production sound workflow

hi!

how do you guys handle multitrack production sound? do you insist on getting a mono/stereo mixdown track and get rid of the other tracks or do you keep the 3,4 or 5 tracks? I think FCP is a bit uncomfortable if you have to handle more than just a stereo audio track per take (trimming, moving etc). how is the procedure then to let the sound editor get all of his audio tracks in sync?
should he use a autoconformer like TITAN?


thanks a lot,
markuse

FCP should be fairly comfortable:

as long as all the tracks are linked, the clip behaves as one.


nick

hey nick,

of course you can link all the audio files but you also want to link audio to picture. so then if you wanna trim the audio different than picture to make sound for example overlapping the previous scene which is just an ordinary edit action you have to unlink the clips and then it's necessary to trim EACH audio clip. that's what I think is uncomfortable.

markuse

> of course you can link all the audio files but you also want to link audio to
> picture. if you wanna trim the audio different than picture to make sound
> for example overlapping the previous scene which is just an ordinary edit
> action you have to unlink the clips and then it's necessary to trim EACH
> audio clip.

Not necessarily. You can leave video unlinked from multiple tracks of audio, no problem. That's how I approach editing -- though Nick would frown upon that because he prefers to have video linked to audio. I always unlink all video from the audio while leaving all synced audio tracks linked to speed up dialogue editing. You can also leave video and audio linked while using Linked Selection (SHIFT-L) to toggle linking temporarily, or override link for single operations by holding down OPTION before dragging the edge of a clip.

derek wrote:

"leave video unlinked from multiple tracks of audio, no problem. That's how I approach editing -- though Nick would frown upon that because he prefers to have video linked to audio."

i frown on that because you're just making more work for yourself

you have to unlink every clip you bring into the timeline.
that's 100's, possibly thousands of actions, and almost impossible to revert to a linked state
if you just turn linking off in the timeline, that's one action,
and one other action to revert.


mark wrote:
"if you wanna trim the audio different than picture.... it's necessary to trim EACH audio clip"

i see what you mean:
linked or merged multitrack audio DOESN'T behave like stereo audio.
you DO have to select each track
that IS a bad state.
very bad design.

only time i'd cut multitrack before (more than stereo) it was audio only, so this wasn't an issue

here's how i;d deal with it:
select an edit, so that ALL tracks are selected.
hold CONTROL+OPTION+APPLE and select the picture edit.
this will de-select the picture.

so it's TWO clicks (select all, de-select picture)
rather than 4 clicks, say, for 4 track audio

AOTHER WAY
would be to use the GROUP tool, but it has it's own problems
that's G (for Group)
then hold OPTION and lasso your audio edits
problem is the trim window will open, and your playhead will jump to the selected edit point
this could be a good thing, but as there's no way to make the trim window NOT open, i think it;s kinda bad.

YET ANOTHER WAY
would be to lock the picture tracks,
but this will only work if you are doing a roll edit on the audio.
a ripple edit will put all subsequent clips out of sync.

with multitrack audio becoming more prevalent,
FCP needs to address this.

the other way is to work with stereo mixdowns,
but it does become difficult to conform back to the multitrack version.
there WAS a product from Gallery software that would help with this.
[www.gallery.co.uk]
i cant remember the name, it was used by the Murch team on Cold Mountain, & possibly Jarhead , but it's not available anymore

cheers,
nick

thank you very much derek and nick I'm gonna check your tips.

i think you mean gallery metaflow.
but there is another software: synchro arts TITAN.

Titan does an auto conformation by reading out the filenames and timestamps off the audio files (via reading an EDL) and resyncs matching audiofiles of a folder of choice. the result is a pro tools session containing all the multitrack audio. well it's a little more difficult but that's the idea.
unfortunately it's not that cheap.

it's important to export audiofiles that still include the production timestamp and you have to be careful with the clipnames. so the best would be to export a referenced omf but that's the problem with FCP. If you export an XML and the audio software is Pro Tools you get problems because ProTools is not capable of reading XML files. You have to use XML-Pro (by www.gallery.co.uk) to convert the XML to a ProTools session.

