Destructive Record/Punch In - Pro's and Con's

[LEX]

Thanks Vin. Wow, this thread has grown into a beast. Funnily enough, Nuendo 4.3 might be my last version. If I upgrade to Nuendo 5 that will definitely be my last. Especially if it's 64 bit Windows 7 viable.

I'm moving my studio to the dark side. Protools HD. Too many of my clients and most other tracking and mixing studios in Sydney are Protools. They just look at me funny when I say Nuendo. I've been a Steinberg supporter for a long time but now it's time to move. I need to be able to send session files anywhere in the world without any AAF, OMF2 or any other translator grief.

I know, I know, you'll all call me crazy. This is business, and the majority of pro studios are running Protools.

Maybe, down the track, I'll run an emulated copy of Nuendo 6. Just to see how it stacks up. :icon_razz:

Cheers

While Nuendo might be a "cheaper" less expensive option, in the long run a PT HD will bring you bigger returns on your investment.

I don't know if you are just Post, or Music and Post, but when you start moving into bigger and better projects, everyone is just on ProTools.

It comes to wonder, did Steinberg really ever want to push Nuendo more or do they not care to explore the "reality" of the Post market and the user needs.

I'll give you 3 examples of what they could have concentrated on.

1) User configured 5.1
2) mono to stereo, mono to 5.1 channels.
3) Destructive record/destructive punch

Just to name a few "necessities" in the Post world.
Let's not forget VIDEO.

You'll probably see your investment pay itself off pretty quickly.

And, should there be a "Nuendo" client in the future, you have it.

LEX

[Animus]

Lex, why is #3 important? I am curious.

[tunetown]

Exactly. I recon it's a no brainer too. It's a shame. I do enjoy Nuendo and it feels so natural. No doubt I'll have a learning curve but that's ok.

How's it run with Vista64 and UAD2 on an i7?

Sorry for the off topic Mods. Move to a new topic if you feel it necessary. Although, Steibergs policies and lack of pro market support is part of the reason for the move.

Cheers

[Animus]

It's sad really about people moving to Protools as Nuendo is potentially a superior product. It pisses me off to no end that Steinberg want to to take it into a "post" direction and not a "be all do all" solution a la Protools. I am sticking to Nuendo for the foreseable future but I have my eye on the Studio One development and also there is word that Digi are working on something "big" and native. If Protools gets there midi up to par and full ADC in a native solution (and step back from proprietary crap) I would probably switch in a heartbeat.

[tunetown]

It's sad really about people moving to Protools as Nuendo is potentially a superior products. It pisses me off to no end that Steinberg want to to take it into a "post" direction and not a "be all do all" solution a la Protools.
Agreed. The status of Nuendo has really been diminished by the whole post debacle. It never really took off in pro circles. Damn shame. It could be the great DAW. Totally unrealised potential.

Cheers

[LEX]

In post, you are punching in on stems all the time.

Once you've done your mix, and laid back stems, you usually review and make changes.

It saves time to lay back the stem once the reel is done mixing and punch in at the spots you or the client wants to change.

Now, with time and budgets getting smaller, what happens when you punch in non destructively.

You have hundreds and hundreds of pieces of files. Now in Nuendo, that maybe fine, but not when the client wants the audio files of the stems.

So with Nuendo, you'd have to take the time to bounce down the stems to a single file.
While you can do it non realtime, there is no quality control of the bounce. So once it is bounced, you have to listen to it to make sure there aren't any ticks, drop outs ect.
Once you've fixed that, you have to do the process all over again which can eat up hours.

We've had clients all the time want to change a music cue, or the producer wants to make a change after we've finished mixing.

Punching in destructively overwrites the file, but you still have 1 file per track instead of many.

This speeds up the process of delivery and dubbing by days if not weeks. And if a client has 1 week in the budget to mix, but you need to add 3 days to it because Nuendo is non destructive, then I wouldn't be surprised if they went elsewhere.

Clients have deadlines with alot of things. Just as stages have clients behind clients.

This is one area SB doesn't seem to understand.

LEX

[LEX]

Agreed. The status of Nuendo has really been diminished by the whole post debacle. It never really took off in pro circles. Damn shame. It could be the great DAW. Totally unrealised potential.

Cheers

It would have, but the software engineers were trying to tell the Post engineers "how" Post should be done.

That was the mistake.

2nd mistake was there were plenty of Post users who mentioned several needed features that was ignored.

They easily could have gotten a PT system to see how people were doing it and adapted, just like PT 8 took many features from Nuendo that PT users wanted.

LEX

[Audiocave]

Excellent explanation, I finally understand the importance of that feature.

Question: When you do that do you back up the file first just in case you "miss"?


In post, you are punching in on stems all the time.

Once you've done your mix, and laid back stems, you usually review and make changes.

It saves time to lay back the stem once the reel is done mixing and punch in at the spots you or the client wants to change.

Now, with time and budgets getting smaller, what happens when you punch in non destructively.

You have hundreds and hundreds of pieces of files. Now in Nuendo, that maybe fine, but not when the client wants the audio files of the stems.

So with Nuendo, you'd have to take the time to bounce down the stems to a single file.
While you can do it non realtime, there is no quality control of the bounce. So once it is bounced, you have to listen to it to make sure there aren't any ticks, drop outs ect.
Once you've fixed that, you have to do the process all over again which can eat up hours.

We've had clients all the time want to change a music cue, or the producer wants to make a change after we've finished mixing.

