the CPU isn't maxed and when you don't use autotune, then that is a valid and solid way of working. Everything has its limits and its optimum, surely it is a case of what works for you.

But how do we know that?

The automation works fine when you print to a stereo track.

And how do I know when it's maxed out?

It works fine in playback.

And Autotune running live is just something I caught.

Who knows what else is out there?

[DaveC]

How do we know?

We listen. Fair enough - if you have picked up on something and decided you don't trust it. But you have to listen carefully to PT all the time anyway even without BTD - sometimes you get a click at the start of a rendered fade, on another pass you don't.

In 3 years of careful listening to mixes being BTD, they have always sounded the same later as they did when they were being printed, and there have been no automation glitches which I (or clients or labels etc) have noticed. It seems to work perfectly with my recording methodologies.

I will stop using BTD if it lets me down, or if there is good documented evidence to mistrust it. But at the moment the reports could be to do with versions of PT/plug-ins, or platform dependent, or working method dependent.

Or I might stop using it because the point about fixing a bridge later sounds like a nice timesaver!



(PS - I agree that printing on playback to a second computer sounds like one of the safest way to mix, but not all of us have 2 audio computers all the time)

[weedywet]

I didn't go LOOKING for a problem and I didn't read about it first on the "internets"


I get to a certain point in my work for the day and then I want to turn to the engineer or assistant and say "okay put it on a cd for me" and that's it.

I don't even want to sit there and listen.
And I certainly don't want to feel I need to "check" it later.

When I save or copy or print a Word file, i don't feel I need to read through it to see if everything stayed the same.

My EXPERIENCE was that BTD sounded different.
Subtle, but annoyingly worse.

I just felt DISAPPOINTED with the CD.

So I tried recording back in, and surprise. It sounds fine.

You don't find that?
Then you should BTD.

by the way, I wrote down what I found on a piece of White Paper, if that helps anyone.

[archtop]

In my BTD vs Record to 2 tracks data gathering that I just did
to gather my own data only seems to cloud the water.

I took and old mix and did stupid automation with it;
all the faders flying up and down and all the pans moving like crazy and recorded the mix to 2 tracks and did a BTD with automation on and off.

you still with me.

Could not get them to null with phase flipage.

even after the proper 51 sample alignment.

but I could get all scenarios to dim by a good 30 db.



My conclusions:

Automation had no effect.

BTD and record to 2 track are not exactly the same,
no null.


but getting a 30-40 db dim tells me the're pretty freakin' close. Right?.

[ggunn]

I'm just a lowly PTLE user reading up and trying to understand more about what I'm doing. I always print to a new track because I build my mixes on an external desk, but...

Is it possible that those who say BTD makes no diff and those who say it does could both be right on their particular hardware? The software may be the same, but there's diff's in CPU speed/type, internal bus speed, overhead from OS and other apps, yadda, yadda... is it possible that that stuff could affect some folks on BTD and not others?

I'll hang up and listen...

[Charles Dye]

by the way, I wrote down what I found on a piece of White Paper, if that helps anyone.

:lol: :lol: :lol:

Very funny, ww.

If ya heard a problem I totally believe you. Can ya describe it a bit more? Not challenging you @ all, just curious.

[weedywet]

Hey Charles, feel free to challenge me!

If it's to a duel, I choose electric foils.


What I hear is a subtle thinning or hollowness.
Like the kind of difference one hears between a 24 bit copy and a 16 bit copy.
Less depth, less substance.

I actually experience it PHYSICALLY, as though there's just something WRONG, or missing.

it sounds more DIGITAL.

[weedywet]

but getting a 30-40 db dim tells me the're pretty freakin' close. Right?.


maybe.

but as I said, I don't want CLOSE.
I don't want to even have to THINK about it.
I just want a copy made that sounds like what I left up.

not ALMOST like it.