I also tried the following:
I made a subclip containing the video clip and all the 4 audio clips. when dragging it into the timeline again the subclip showed up as one video clip and two audioclips, linked. but actually all four audio tracks were working. when I open the subclip by double clicking on it I find all four audiotracks. that's pretty comfortable because you can edit all 4 audio clips by just handling a stereo pair. i hope you can follow. so that seemed to be fine...

but:
because I need to export the sequence unembedded (referenced) anyway I tried to export an XML to Logic. But the problem is that the subclip is somehow not transfered via XML. It is transfered via FCP OMF. Do you know why?

cheers,
markuse

yeah, metaflow was what i couldn't remember, thanks.



logic:
oh, i saw that other post re: logic,
but i didn't have an answer.

for the film i;m working on now,
we tried to work with logic.
it wasn't really supposed to be my job to get it into logic
but the sound guy who was supposed to be doping it, didn't know much about film,
and knew nothing about the FCP / Logic interface.
OMF seemed to work,
but then the sound guy said he wanted the volume graphing,
so we tried XML,
and there were problems with that.
i cant remember the details, sorry,
but they were enough to make me dump logic as a solution,
and in the end we went with a different approach altogether.

so i cant advise on that,
other to say logic doesn't really seem quite there yet for video post production.
(although it could have been a "personnel problem"

we DID have subclips in our show,
but not many.
maybe none in the first spool, which is all we tried before giving up.

how's that for helpful...not!

try the creative cow Logic forum:
[forums.creativecow.net]

good luck,
nick

>>how do you guys handle multitrack production sound?<<

Record .bwf audio. All your tracks are in one file with original TC.

Clay

yes poly BWAVs are a very nice thing. all organizing is much easier. I didn't try it with FCP so far. Hopefuly the day will come when the conversion isn't necessary anymore. I mean it can't be such a big deal to correctly read BWAVs. I heard of sound guys having problems with .mov audio files while using auto conformer software. well there is always something...


cheers,
markuse

Clay,
do you know how the poly BWAVs behave in FCP?

i created a poly QT file, and in FCP it behaves as described before:
you need to click on every track when selecting to trim.
(unlike a stereo pair, with the green arrows, where you only have to click on either side for both to be selected)

cheers,
nick

Hi Nick,

.bwf poly files (once converted to QT) behave like a group of mono tracks. Let's say you had a set of 4 tracks in the timeline. You can link them all to the appropriate video track, or create 2 sets of stereo pairs and link those to video as well. Select and drag either end to trim the whole lot. You can also drag the linked selection from the timeline back to the browser to make subclips and retain all tracks, linkinkg setups and tc's. Exporting via Batch as usual with in and outs selected works too.

To trim a poly audio selection: What would be highly useful is a Group command to do what you're talking about (put that way up on the upgrade wishlist). Only workaround I know is to make stereo pairs (as above), which unfortunately does nothing more than save a few mouseclicks.

Maybe interesting general note: I recently heard about a .bwf workflow where they piggybacked 3 4-track machines and generated one 12 track file for 5.1 mix. Very cool. And then of course there's DEVA. Interesting read over at zaxcom.com if you're not already familiar with it. It's FCP compatible via the QT conversion. With multitrack coming at us rather quickly now, FCP needs full .bwf support. (end of sermon)

btw >>i created a poly QT file<< How exactly? Could you elaborate?


Clay

"btw >>i created a poly QT file<< How exactly? Could you elaborate?"

i used two methods, using some files for 5.1 music tracks from a film a did a while back.

1.
laid the tracks up in FCP,
set teach track to have a discrete audio output,
and exported.

2.
opened all the files in QT,
and did a copy/paste...
sort of.

i picked one as the "master"
select another,
apple a to select all,
apple c to copy.
over to the "master",
Edit Menu/"Add to Movie" or Option Apple V

did that 5 times, then did a "save as"

you're not wrong about bwf support for FCP.
it;s crazy that it doesn't exist right now!

cheers,
nick

ClayC wrote:

> >>how do you guys handle multitrack production sound?<<
>
> Record .bwf audio. All your tracks are in one file with
> original TC.
>
> Clay

Hey Clay,

did you mayby have to export Audio to ProTools at projects you worked with converted BWAVs in FCP? How did you do that exactely?

I just think of the following :
ProTools is not able to link QT files, they must be converted to either (B)WAV, AIFF or SD2. so it wouldn't be possible to share referenced session data, you have to export embedded OMFs. That again would make dialogue editing while picture editing is still in progress very uncomfortable since the sound editor frequently gets audio footage he already has (via the embdded OMF) I dont know how XML Pro behaves in that case.

cheers,
markuse

You can try a 'virtual mixdown' by nesting each audio clip. So you got both: the access to all original tracks and the 2 track mix of it.