Punching in destructively overwrites the file, but you still have 1 file per track instead of many.

This speeds up the process of delivery and dubbing by days if not weeks. And if a client has 1 week in the budget to mix, but you need to add 3 days to it because Nuendo is non destructive, then I wouldn't be surprised if they went elsewhere.

Clients have deadlines with alot of things. Just as stages have clients behind clients.

This is one area SB doesn't seem to understand.

LEX

[Animus]

thanks Lex for the explanation.

[LEX]

No need to backup anything. If you are fixing it, you are punching in.

If you miss the punch, you roll back and punch again.

LEX

[TAFKAT]

I vote we move the Post related stuff to a new thread, Man, I am running out of thread title ideas.. LOL

Give me an idea on what the thread title should be for the Post related discussion, and I'll create the thread and move the posts.

[Animus]

Are you bored Vin? hehe Knock yourself out. These are called "threads" not a "thesis". :icon_lol:

[TAFKAT]

I am chairman of the bored Stacey, you know that...

[Animus]

:D

[paulwr]

I am chairman of the bored Stacey, you know that...

VERY good one. I'm impressed!

-Paul

[Livewire]

Yea, I really wish nuendo had destructive punch in. Ever since I've been using the new Fairlight dream system, it's tape mode has saved a bucket load of time on output. If only SB implemented non-destructive clip based eq as well (without having to use undo history) it would also make dialogue mixing quicker.

[dcwave]

If only SB implemented non-destructive clip based eq as well (without having to use undo history) it would also make dialogue mixing quicker.

AMEN!!!!

[RiffWraith]

One of the things that I really like about Cubase is that it is NON-destructive.

That said, LEX has an excellent point about workflow. So are you guys wanting a switch, or an option to do either/or?

[LEX]

One of the things that I really like about Cubase is that it is NON-destructive.

That said, LEX has an excellent point about workflow. So are you guys wanting a switch, or an option to do either/or?

Most other daws, even hardware recorders, have destructive and non destructive record.

The fact that SB apps don't is a bit retarded considering the technology for both has been around for almost a decade.

LEX

[kdm]

Yea, I really wish nuendo had destructive punch in. Ever since I've been using the new Fairlight dream system, it's tape mode has saved a bucket load of time on output. If only SB implemented non-destructive clip based eq as well (without having to use undo history) it would also make dialogue mixing quicker.

+10 on both accounts.

[TAFKAT]

Moved related posts to new thread..

[LEX]

I have 2 schools of thought on this.

1) Steinberg doesn't know or can't program destructive record with sample accuracy.

2) They figure that there is no need to have individual files as a whole for other programs.

Except, that depending on the layback house or delivery requirements, this is completely opposite.

Just like tape, you need to be able to monitor your punches and your bounces.

LEX

[tunetown]

This seems like a no brainer. A little switch enabling destructive or non-destructive.

[LEX]

This seems like a no brainer. A little switch enabling destructive or non-destructive.

Alot of things are no brainers, that's the problem.

SB doesn't get "basic concepts". So they are often left out.

Like I said before, SB's approach is the software engineer tells the engineer how to do his job.

LEX

[olamo]

Like I said before, SB's approach is the software engineer tells the engineer how to do his job.
LEX

A very strange approach indeed.. It's a thin line between innovation and stupidity.

Ola

[MattiasNYC]

Well, first off even PT didn't have destructive "punch record" for the longest time. And by that I mean that when you did a punch in the middle of a file it wouldn't simply replace that area and automatically consolidate it the way it does now.

Secondly, I'd say that to be fair to Nuendo, the non-realtime mixdown is fantastic. Sure, the ability to do a destructive punch in PT is perhaps useful, but then again PT can't do faster than realtime bouncing.

And this really boils down to trust in the DAW one is using. Lex, you mentioned that you'd have to check your non-realtime mixdown in Nuendo; well, I'd say that whenever you do something in a DAW you'll have the choice between checking it or trusting the app. PT has thrown curveballs at me several times and now I have to visually and audibly check all these punch ins, x-fades and consolidations every single time I do an edit/punch record.

(And to clarify, I'm not talking about a design flaw in the app but about the app either having bugs or screwed up preferences. As an example, in PT 7.4.1 I had instances where the recording into a track (from aux) in PT would result in one "delay". Punching in on those areas would delay the signal further by a hair, making the default x-overs too "early" so that one would perceive a frame of silence before the new recorded area was heard. It wasn't audible at the time of the punch/record, only on subsequent playback over the punch-in/-out points leading to the risk of delivering a faulty file if trusting ones ears on record.)

So to recap:

I think this is an area where the two apps, PT and Nuendo, are actually "tied". It's all a matter of workflow and how one trusts various parts of the respective apps.

[LEX]

Well, first off even PT didn't have destructive "punch record" for the longest time. And by that I mean that when you did a punch in the middle of a file it wouldn't simply replace that area and automatically consolidate it the way it does now.
It still had destructive record. You could still punch into it. Destructive punch you didn't need to set and in or out point.
So, it was still faster.


Secondly, I'd say that to be fair to Nuendo, the non-realtime mixdown is fantastic. Sure, the ability to do a destructive punch in PT is perhaps useful, but then again PT can't do faster than realtime bouncing.


Nope. Can't do it.

And here's the reason:
http://nuendo.com/phpbb2/viewtopic.php?t=21362

You MUST listen to your material. Any rogue tick or problem, you'd be fucked.
While the non realtime bounce is nice, I've heard issues before after the fact.