[Charles Dye]

Can ya clarify ur original scenario a bit more, ww, cuz it was a CD u said u were disappointed with + ur description of the problem is that it sounds like 16 bit instead of 24.

[weedywet]

well, what i meant was that I do the bounce or the re-record most often to take rough mixes home.

so they BOTH end up on CD.

but I noticed that when the rough was done as a BTD there was something oddly hollow about the mix, whereas when it's recorded onto a stereo track the resulting CD sounded better.

so the final conversion to CD was done the same way in both scenarios.

[Kenny Gioia]

maybe.

but as I said, I don't want CLOSE.
I don't want to even have to THINK about it.
I just want a copy made that sounds like what I left up.

not ALMOST like it.

Exactly.

I've been listening to the damn thing for 8 hours.

When I'm done, I'm done.

I'll listen back to the first 10 seconds to make sure it's there, but that's it.

What used to suck about mixing on a real console is that mechanical changes can happen.

Not anymore. Convenience.

[DaveC]

listening to the final print is really another issue. I certainly know the feeling of 'it's down, print and forget'... but after listening to the damn song 1000 times, doesn't it make sense to pay attention one more time on the last and actually most important play through? However you print it?

Had to on analogue at the studios where I worked - make sure none of the buttons or pots suddenly developed a crackle or no mysterious modulating noise suddenly appeared on the stereo bus! Anomalies may be a lot less with PT, but they are still there, on playback as well as BTD.

[weedywet]

That's what my engineer or assistant is for.

It's not unattended.

but with BTD I can swear even IMMEDIATELY as it's running that i get that "hey, wait a minute" feeling... as though something's not right.

[Jason Phair]

I don't know a whole lot more than jack shit about Alsi.


Samplitude's bounce to disk seems to be just fine though...


Hey, fuck you! Someone asked about other platforms!

[DaveC]

As Charles Dye and others have pointed out, there are two distinct issues with BTD being discussed here: the sonic integrity of BTD, and automation glitches in BTD.

To test sonic integrity on a system, the bussing, the path and the audio conversions need to be the same. If you are converting the format during BTD to interleaved or a different sample/bit rate, then you are not testing the BTD process, you are testing the conversion.

Different systems and different platforms may be different, but I thought it best to test my home system. This is PT LE 7.3.1 on PC (XP), with an 002. Track count = 21, lots of waves plug ins including SSL channel and convolution reverb. first half was with no automation and second half with some wild volume changes. (this is NOT meant to be an exhaustive test of automation glitches - more to see if BTD was rife with such errors). All the tracks were sent via a stereo bus to an aux with an L3 on it, and then via another stereo bus (17 & 18 in this case) to a stereo track to record the mix.

I recorded the mix twice - so I could check that at least these 2 would null, and to check that my nulling procedure was not introducing any sort of errors. Then I performed 2 identical bounces.

New session, loaded in all 4 stereo audio files. Stuck the time delay plug on all channels, and used that to reverse the phase or not.

Results? On my system, any pair of audio nulled completely. Total digital silence. Didn't even need to move the audio - it was already aligned (I had selected the same region to record and to bounce, so this is not a total surprise)

Archtop - how did you bounce and mix your test tracks? How did you achieve your phase flip? And what version/platform are you using?

Anyway, I am happy that on my system, I have sonic integrity with BTD, and that at least generally, automation glitches are not all over the place.

[archtop]

Archtop - how did you bounce and mix your test tracks? How did you achieve your phase flip? And what version/platform are you using?

.

I did a few different bounces too, one recording thru digi's
mic/line 1 & 2, some thru my lucid converter, and BTD, interleavened, and dual mono.

Using 1 band E.Q. on all channels for fliping phase
PT le 6.4 on a mac G-4.

Similiar results on all tests.

No null, but 30-40 db dim.

[DaveC]

ahhh - your BTD a digital bounce, your record path is different because it is going through AD and DA convertors which will affect the sound, and also be adding a delay (I think).