To work with "stupid" other edit packages rename the the .mov to the original .bwf or .wav - then use an EDL or just rename in the EDL. This works fine.

Andreas



Post Edited (05-10-06 06:03)

Some workflow tools for FCP [www.spherico.com]
TitleExchange -- juggle titles within FCS, FCPX and many other apps.
[www.spherico.com]

>>did you mayby have to export Audio to ProTools at projects you worked with converted BWAVs in FCP? How did you do that exactely?<<

It's very easy. Once you've finished up your edit, put 2-pops at the heads of all the audio tracks, then select all audio tracks and hit Export> Audio to OMF. Burn the OMF file onto a Data CD and send it out to the recording studio. Works with ProTools just fine.

Clay

Nick,

thanks for that. Get the general idea.

How 'bout this though: Imagine being able to export> quicktime movie> audio only> options> poly (# of tracks, bit depth, sample rate, blah etc.). Maybe Andreas will add that to the next bwf2xml tool upgrade. If one can get the converted bwf audio in, there's gotta be a way back out from the timeline or browser or both, either as QT poly or back into .bwf poly. Then we'd have a mulitrack alternative to OMF that would definetly work in any situation. And, be able drop the QT poly file into Soundtrack for a 5.1 or 7.1 mix.


Clay

> To work with "stupid" other edit packages rename the the .mov
> to the original .bwf or .wav - then use an EDL or just rename
> in the EDL. This works fine.

thanks Andreas!

So I tried renaming the audiofile from .mov to .wav (was a .wav originally).
unfortunately the "stupid" edit software wasn't able to relink it. but I tried it with Sebsky Tools. Sorry

Are you sure that works with BWF2XML?

markuse

Hi Markuse,

Sorry I was a little bit short and somehow wrong in my reply. The trick with the renaming is:
Keep your original BWFs somewhere, then work with the converted QT files in FCP. If the QT files are named .wav or .bwf FCP doesn't care, it just works fine. After you're done with editing export the EDL, which refers just to file names.
The "stupid" edit software will take everything with the names in the EDL as long as it fits to the file format it can read. So just feed it with the original BWFs. The "stupid" edit software has some problems with EDL though, because of the EDL format since EDL (audio) may can't keep that many tracks.

I will update BWF2XML soon to create QT files which keep their current extension, and to create "virtual mixdowns" and maybe "real mixdowns" as well (for the last I got quite a lot of requests from Avid users). Also Titan naming conventions might be included in the next update.

@ Clay:
I know how to convert the QT files in theory, but this can't be done directly from FCP - only thru an additional step. The bad thing is that I'm still a VFX guy, who does know too little about QT programming.
But since some more people really pay for BWF2XML, it looks like I can hire somebody to build a obj-c class

Andreas

> Keep your original BWFs somewhere, then work with the converted
> QT files in FCP. If the QT files are named .wav or .bwf FCP
> doesn't care, it just works fine.

thanks, that makes sense. so after all the question is which application is the stupid one. the one that doesn't read QT files or the one that doesn't notice filename change ...

cheers,
markuse

> thanks, that makes sense. so after all the question is which
> application is the stupid one. the one that doesn't read QT
> files or the one that doesn't notice filename change ...
>
... aaah, good question, I think it's the one which doesn't make use of actual software technologies.
Depending on the direction you're going, it might be be a different one - if Titan is not able to read QT it's stupid, if FCP is not able to read BWF (with the data included) it's stupid as well.
But I tend to say Pro Tools and Titan are 'stupid' a little bit more than FCP, if Pro Tools claims to work with BWFs, why does it kick out out all other metadata beside timecode when converting BWF files - I do not see any sense there.
When QT 5 came out, I had some discussion with Apple about BWF and the ability to read those files, they listened and made it possible to read this kind of file type in a very basic, simple way, though it didn't make the way to the FCP team of Apple - they are still listening and watching, so stay tuned.
That's why maybe sometimes it's only the user that is the 'stupid' one, cause he doesn't see that he is working with a 'stupid' computer and a 'stupid' software, where both of them do exactly what he tells them to do ;-).

Does that make sense?

Andreas

P.S.
I forgot to mention: Often the user is the "clever" one, cause he realizes how "stupid" his software is.



Post Edited (05-11-06 04:22)

Some workflow tools for FCP [www.spherico.com]
TitleExchange -- juggle titles within FCS, FCPX and many other apps.
[www.spherico.com]