To ensure absolute Quality Control, you must listen to the layback. Non realtime bounce is unacceptable.

Just so you know, most Post places had MX2424's or MMR's for layback because of the destructive punch not in PT.
But, with either MMR or PT, you can listen to your playback as you cannot trust the application to always do the job.

Your ears do the job, and with Nuendo's rep, I wouldn't trust it AT ALL!

LEX

[colony nofi]

Ah - but even when you "listen" as pro tools HD mixes down, and it sounds right... I've had errors in the resulting file that I would have never found if i didn't re-listen to the files themselves. (Audio-ease snapper is great for this...)

So, in my mind, its quicker in nuendo... bounce (non-realtime) and listen back thru snapper (and you have visual feedback of peak level etc to check for accidental overs....)

In PT - mix out the file in realtime, and then spend the time again listening to the master file.

Sure - I don't do this second check in a fast turn around environment (I know the mixes will be checked before going to broadcast anyway) - but for film scores, or doco mixing, I ALWAYS do it.

As for destructive record... I love it.
(radar anyone...man they know how to crossfade quickly and elegantly)

N4 could REALLY do with it for mixing / stem output etc etc.

[nikki-k]

Non-real time is huge for me. Bounce individual tracks down.. such as VI's.. absolutely. And as mentioned, no matter what, MUST re-listen after a bounce. Although a real-time listen during a bounce may sound 100% fine, I have found problems can be present in the resulting file.

IMO, Avidesign and Steinberg are the same breed. And the apps each have their own long lists of pros and cons. I just wish Reaper were not a sprawling mess. Ah, well.

[MattiasNYC]

It still had destructive record. You could still punch into it. Destructive punch you didn't need to set and in or out point.
So, it was still faster.

Hmmmm.... so that was the only difference in v7? I thought it was more severe.... like replacing whatever audio was on the track you recorded destructively on again. (I know it had destructive record for the longest time, the issue was what was destroyed and what was consolidated into a new file).



Secondly, I'd say that to be fair to Nuendo, the non-realtime mixdown is fantastic. Sure, the ability to do a destructive punch in PT is perhaps useful, but then again PT can't do faster than realtime bouncing. Nope. Can't do it.

And here's the reason:
http://nuendo.com/phpbb2/viewtopic.php?t=21362

You MUST listen to your material. Any rogue tick or problem, you'd be fucked.
While the non realtime bounce is nice, I've heard issues before after the fact.

To ensure absolute Quality Control, you must listen to the layback. Non realtime bounce is unacceptable.

Just so you know, most Post places had MX2424's or MMR's for layback because of the destructive punch not in PT.
But, with either MMR or PT, you can listen to your playback as you cannot trust the application to always do the job.

Your ears do the job, and with Nuendo's rep, I wouldn't trust it AT ALL!

LEX

Well, I actually think you reinforced my point: "You MUST listen to your material." Period. This goes for both DAWs.

You had problems in Nuendo (I presume - or at least you don't trust it), and I had issues with Pro Tools, and I don't trust it (enough). Thus we both have to listen to the final file when doing our work. So that issue isn't in favour of either.

[MattiasNYC]

Ah - but even when you "listen" as pro tools HD mixes down, and it sounds right... I've had errors in the resulting file that I would have never found if i didn't re-listen to the files themselves. (Audio-ease snapper is great for this...)

So, in my mind, its quicker in nuendo... bounce (non-realtime) and listen back thru snapper (and you have visual feedback of peak level etc to check for accidental overs....)

In PT - mix out the file in realtime, and then spend the time again listening to the master file.

This was the kind of work-flow I was getting at.....

[LEX]

Ah - but even when you "listen" as pro tools HD mixes down, and it sounds right... I've had errors in the resulting file that I would have never found if i didn't re-listen to the files themselves. (Audio-ease snapper is great for this...)



I assume you mean if you are mixing in PT. With 3 PT stages, and just about every stage around town here has PT HD as a stem recorder, I've never had a single problem.

In fact, When I do punch and it is in a drone, I can hear the bad punch. So I go in earlier.

We've got a Pro Control Room that we do dialog predubs in, and there has yet to be an issue recording back in, in about 80 plus predubs.

I can only assume these "issues" were earlier versions of ProTools, so I can understand the "reluctance".

But, when users complain about bad non realtime bounces in v4.3 of Nuendo, well that is a pretty new version and is down right scary.

LEX

[MattiasNYC]

I assume you mean if you are mixing in PT. With 3 PT stages, and just about every stage around town here has PT HD as a stem recorder, I've never had a single problem.

In fact, When I do punch and it is in a drone, I can hear the bad punch. So I go in earlier.

We've got a Pro Control Room that we do dialog predubs in, and there has yet to be an issue recording back in, in about 80 plus predubs.

I can only assume these "issues" were earlier versions of ProTools, so I can understand the "reluctance".

But, when users complain about bad non realtime bounces in v4.3 of Nuendo, well that is a pretty new version and is down right scary.

LEX

Well, once again to be fair and consistent Lex, isn't this the same principle that so many critics of Nuendo/Steinberg "disqualify"? The short version... sort of:

"Well, I don't have this problem so it must be your individual system".

Now, we can accept that as a valid argument, or not. I think it could be. And so it then cuts both ways.

The link you posted to with the one user having a problem with v4.3 doesn't then prove that there are problems in general with the software (although it certainly could be the case). And just as you mention above, there seem to be other users without that problem on the same version.