To compare like for like, either send the mix out of digital outs and back in digitally (via spdif or light pipe), or send the mix to a bus and record into a new stereo channel via the bus.

[weedywet]

ahhh - your BTD a digital bounce, your record path is different because it is going through AD and DA convertors which will affect the sound, and also be adding a delay (I think).
...

not if your 'record path' is acheived by sending the mix via internal busses to a new stereo track.

no A-D or D-A

and I ddidn't convert during BTD.
in both cases exported to CD format AFTERWARD

[Extreme Mixing]

I think he was talking to Archtop on that one...

Let's face it, If you put the tracks up, phase reverse one of them and don't hear anything, then they are identical as they can be. There are reasons to RTD, but they are mostly work flow rather than quality related.

You should work the way you want. Sometimes believing something to be true is as important as having it actually be true.

Steve

[Charles Dye]

Totally agree w/ u on every point, Steve. :Thumbsup:

[PSN Big Al]

Totally agree w/ u on every point, Steve. :Thumbsup:

I'd just like to say that I have no idea what this thread is about,
but I totally agree with Charles. :Thumbsup:


.

[Charles Dye]

dude, ur funny. :lol:

[Mundox]

Greetings.
Just a related quick question.
On a 24 bit session, after we record on track and export as 16bit, does it get dithered during the Conversion process?

[Nerdyrocker]

Mundox...
Pro Tools will apply dither when you export a 24-bit file and choose that you want the file to be 16-bit. I am assuming you are using the 'Export Selected As Files' option found in the regions drop down menu (or Shift+Command+K for the shortcut nerds).

As a side note...
I rather enjoy printing my mixes on a separate track in the same session. I create the track, change it's disk allocation to be a folder inside the audio files folder called '!BOUNCES' and then name the track before I record on it (Song 1 Ruff 2.16.07). When it comes time for a different mix, I use the playlist selector on that track, name the new playlist (Song 1 Final Mix 2.16.07) and record. That way I have EVERY mix that I do for a song in the same place as the song itself, as well as all of my mix files themselves living inside of the folder '!BOUNCES' that I created. The ! is there to place the !BOUNCES folder at the top of the list inside the Audio Files folder.

I think that makes sense if you read it 3 times.

[Charles Dye]

Only had to read it veeerrry slowly 2 Xs. :D

[5down1up]

I am having the same issue here with my PT Mix System.
The bounced file does sound different compared to the session thats playing. The solution for me is recording the mixdown to a stereo track as well. And thats doing it for me.
( takes same time, takes same space & sounds better ) :grin:

[robmacki]

Greetings.
Just a related quick question.
On a 24 bit session, after we record on track and export as 16bit, does it get dithered during the Conversion process?
In Pro Tools, yes. And there has been much discussion on the limitations of and the unselectability (is that even a word?) of the dither Pro Tools uses upon Export from within the Session.

[Nerdyrocker]

A little known fact...
Pro Tools DOES apply dither automatically when you export a file.
Pro Tools DOES NOT apply dither automatically when you choose Bounce to Disk. BTD does not dither to give us the freedom to choose the dither plug that we want.

Maybe this will change some people's work flow?

[Mundox]

Hmmm. Very peculiar. I have read the opposite on couple of other forums.
Is there a link to a document about this somewhere?

[DaveC]

To the guys who get audible sonic difference between BTD and recording to disk:

1 - what version of PT are you using, and what platform

2 - can you do a A/B test and see what the dB difference is between the 2 versions?

I am not doubting for an instance that PT may respond differently in different cases, but myself and one other poster have found our BTDs null completely - ie are sonically identical (on our systems)

PS - this is not an arguement to use BTD, more a bit of scientific enquiry. I am not the possible automation glitches - that's a good reason not to use BTD, but not the issue in question.

[Nerdyrocker]

It is in the Pro Tools Reference Guide (the manual that comes with an HD system, 750 some odd pages of good reading) for 7.2.

