Elastic Audio - X-Form - rendered/committed regions Q?

[jeremyroberts]

Is there any benefit (less DSP usage?) to disabling Elastic Audio and committing the regions once X-Form has completed its render?

Is there any benefit to leaving X-Form enabled (regions rendered) vs. disabling the EA properties of the track and letting it write new audio files ?

Or an alternative workflow, drag rendered regions to new track (writing new audio) then hide and make inactive the EA track. ???

Is there a "best" workflow for this? Assume that the X-Form rendered track sounds exactly the way you want it to.

Thx in advance for any guidance.

J

[midnightrambler]

My opinion (FWIW) is to commit as soon as possible, due to the nature of Elastic Audio generally being "a bit dodgy", to put it bluntly. I've rebooted sessions with missing regions, dropped regions, regions not playing as they should, etc etc. I would have thought for long term archiving this would be the most fool-proof as well.

I quite like your idea of keeping the X-Form track and putting it on a new track - but it renders so fast when you do this, compared to "committing" properly, that I can't help wondering if it's not doing something correctly when it does this. I always save a session with "PRE-COMMIT" track so I can import it to the newer session if necessary.

[jeremyroberts]

Please PTI if this is has been discussed for years -- but WHY does committing regions that have ALREADY been rendered by X-Form take an equal amount of time to commit as the initial rendering. If the audio has been rendered, shouldn't it simply be a faster than real-time process to lock it in as audio?

If I have a track with 1 region and X-Form has already rendered, committing now takes JUST AS LONG as the initial render. YIKES!!! What is this about?

Wouldn't it be faster to just bus the track and bounce it over?

What is technically happening during the "commit" process? Is there more going on than meets the eye?

Is a rendered X-Form region really fully rendered, or is only a placeholder?

Confused.

DTS?

[1m1]

I have been wondering exactly the same thing Jeremy. X-Form is great and totally worth the rendering time, but having to do it twice... I have plenty of time to sit and ponder this question.

[meech]

How long is the first render supposed to take? Whenever I've tried it, it just has taking an enormous amount of time and I could never stand to wait that long for it to render.

[Sean Russell]

Originally Posted by midnightrambler

My opinion (FWIW) is to commit as soon as possible, due to the nature of Elastic Audio generally being "a bit dodgy", to put it bluntly. I've rebooted sessions with missing regions, dropped regions, regions not playing as they should, etc etc. I would have thought for long term archiving this would be the most fool-proof as well.

I quite like your idea of keeping the X-Form track and putting it on a new track - but it renders so fast when you do this, compared to "committing" properly, that I can't help wondering if it's not doing something correctly when it does this. I always save a session with "PRE-COMMIT" track so I can import it to the newer session if necessary.

This is how I do it, all the way up to the save-as PRE-COMMIT - I, too, have been bitten by strange re-opens with missing X-Forms, and that kind of thing will only happen to me once, maybe twice before I learn to commit as soon as I can, with plenty of save-as's along the way :)

But I'm confused by some of the later posts.

When I X-Form a set of drums, for instance, it's time for lunch I don't like touching the Mac when the render is happening - no touching the system for at least 20 minutes!

Before I EA, I *always* comp the drums with fades and make sure everything's perfect, then save-as, then consolidate, then wipe the extra drum playlists, then turn on EA, then save-as, then fix, then save as, then X-Form, then commit. No second ridiculous wait time. Lots of breadcrumbs to go back if needed. Usually I don't.

On average, I usually get my drums fixed via close drum mic's with Rhythmic and far / overheads with Polyphonic. When they're right where everybody wants them, I then turn on X-Form. Lunch. Then I commit. It seems to happen a lot quicker.

When I come back from lunch, and I change the X-Form to 'None' or whatever, it only seems to take a few seconds for it to commit the Audio to new files. I get the revert / commit dialog, pick 'commit', hit go, and get setup to track bass. (In my usual, song-per-day mode). It usually fills in the regions pretty quickly.

My typical cutting room CPU is a 2009 Nehalem Quad Core with 8 GB RAM, 8.0.3 cs2, HD-3 and I usually cut 24/88.2k to the internal SATA that shipped with the MP - a 640 GB WD - which I have since wiped and use as a tracking drive. It works great, but I still hate WD in general, especially their chassis / FW/USB drives. But I don't wait twice for the same time for the X-Form render. I seemed to remember waiting longer in the past, but for whatever reason, I don't anymore.

Does my workflow differ from any of you that are experiencing super long wait-times?

[midnightrambler]

Hmm interesting - the main difference I can spot with my workflow is that you consolidate before you turn on EA. I must start doing that and see if it makes any difference. Am I right in saying that you don't do any sort of editing to the consolidated file apart from EA adjustments? I must admit I have a tendency to keep on chopping stuff about even when EA is on, which probably doesn't help the cause.

[Sean Russell]

Originally Posted by midnightrambler

Hmm interesting - the main difference I can spot with my workflow is that you consolidate before you turn on EA. I must start doing that and see if it makes any difference. Am I right in saying that you don't do any sort of editing to the consolidated file apart from EA adjustments? I must admit I have a tendency to keep on chopping stuff about even when EA is on, which probably doesn't help the cause.

Chopping with EA will definitely increase the render time.

I consolidate, EA, fix, commit, then re-chop, when needed. I like to have it all rendered before the producer suddenly wants 4 bars out of the bridge and the outro is now the intro

[jeremyroberts]

But what about the "commit" vs. "render" question?

Is a rendered X-Form region identical to what it will become when committed? If so, let's say I have 6 stereo pairs of rendered x-form tracks that Im want to commit -- am I not better off making 6 tracks, bussing and RTD? (if in fact there is no diff)

IF there is a difference, then what exactly is happening.

If there is no difference, why does it take an equal amount of time to commit?

DTS, please clarify. The documentation is not clear.

Thx.

[midnightrambler]

Originally Posted by jeremyroberts

But what about the "commit" vs. "render" question?

Is a rendered X-Form region identical to what it will become when committed? If so, let's say I have 6 stereo pairs of rendered x-form tracks that Im want to commit -- am I not better off making 6 tracks, bussing and RTD? (if in fact there is no diff)

IF there is a difference, then what exactly is happening.

If there is no difference, why does it take an equal amount of time to commit?

DTS, please clarify. The documentation is not clear.

Thx.

+1

All I know is, if you do both, the files will not cancel when you invert one - so there's something going on...