In addition, it doesn't really matter if the version is "new" per se (and I take it literally - assuming you didn't mean "recent"), because any bug will always at some point be present in a "new" release. PT actually also had a bug where destructive punch-ins resulted in no audio in the punched area. Obviously Digi takes that more seriously and addresses the issue faster than SB likely would have.

But still, I'd say that the two are definitely tied on this issue.

[kdm]

I've had enough problems with errant bounces (and it only takes one) that I never trust Nuendo to bounce offline for a critical or final mix.

If I absolutely need to speed up the render time for something like a 15-20 minute corporate promo, I might run two bounces, then invert-phase sum them and check the sum-bounce for clicks/pops, but for shorter pieces and longer features/docs, realtime is the only reliable solution.

That said, I've done spots and shorter corporate mixes non-realtime and been fine, but that was only with a longtime client that I know I can ftp over a remix in 5 minutes if there is a problem, and on smaller mixes, the chances for problems are much lower. On a feature? I never trust it to offline.

I too would like to see destructive record for changes to final mixes, though I would still listen to the punch section in the final - but faster than running down the full length mix in realtime.

[LEX]

Well, there are a number of other things as well.

In PT, if you have a bunch of clips, you can do a "consolidate" which creates new audio file without bouncing.

Basically like gluing, but it creates an audio file that can be used anywhere.

Also, when it comes down to destructive, I do alot of shows that have multiple versions in which the stems are conformed to accommodate act breaks.

But often times, we get QC reports after the 2nd version has been done.
I fix the original stem, copy/replace in the audio files folder for V2, and the stem is updated.
Takes about 5 minutes to transfer the files and relink them.

In Nuendo, you'd have to rebounce the audio files and rename them.

In terms of "some systems" comment, when there are 6 different PT systems all running fine, plus all the ones all over town both PPC, Intel and I've yet to hear of any issue, I have to wonder where the "defense" really comes in.

To make a stronger point. I've had non realtime export errors in every version of Nuendo or Cubase.
This is across about 7 different MAC's and even more different PC's with an array of different audio cards.

In ALL cases, it would happen. Not ALL the time, but when you are on a deadline, trust Nuendo to function properly, only to find you delivered an audio file with a "tick" in it, confidence is low.

I recently, within the past week, experienced it with Cubase 5. Built in synths, 9 seconds long, some automation, no effects.

Non realtime bounce produced 2 "ticks" in the audio file. I did a realtime bounce and there was no tick.
Never tested to see if it repeated on another non realtime bounce.

LEX

Just thinking, maybe it is a Floating Point math error in Cubendo and it can't always round the numbers correctly.

[MattiasNYC]

Well, there are a number of other things as well.

In PT, if you have a bunch of clips, you can do a "consolidate" which creates new audio file without bouncing.

Basically like gluing, but it creates an audio file that can be used anywhere.

Which you can also do in Nuendo, no?


Also, when it comes down to destructive, I do alot of shows that have multiple versions in which the stems are conformed to accommodate act breaks.

But often times, we get QC reports after the 2nd version has been done.
I fix the original stem, copy/replace in the audio files folder for V2, and the stem is updated.
Takes about 5 minutes to transfer the files and relink them.

In Nuendo, you'd have to rebounce the audio files and rename them.

I'm not sure I understand the process or terminology, because it doesn't sound all that different to me in terms of time spent and difficulty/ease of process!?


In terms of "some systems" comment, when there are 6 different PT systems all running fine, plus all the ones all over town both PPC, Intel and I've yet to hear of any issue, I have to wonder where the "defense" really comes in.

To make a stronger point. I've had non realtime export errors in every version of Nuendo or Cubase.
This is across about 7 different MAC's and even more different PC's with an array of different audio cards.

In ALL cases, it would happen. Not ALL the time, but when you are on a deadline, trust Nuendo to function properly, only to find you delivered an audio file with a "tick" in it, confidence is low.

I recently, within the past week, experienced it with Cubase 5. Built in synths, 9 seconds long, some automation, no effects.

Non realtime bounce produced 2 "ticks" in the audio file. I did a realtime bounce and there was no tick.
Never tested to see if it repeated on another non realtime bounce.

LEX

Just thinking, maybe it is a Floating Point math error in Cubendo and it can't always round the numbers correctly.

Hmmm.... I'll have to check this myself.

When I did encounter problems before they were due to the UAD issue. Since then I haven't encountered a problem and so I'll try this later to see if this happens to me as well.

However, the point still stands in terms of the principle: If your argument is that there can't be a problem with PT because it works for you, then you should accept the same line of reasoning when people say the same about Nuendo.

[LEX]

However, the point still stands in terms of the principle: If your argument is that there can't be a problem with PT because it works for you, then you should accept the same line of reasoning when people say the same about Nuendo.

No. My point is, while there may have been problems in the distant past with PT, probably V5, v6.4, they have been fixed and fixed for a very long time.

While with Cubendo, users are still experiencing these issues on many different systems even without UAD.
I didn't have a UAD on my laptop, and only used 3 builtin VSTi's and no effects and there was an issue.

Dedric has experienced it as well and I believe is on the current version of Nuendo.

Saying Nuendo is fine, is a problem. We experienced all of this with Solo/Mute and Video.

It would be nice if SB and the "defenders" would acknowledge there are problems and fix it.