I just finished reading it yesterday. I had heard mixed things from other forums as well but here it is from the 'horses mouth' as they say.

[robmacki]

Hmmm. Very peculiar. I have read the opposite on couple of other forums.
Is there a link to a document about this somewhere?

DUC Thread on Dithering and Export (http://duc.digidesign.com/showflat.php?Cat=&Board=UBB16&Number=1098423&Forum=,f16,&Words=%22Dithering%20question%22&Searchpage=1&Limit=25&Main=1098423&Search=true&where=bodysub&Name=&daterange=1&newerval=3&newertype=m&olderval=&oldertype=&bodyprev=#Post1098423)

PT Reference Guide pp 151, 557, 636

hth

[Mundox]

Thanks a lot guys, I'll give it a read.
:Thumbsup:

[5down1up]

To the guys who get audible sonic difference between BTD and recording to disk:

1 - what version of PT are you using, and what platform

2 - can you do a A/B test and see what the dB difference is between the 2 versions?



1 - PT 6.4.1 on a G4 dual 800, OSX 10.3.9
2 - Importing the bounced file back into the session and comparing it to the original, it just does not sound as good as the session does. The only time i hit bounce to disk is when all 64 tracks are gone :Roll eyes:

[EricTheKid]

To the guys who get audible sonic difference between BTD and recording to disk:

1 - what version of PT are you using, and what platform

2 - can you do a A/B test and see what the dB difference is between the 2 versions?

My test:

1 - PT HD 7.2, Mac OSX 10.4.8 - Dual 2.5 GHz PPC

2 - I did A/B the results and got significant dB reduction, but not a full null (during heavy automation)

Interesting thing - it would null in sections of little automation but no null in sections of the song with heavy automation.

I could actually hear it struggle and then succeed at different parts of the song.

[DaveC]

My test:

1 - PT HD 7.2, Mac OSX 10.4.8 - Dual 2.5 GHz PPC

2 - I did A/B the results and got significant dB reduction, but not a full null (during heavy automation)

Interesting thing - it would null in sections of little automation but no null in sections of the song with heavy automation.

I could actually hear it struggle and then succeed at different parts of the song.

Interesting results. Maybe the sonic differences are all something to do with automation - I guess my mixes are usually fairly minimal on automation (fair number of level rides and the odd sweep of a res filter now and again, but nothing to drastic).

[EricTheKid]

Just another note - mine was not a session with a ton of tracks, either. (Just 8 tracks, total). But heavy automation on the 2 guitar tracks in my test.

It wasn't until later in the song (when I backed off on the panning automation) that they went to null.

[DaveC]

I'll have to do some more tests on my system with some wild automation - might be a few weeks before I get a chance, depends how busy the upcoming session is.

[glitchfactor]

I have had an issue with BTD causing issues that Aux Rec didn't.

My problem, I THINK, had to do with MANY quick edits.
I had a song in PT with a drum part sliced up into 128 and 64th notes (stutter edits).

With BTD the edits wouldn't play perfectly on beat or "on grid"

When I recorded to another track it was fine!

Now I will mention that this song had a fair amount of automation and a track count of 40.
Which seems to prove the article that was posted a few pages back.


PT HD3 6.9 733Mhz G4 w/ 1GB RAM (Why won't they upgrade where I work? For the love of God?)

Justin

[6x2]

I have had an issue with BTD causing issues that Aux Rec didn't.

My problem, I THINK, had to do with MANY quick edits.
I had a song in PT with a drum part sliced up into 128 and 64th notes (stutter edits).

With BTD the edits wouldn't play perfectly on beat or "on grid"

When I recorded to another track it was fine!

WTF?

Are you saying BTD would cause audio regions (i.e. your edits) to play at different locations than during normal playback? Am I missing something crucial here?

Sooo... Did you bounce ONLY the drums for a submix or whatever (obviously checking that you bounced without any plugin delay at the master) or are you talking about the whole mix? Did it cancel phase-flipped? Was the difference audible? Did you try it with consolidated edits?