LEX

[MattiasNYC]

I'm not saying it's fine. I'm also not saying there isn't a problem.

Other than that I'll just reiterate that the principle of your argument IS the same. Go back and re-read what I wrote in conjunction with your posts. The argument is the same in principle. That's all I'm saying.

[colony nofi]

No. My point is, while there may have been problems in the distant past with PT, probably V5, v6.4, they have been fixed and fixed for a very long time.
The issue I had was with Protools 7.4HD... and digi/avid guys know ALL about it. We've had them in showing them / repro-ing the problem.

Anyhow...

From where I sit...
* It would be great if cubendo had destructive record. No doubt it is very useful in some post pro circles
* However, there are plenty of work arounds for this, some mentioned here, others not. (I use a series of macro's and QuicKey assignments together with the "bounce selection" command) Sure it is a different paradigm of working, but... it still works! You get the same result, and in many cases, it's just as quick in nuendo.
* ALL audio mixes need to be checked. Realtime or not. Even if it is when you are striping back to tape, or attaching the audio to the HD Video file inside quicktime.

b.

[olamo]

ALL audio mixes need to be checked. Realtime or not. Even if it is when you are striping back to tape, or attaching the audio to the HD Video file inside quicktime.
b.

+1

Is it more safe to render in Realtime? probably yes..
Do you need to check your mixes afterwards, regardless of realtime/nonrealtime? Yes, no doubt

PTHD did occationally render with errors, seldom but it happened to me several times during a period over three years - more often PT crashed during the bounce process than errors silently sniking in.

I also used offline bounces from Nuendo in the same period and honestly did not experience a huge difference between the two in this regard.

The lack of offline bouncing as an option in PT was getting on my nerve more than once ( and still does.. ), it should at least be an option. I agree that we should have destructive record in Nuendo, but the option to bounce offline is a good thing IMO.

Ola

[LEX]

I think we are all talking about 2 different things here.

Seems like everyone is talking about PT Bounce vs Nuendo bounce.

I am talking recording. In other words, I'm not going to "render" a mix.

I am going to have input channels I record to, destructively.

In Post, you stop and start often and punch in. I have a 5.1 record channel stems set up and I record in realtime.

Realtime "bounce" does nothing in terms of use for me in this case. I need to be able to monitor what I am recording and be able to destructively punch into those files.

Destructive record (hence the title) vs offline/realtime bounce.

2 VERY different things.

LEX

[colony nofi]

Seems like everyone is talking about PT Bounce vs Nuendo bounce.

I am talking recording. In other words, I'm not going to "render" a mix.

I am going to have input channels I record to, destructively.

In Post, you stop and start often and punch in. I have a 5.1 record channel stems set up and I record in realtime.

Realtime "bounce" does nothing in terms of use for me in this case. I need to be able to monitor what I am recording and be able to destructively punch into those files.

Destructive record (hence the title) vs offline/realtime bounce.

2 VERY different things.
Hey Lex!
Don't know if you were referring to me, but I'll share my workflow anyway to show how I get around the lack of destructive record. And don't get me wrong - I think the feature is sorely lacking... but...

So I put my stems into record ready in nuendo. Usually around 24 mono tracks.
Arm those tracks
Hit record.
Hit stop...
Go back. Fix up something on the mix
Set punch in
Hit record
hit stop...
etc. All day long. End up with piles of regions on top of each other. BUT. The most recent is on top, which is what I go for. Select the whole lot, and "bounce selection". (I have this as a key macro). Then I select all the other audio files in the pool and chuck them out. Takes about 2 mins.

Does that make sense? Sure, its not exactly the same as using destructive record but it gets the job done no?

Cheers! B.

[Sam]

not really the same thing brendan - the bounce results in an entirely new file you have never heard before so you must QC the whole thing again anytime you make a new change and rebounce....with destructive punches you are actually altering your source/mix file as you go with no need to bounce, and what you are hearing in your session as you punch in sections is exactly the same file you will deliver - so if it sounds right as you go and you check your punches as you as any sane person would - then your delivery files is what you were listening to when mixing....hope that is what you were asking mate...

[Captain Caveman]

I always thought that Bounce selection didn't export the audio through the engine in a way that could result in clicks and pops, but did it in a more "joining up the files" kind of way. It does multiple tracks so quickly that it doesn't seem to anyway.

so if what I am thinking isn't completely off the mark, in the same way that destructive record leaves the stuff before and after the punch in/out points alone, bounce selection does too because it is just copying the data.

Am I tripping?

[MattiasNYC]

I always thought that Bounce selection didn't export the audio through the engine in a way that could result in clicks and pops, but did it in a more "joining up the files" kind of way. It does multiple tracks so quickly that it doesn't seem to anyway.

so if what I am thinking isn't completely off the mark, in the same way that destructive record leaves the stuff before and after the punch in/out points alone, bounce selection does too because it is just copying the data.

Am I tripping?

That was my understanding as well.

[MattiasNYC]

not really the same thing brendan - the bounce results in an entirely new file you have never heard before so you must QC the whole thing again anytime you make a new change and rebounce....with destructive punches you are actually altering your source/mix file as you go with no need to bounce,

I'm not so sure that's a conceptually reasonable way to think about it, in the context we're discussing.

Think about it this way:

ProTools - You record the first file. Let's say you get about 10min into the show. Then it is decided that something needs to be tweaked. So you stop recording. You now have file "A". You roll back, say, 2 min and trim your mix while recording back onto the same track again. Now you get another 5min into the show. You stop. Your file "A" is now a 15 min file, as opposed to what it was before, 10min.