Sounds really far-fetched. I'd like to see this problem repeated.

Anyway, I've had more trouble with record to track, since ProTools seems to randomly delay one channel of a stereo track when recording while using delay compensation. I've even had a mook accept a mix with 400 samples of delay in the left channel ("This sounds really wide, cool!") I didn't say anything, but fixed it after he heard it... :lol:

6x2

[DaveC]

My test included a hell of a lot of drum edits (and some panning automation) - and nulled completely. Maybe this (BTD) is the one thing that PC can do better than mac?:icon_eek:

[glitchfactor]

WTF?

Are you saying BTD would cause audio regions (i.e. your edits) to play at different locations than during normal playback? Am I missing something crucial here?

Sooo... Did you bounce ONLY the drums for a submix or whatever (obviously checking that you bounced without any plugin delay at the master) or are you talking about the whole mix? Did it cancel phase-flipped? Was the difference audible? Did you try it with consolidated edits?

6x2

I guess I wasn't as technical about it as I should have been, but I could definitely hear a difference between the bounce and playing the session itself. It was subtle but still there. Of coarse that really means nothing. Taking it from a guy that you've never met before. AND i'm new to the forum :Wink:

[runrunrun]

Can anyone enlighten me?
I'm mixing in ProTools 7.4.1

What is the difference between:

1) using "Bounce" to create my final stereo mix file

2) Creating a stereo audio track whose input source is the output of my Master Fader, set to 2 unused buses ?

This would allow me to make a real-time "tweak" of maybe one or two elements of the mix as the track rolls by in real time.
It bothers me to have to sit there with my arms crossed, while PT bounces. Any sonic difference? Any reason I shouldn't do it the second way?

Thanks!

[Stick]

It's hotly debated, but I've switched to #2, only because of the reason you state... it's much easier to tweak, and/or punch a little change. It's also handy to always have the latest mix right there in the session (and thus the Audio Files folder).

I used to BTD all the time, and never heard any difference.

[glitchfactor]

I've found that PT bounces to disk isn't as accurate as recording to a new track. I had an industrial/techno tune with a lot of stutter edits (64th note divisions) that wouldn't play in time with BTD, but when i recorded to a new track it sounded fine.

[Charles Dye]

Been discussed @ great length here. Please use search function.

[Mixerman]

Been discussed @ great length here. Please use search function.

Oh come on. Who can be bothered with that! I've missed this discussion myself, and I find it interesting.

(uh oh...something tells me we're going to be moved).

I've found in Logic, that if I print back into the DAW while I'm making the mix, that it's not as good as when I print it to the Radar, and then transfer it in separately.

Enjoy,

Mixerman

[Cheech]

I've found in Logic, that if I print back into the DAW while I'm making the mix, that it's not as good as when I print it to the Radar, and then transfer it in separately.

Enjoy,

Mixerman

Do you record it digitally to RADAR, then use RADAR to do the sample rate conversion and what not?

[Mixerman]

Do you record it digitally to RADAR, then use RADAR to do the sample rate conversion and what not?

No, I'm going into the Radar analog through summing boxes and an SSL compressor, and then out of the Radar digitally into Logic. But when I just use the Radar as the converters, and record the digital 2-tracks back to the DAW simultaneous to playing the mix, I find it to be less accurate than printing to the Radar and the digitally transferring it (to a spot on the palette that there is no waveforms). There really should be no difference.

Now, I don't know if I'm imagining this or what. I suppose that's a possibility. But at the moment, I don't think so. When I have a second to experiment, I'll try to confirm whether it's my mind playing tricks, or an actual problem.

Mixerman

[tannoy]

Hi,

so Mixerman, how is it going with the human suffering due to the change to mixing itb ? There were no news for some time...

(Sorry for a little off-topic, just curious)

best,

Tannoy.