BUT, the revised file is based off of your original, right? So it's really a matter of splicing new data onto/into a previous file.

In Nuendo we do the same thing. You record 10min. Stop. You have file "A". Roll back, tweak while recording and stop - file "B". You consolidate the files into a new file, "C". But what's really happening is that B is spliced onto/into A, yes? So sure, you end up with a new file, and as a result have 3 files total. But I have to wonder if the process practically speaking is of any significance whatsoever as far as reliability goes when compared to the PT way of doing it.

I'm just not seeing how it would, mathematically, make any difference. In both cases you have information already to which you append/splice new information. How many ways can that really be done? And if the only concern is copying information as opposed to only appending/splicing it, then we have much bigger issues to worry about instead. After all, if a simple file-copy command would so frequently result in erroneous copies, the whole world would be in a terrible state (think copied files in financial institutions, legal institutions, government etc)

And at the end of the day the destructively recorded into file you haven't heard in its entirety either. Just like the consolidated "file C" it is actually "new" to you, and they're both supposed to do the same thing.


and what you are hearing in your session as you punch in sections is exactly the same file you will deliver

That doesn't really jive 100% with:


- so if it sounds right as you go and you check your punches as you as any sane person would -

Maybe I'm nitpicking though.

To reiterate a point I made before though, it seems to me as if we can carry on this conversation on a couple of different "tracks"; one being one discussing facts and absolute behavior of the various DAW's from a technical standpoint, and the other being one where we try instead to think about how our workflows and emotions affect this issue.

It doesn't seem as if we have too much to go on when it comes to the former, but with the latter I'll point out, once again, that some would never even release the final files - i.e the ones recorded (destructively) into and/or consolidated files - without listening through all of it one final time. IF that is ones workflow, stemming (pun) from ones lack of 100% trust in DAW's, then the difference is close to zero. Because one can then either add to ones workflow the faster-than-realtime export in Nuendo, and do final QC to that file, or simply consolidate all the spliced tweaks into a new file and do final QC to that file. Either way one would have to do the same in PT with the file recorded destructively. Same consumption of time really (except for faster-than-realtime export which adds a wee bit). Only practical difference would be more hard drive space consumption in the case of consolidation, but only until you delete the superfluous and now "old" and unused files.

PS. Still need to test this.... Perhaps today I'll run some tests....

[Captain Caveman]

Good points m2, the whole premise of this thing is that bounce selection can produce clicks. I've always looked at it as a file-joiner-upper and have never even thought about clicks n pops, let alone had any. I've searched Cubase.net for bounce+selection+clicks and got a result page of 6 threads, none of which are about this non-existant issue.

If in this example a seperate record folder is selected for the stem tracks then all it means (if all are recorded and bounced simultaneously) that the files with the highest (and same) "-x" number are the ones to keep. A fancy pants batch renamer would remove the "-x" from the files for the ol' FTP.



But often times, we get QC reports after the 2nd version has been done.
I fix the original stem, copy/replace in the audio files folder for V2, and the stem is updated.
Takes about 5 minutes to transfer the files and relink them.

In Nuendo, you'd have to rebounce the audio files and rename them.
As we have clarified that bouncing in Cubendo isn't the same as exporting is the issue now basically how long it takes to rename the files?

[LEX]

Good points m2, the whole premise of this thing is that bounce selection can produce clicks. I've always looked at it as a file-joiner-upper and have never even thought about clicks n pops, let alone had any. I've searched Cubase.net for bounce+selection+clicks and got a result page of 6 threads, none of which are about this non-existant issue.

If in this example a seperate record folder is selected for the stem tracks then all it means (if all are recorded and bounced simultaneously) that the files with the highest (and same) "-x" number are the ones to keep. A fancy pants batch renamer would remove the "-x" from the files for the ol' FTP.


As we have clarified that bouncing in Cubendo isn't the same as exporting is the issue now basically how long it takes to rename the files?

I think we are getting to an understanding here.

Though I haven't seen it posted, I've had bounce pops as well.

Yes, and the file renaming. Kind of a sore spot as well.

I think I've made my points several times, so I'll just read back in when there are updates.

LEX

[Daryl]

Good points m2, the whole premise of this thing is that bounce selection can produce clicks. I've always looked at it as a file-joiner-upper and have never even thought about clicks n pops, let alone had any. I've searched Cubase.net for bounce+selection+clicks and got a result page of 6 threads, none of which are about this non-existant issue.
I've never had any issues with bounce. Obviously I can't comment on anyone else's system, but I would never re-listen to anything that I've bounced, because there's never been any problem.

D

[funkcity]

It took Digi 10 years to get Destructive Punch right!

The reason we the Film Post folks asked for it was that traditionally we recorded on 35MM Mag film or 24 track tape or DA-88s etc..
When you were done you just delivered the medium... It was the master. The source elements were still available and console automation saved the mixes.

The next step in this film-style recording was the era of the non-linear record-to-hard-drive "Digital Dubber"
The Akai DD8, Tascam/TimeLine MMR8 and the Fairlight MFX did there best to emulate the tape and film machines.

But as time marched on ProTools became even more embedded as the de-facto standard. And the Macs got faster. And the networking became easier, and storage got much cheaper.....