[Charles Dye]

Much was said in the earlier thread (now merged). Some great points made for each approach. It's pretty lengthy discussion, but please feel free to continue here. Just wanted to prevent rediscussion of the same things.

Carry on... :D

[Kenny Gioia]

I never use BTD.

It seems crazy to me to print your mix in a way that you haven't been hearing it while you were mixing.

Although it might be identical. Who knows?

I just don't see the benefit of using it.

[Bob Olhsson]

The ability to compare or even comp between/punch into different mix versions is powerful.

[tannoy]

Hi there,

I'm not using ProTools, I'm into Cubase - but I think this is a platform-independent (non?) phenomenon.

From my point of view there shouldn't be any audible differences between the 2 methods - I think the processing/internal maths are the same in both cases, so why should it sound different ? I'm not sure about the automation thing, till now I didn't experience any problems -but maybe it occurs at certain setups.

So does anyone really hear a difference (automation issues aside) ?

Best wishes,

Tannoy.

[6x2]

No, there shouldn't be a difference. But there is. At least in automation in CPU-heavy sessions in PT LE. I've not tested in HD.

BTW, Cubase is a weird one, cause as far as I remember, I couldn't even LISTEN to the bounce while it was doing it! That was still a few SX versions back. I suppose they have fixed this?

6x2

[tannoy]

Hi,

yes, the newer Cubase-Versions (as far as I know SX 3.1.1.x and higher) support realtime bouncing where you can listen and adjust changes while exporting the mix but also they are still able to do the faster 'underneath' bouncing where you don't hear a thing.

I'm still using the 'vintage' Cubase VST/32 - there is no built in real time audio export, if I want to do it, I've got to route the audio internally for that purpose. But to be honest, I've never done that - I make my adjustments before bouncing and use the automation.

So for me there's no reason to hear something while bouncing to disk - with Madonna's words: "I've heard it all before..."

But that's just me - I guess there are many philosophies about this...so anyway how it's done, if it works, it's ok.


take care,

Tannoy.

[tamasdragon]

There is no sonic difference between the methods, but record back to track has several advantages as mentioned above.
Tamas Dragon

[6x2]

There is no sonic difference between the methods,

Apart from the automation irregularities...

but record back to track has several advantages as mentioned above.
Tamas Dragon

Yes it certainly does.

To be honest, the automation inaccuracy is not that catastrophic most of the time, it's more the thought of not getting what you hear which is upsetting. That's why I record to track.

6x2

[Oreganmusic]

I always record to track.....just doing what has always been done (minus the tape) :Wink:

[Charles Dye]

I've merged the 2 threads.

Please read the previous posts. There are some good ones. Many of the same points were discussed.

Thx!

[6x2]

Aaaah... I was wondering how those 98 new posts got there in a matter of hours... :D

6x2

[Chris Lambrechts]

I use both ... BTD ... print to disc.

For me it depends what the final purpose is and what it is going to be used for.

Before I give a few examples ... I just wanna say that on a healthy system I don't think there should be any differences.

It is not like I test this with every mix I print but I have done pretty extensive testing using both methods and fwiw ... every single one of my tests I have ever done ... both the BTD and the printed and even the exports ... they have ALLWAYS cancelled out. (keeping logical conditions in mind of course as there are : you can't export an audio file that has a hardware insert on it and then do a 'null' test against a bounce .... duh right ?


Anyhow :

- typical BTD scenarios :

I need to send an mp3 ... BTD to mp3
I need to burn a quick CD to listen to on my way home ... BTD to stereo interleaved 44.1/16
I need to turn a snare hit into a sample .... BTD
I do home mastering ... stereo session with printed mixes ... BTD

typical 'print' scenarios :

Oh and fwiw ... if I print ... I print to a seperate folder in the disc allocations menu

I need to prepare to go to mastering ... print .. multi mono files.

midi tracks ...... bus ..... print.



Chris