Tape Mode on the MMR8 essentially set the stage for PT Destructive Punch.
However.. It took Digi 10 years to get Destructive Punch right and of couurse when they copied the feature they actually did it better! (sound familiar?)
You actually write zeros to the drive first by DPEing them (Destructive Punch Enabling)the length of the project with the correct amount of tracks selected.
You then fill those zeros with new data when you record... and its just like tape. And you can deliver this directly for DCinema use, Dolby, DTS, Sony mastering, archival etc... with no rendering or bouncing required!

Also.. For You Live recording folks.
DESTRUCTIVE PUNCH IS EXACTLY WHAT YOU NEED WHEN RECORDING LIVE!
Because you've actually allocated physical sectors on the drive, you can pull the power cord and lose no data! I've done it!

One day we tried another beta of PT DP and it worked!
Also the Intel Macs made their arrival and CommandSoft/FiberJet figured a way to make a RAID array SAN look like a local drive on ProTools and the dubbers went away and we record straight to a SAN with Destructive Punch... We’ve been doing this now for the past 2 years now and it’s been flawless!

Our Scoring stage and the ADR Stages do record non-destructively as we do NOT want to lose live performances or alternate takes.

[MattiasNYC]

some interesting info there..... thanks.

[psvennevig]

Lex?

Why is there a problem.

Maybe I'm stupid, but whats the problem of selecting the bits an pieces in a stem (or multiple stems) after a lengthy dubbing session and choose "audio bounce selection" ??
It works, nor "chips and stuff" and it takes under 2 seconds.
AND compared to PT you can adjust the x-fades if you want before you go.

Am I missing something?

/pål

[LEX]

Lex?

Why is there a problem.

Maybe I'm stupid, but whats the problem of selecting the bits an pieces in a stem (or multiple stems) after a lengthy dubbing session and choose "audio bounce selection" ??
It works, nor "chips and stuff" and it takes under 2 seconds.
AND compared to PT you can adjust the x-fades if you want before you go.

Am I missing something?

/pål

Read here:
http://cubendo.com/showpost.php?p=18518&postcount=50

I should also point out, this could possibly lead to disk fragmentation too. Lets say you record 48 tracks worth of stems, do a days worth of punching in and out, bounce, get rid of old files and you continue like that for the next 3 days.
By the end, the 48 tracks are closer to the inside of the disk than the outside, and will eventually slow down the drive.

I don't understand everyone's reluctance toward destructive record. There are alot of advantages to having it.

Why not give us both? They both have their purpose and function as funkcity pointed out.
But by SB methodology, it is "their way or no way".


LEX

[funkcity]

to be fair...

I use Nuendo at home and am setting up a Nuendo studio for my friend and a Cubase studio for my brother. Original music recording and mixing is the drill for these rooms.

The major Film Studios in Hollywood (Fox, Universal, Sony, ToddAO, Disney, WB) all have huge investments in their mix stages with Harrison, Euphonix and AMS consoles. (And ICON rooms or ICONs used as sidecar consoles are used increasingly) Lots of storage and lots of ProTools 64~96 track players and recorders. These are the studios that are all primarily running destructively on their mix/dub stages.

The movie we are working on right now has delivery requirements for Sony SDDS, DTS, Dolby SRD, and Dolby Stereo Lt-Rt, IMAX, and DCinema. Some are delivered for encoding at: at 44.056, 47.952, 48K, etc...

We run all these at the same time with different PT recorders in Destructive, back them up and electronically "Smart Jog" them to anywhere in the world.

This is a major time saver.

Could Nuendo do this? Absolutely, but The Steiny folks need to want this business enough to to take the next step!
Now that Martin and Lars are gone it has yet to be proven that they want this type of post business

I can only imagine the meetings at Yamaha in Orange County with the Japanese, Americans and the Germans! :icon_biggrin: Can this enormous talent be harnessed?

[psvennevig]

Read here:
http://cubendo.com/showpost.php?p=18518&postcount=50

I should also point out, this could possibly lead to disk fragmentation too. Lets say you record 48 tracks worth of stems, do a days worth of punching in and out, bounce, get rid of old files and you continue like that for the next 3 days.
By the end, the 48 tracks are closer to the inside of the disk than the outside, and will eventually slow down the drive.

I don't understand everyone's reluctance toward destructive record. There are alot of advantages to having it.

Why not give us both? They both have their purpose and function as funkcity pointed out.
But by SB methodology, it is "their way or no way".


LEX

Of course we should have both.

I just wondered if its a problem today? And I don't think its a problem with file fragmentation.
Not at all compared to huge multitrack recording at least.

Put 4 pc. 256MB 64MB cache SSDs in a RAID0 :-)
Should get you running.

P

[Daryl]

Well I spoke too soon. I'm trying to bounce 4 stereo tracks, each with 4 edits in them. So far it took 9 minutes just to get the Replace Events dialogue. I wonder how long until it is done?

I think that this is buggy in the same way as the export is, and I think it is to do with having the tempo track enabled. Obviously Post doesn't necessarily need a tempo track, but for the majority of the MIDI users it is essential. How can this be so hard to achieve?

So it has now started bouncing......................:sleeping:

So it took 29 minutes, the irony being that it would almost have been quicker to record each track again to another audio track.

D

[psvennevig]

Major drag and bug, I agree.

[Daryl]

OK, so I did a little test. I switched the Tempo track off. Bounce started immediately. Hmmmmmmmmmmmmm. Why is bounce affected by enabling the Tempo track? Does bouncing take tempo related stuff into account? Time for some investigation.

D

[The Guru]

This has been around since v.4. There was a thread at cubase.net that dealt with the export/bounce times with a Tempo Track. IIRC, the more tempo changes and time signature changes, the longer the wait.

Right now, the only two real options for stems is MEAP or setting up not-connected ports in VST Connections and doing it realtime.

[Daryl]

This has been around since v.4. There was a thread at cubase.net that dealt with the export/bounce times with a Tempo Track. IIRC, the more tempo changes and time signature changes, the longer the wait.

I knew about the export Mixdown problem, but I didn't know that it affected Bounce as well.

D

[The Guru]

Haven't gone through it, but here's the original thread with a repro.

http://forum.cubase.net/phpbb2/viewtopic.php?t=92604&start=0&postdays=0&postorder=asc&highlight=

[LEX]

:pop_corn: :eusa_think:

LEX

[Daryl]

Haven't gone through it, but here's the original thread with a repro.

http://forum.cubase.net/phpbb2/viewtopic.php?t=92604&start=0&postdays=0&postorder=asc&highlight=
Yes, this thread is about Mixdown, not Bounce. Mixdown is not a problem, because I always use real time any way, so I could record to another audio track, if I wanted to. However, Bounce is supposed to be quick, just like PT consolidate. Only it's not......!

D

[The Guru]

As you can see by the dates, it was a while ago. I could have sworn I saw the bounce issue discussed.:eusa_eh: Maybe I just saw it on my system and linked it to the mixdown issue. :eusa_think:

How's about posting a step by step repro in a new thread so we can confirm it/ see if it's still present in v.5?

[MattiasNYC]

Well I spoke too soon. I'm trying to bounce 4 stereo tracks, each with 4 edits in them. So far it took 9 minutes just to get the Replace Events dialogue. I wonder how long until it is done?

I think that this is buggy in the same way as the export is, and I think it is to do with having the tempo track enabled. Obviously Post doesn't necessarily need a tempo track, but for the majority of the MIDI users it is essential. How can this be so hard to achieve?

So it has now started bouncing......................:sleeping:

So it took 29 minutes, the irony being that it would almost have been quicker to record each track again to another audio track.

D

Can you give some more details on the process please? Length of tracks etc...

[MattiasNYC]

Just tested bounce selection from a full length movie on the Dialog track with way over 100 edits over 40 minutes. Finishes in about 8-10 seconds with a replaced events + waveform reconstruction.

Not sure what the problem was in your session....

[MattiasNYC]

Just took a piece of same dialog track and recorded back into Nuendo. Then took same track and did faster-than-realtime audio mixdown. Then phase reversed one of them.

0.........

So far I'm not seeing the problems others are mentioned.....

[Daryl]

Just tested bounce selection from a full length movie on the Dialog track with way over 100 edits over 40 minutes. Finishes in about 8-10 seconds with a replaced events + waveform reconstruction.

Not sure what the problem was in your session....
How many time sig changes? How many tempo changes? I would guess hardly any. :wink:

D

[Daryl]

As you can see by the dates, it was a while ago. I could have sworn I saw the bounce issue discussed.:eusa_eh: Maybe I just saw it on my system and linked it to the mixdown issue. :eusa_think:

How's about posting a step by step repro in a new thread so we can confirm it/ see if it's still present in v.5?
As soon as i have time, I'll try deleting stuff in my tempo track, to see how much has to go before it works as expected. No time at the moment though, as I'm preparing for sessions.

D

[Daryl]

I've just done a few unscientific tests. It seems that the delay on Bounce starting is directly related to the number of tempo change points in the tempo track. With only one point, it is instantaneous, but as soon as it gets up to 4 or 5 the delay goes up to 15 seconds. My Tempo tracks probably have over a thousand tempo points, so this is why it took 9 minutes to start. As to why it took another 20 minutes to Bounce, I really don't have the time to investigate further. I think I may have to sort this project out in N3.

I will also investigate whether or not it is fixed in C5, because of not, then that will be yet another reason not to upgrade to N5.

D

[The Guru]

Yeah, add a few time signature changes and it will get even worse. Bounce and non-realtime mixdown are cousins if not twins, IMO.

[Daryl]

Yeah, add a few time signature changes and it will get even worse. Bounce and non-realtime mixdown are cousins if not twins, IMO.

Time sigs don't seem to make so much of a difference, if any. I have about 25 time sig changes, and it is still instant Bounce, as long as there is no more than 1 tempo point in the project. I still don't understand why Bounce takes anything tempo related into account. It's only glue.......

However, do remember that Nuendo 4.3 is a newer version than any Cubase 4 version, so things might have changed.

D

[MattiasNYC]

How many time sig changes? How many tempo changes? I would guess hardly any. :wink:

D

No, you're right, neither.

I was talking about this issue from the standpoint of destructive record as it's done in post-production, in which case time/tempo changes aren't relevant really.

[MattiasNYC]

Time sigs don't seem to make so much of a difference, if any. I have about 25 time sig changes, and it is still instant Bounce, as long as there is no more than 1 tempo point in the project. I still don't understand why Bounce takes anything tempo related into account. It's only glue.......

However, do remember that Nuendo 4.3 is a newer version than any Cubase 4 version, so things might have changed.

D

Very strange indeed...

[TAFKAT]

Hey Daryl,

Interesting stuff, and it amazes me how it flew under the radar for so long.

Please get back to us re C5.

:pop_corn: