Any ideas on channel spill for Field Recorder "Functionality"?

[trevora]

I think the 'strange clip name' looks like the camera (i.e. video) clip name. Presumably from the sound/vision having been grouped. The BWF metadata should still be in there though...

[Ray JB]

No metadata is carried thru with the MXF files it seems apart from timecode and Family UMID. I see no channel number or name, only the .A1 etc suffix.

I'm getting a little closer to finding out what has been happening with the changed channel numbering, it seems that the recordist has been recording from track 3-8 on the Aaton Cantar and also another Sound Devices 788T. The original recordings have track number 3-8 but AVID brings those in and re-orders them as 1-6, sheesh. This is despite iXML and BWF metadata to the contrary.

It should not be that hard surely!

[Frank Kruse]

Originally Posted by Ray JB

No metadata is carried thru with the MXF files it seems apart from timecode and Family UMID. I see no channel number or name, only the .A1 etc suffix.

I'm getting a little closer to finding out what has been happening with the changed channel numbering, it seems that the recordist has been recording from track 3-8 on the Aaton Cantar and also another Sound Devices 788T. The original recordings have track number 3-8 but AVID brings those in and re-orders them as 1-6, sheesh. This is despite iXML and BWF metadata to the contrary.

It should not be that hard surely!

Ray,

the Cantar has a function called poly-rotate. Normally the machine's mix-tracks are the bottom ones which is a bit of a pain for the editor because it's always on a different track number depending on how wide it records.

Hence they added the poly-rotate feature which tags the bottom mix tracks as the top-most (rotates them to the top and the top ones further down). But it does this only in the meta-data, not the physical track of the poly-file.

So in fact the AVID is doing the correct thing but it can be confusig when the track names are missing because you go by the numbers and you think what is labeled 3-8 shows up as 1-8.

Makes more sense when you see the track names and the actual number os less significant.

F.


From the Cantar manual:

• PolyRotate2 moves the
tracks two ranks up (the T1–T6 isos go to positions T3–T8, while T7T8 are sent to T1T2), then it makes a poly file (AB1234.PR)
stored in an .AAR folder. PolyRotate2 solves the former Avid MCs inability to put tracks where desired, e.g. when importing a mixdown the MC puts it on the timeline A1-A2, unfortunately it will also put the T1T2 iso tracks of a multi- track recording there too!
Consequently if the editor later decides to import the original T1–T8 tracks, the MC will kill the mix tracks he has been using from day one! Thanks to PolyRotate, the original mix- down tracks (T7T8) rotated to T1T2, will replace the 16-bit poly mixdown with the 24-bit originals, while the native iso tracks (T1–T6), pushed to T3–T8, will gracefully fall into place.
• PolyRotate1 performs the same operation as PolyRotate2 but pushes the tracks by one rank only: T1–T7 go to T2–T8 while T8 (the Xb mixdown) is sent back to T1.

hope this helps... so if your guy doesn't record a mix track there's only T3-8 when poly-rotate is on.

Another explanation as to why the track names get lost could be that in old days the Canter would only write the track name info to the iXML chunk which the AVID MC ignored. But they appeared in PT when reading the originals because PT supported iXML.

Then at some point AVID read iXML so maybe the AVID version they are using is quite old and doesn't read iXML or the AVID MC only forwards whatever is in the bext-chuck to the MXF media essentially stripping the iXML info on import.

Many reasons....but best workaround still:

Don't use avid media in the first place ;-)

[Ray JB]

Hi Frank

The displacement of Track numbers and loss of names in the AVID metadata is kinda weird. Apparently the version of media composer is 8.3.1 and the Aaton Cantar X2 software is version 2.44 (the latest). But there are Sound Devices 788T files here with the same displacement of track numbers. Those SD files also start on track 3-8 and are displaced by AVID Media composer to be tracks 1-6.

Inside each poly file I can see basic BWF info and also the iXML chunk contains the correct track names and numbers. Is it possible that AVID Media Composer is not taking in the iXML data? Screenshots posted below of both an example SD788T BWAV poly and an Aaton Cantar X2 BWAV poly.

So what I'm seeing is the total loss of any track name or number once ingested into AVID Media Composer? Obviously ProTools can search in the original location sound folders and find matches based on the Family UMID but AVID seems to merely stamp the tracks as starting from track 1.

As the editor on this show has cut with lots of iso channels rather than just using a mix track, I get a huge number of tracks to wade through if I expand tracks as Frank suggested. Mark and Frank, you were correct when you suggested I could ignore the AVID media and reconform to Original sound but it's fiddly as when the editor has clip gained a clip down to -Infinity rather than muted it, I lose the clip gain when the channel is swapped, meaning I have to compare my new expanded AAF tracks (with correct files and proper names etc) and see that I've not just fired up a clip that the editor never intended to play!

Too complicated for it's own good.

[mfranken]

Originally Posted by Ray JB

Mark and Frank, you were correct when you suggested I could ignore the AVID media and reconform to Original sound but it's fiddly as when the editor has clip gained a clip down to -Infinity rather than muted it, I lose the clip gain when the channel is swapped, meaning I have to compare my new expanded AAF tracks (with correct files and proper names etc) and see that I've not just fired up a clip that the editor never intended to play!

Yeah that's boring, but I've always ignored the volume and clip gain info from an AAF. I'd rather take everything at unity to make sure gains are set correctly and compare to the guide track to see if there is anything that was not meant to be played.

Regarding conforms/assemblies, I know most productions here in Australia just use an AAF - as most editors still leave all tracks in the edit, but I've been surprised by the amount of interest I've had from the UK and US re EdiLoad as they want to conform to original WAVs, using EdiLoad to clean the EDLs and PT to expand the tracks. This way you can get the cleanest conform/assembly linked to original poly WAVs.

Regards,

Mark

[Ray JB]

That's interesting Mark, do you find that these days with file based workflows that you have to correct sync much using an EDL based assembly method rather than the AAF method (which will bring everything into sync as close as possible to what the Picture editor was dealing with)?

I mean I know how it used to be, we used to get a lot of small offsets in the conformed files as compared to the AAF but I don't see that much these days when using an AAF to proceed with.

[mfranken]

Originally Posted by Ray JB

Do you find that these days with file based workflows that you have to correct sync much using an EDL based assembly method rather than the AAF method (which will bring everything into sync as close as possible to what the Picture editor was dealing with)?

The only problem I'm seeing is the issue where a location recorder doesn't start it's recordings on a frame edge and when the WAVs are loaded into a picture editing system, the WAVs will be moved to start on a frame edge. Then when you expand tracks (using an AAF or EDL as the location reference) the location WAVs are up to a frame out-of-sync because PT understands samples. Actually in this case case I feel the audio in MC is out-of-sync and the audio in PT is correct!

I know this occurs with deva recorders, not an issue with SD recorders. Avid said it was addressed in MC7.04 and MC8 with a pref that reads: 'Subframe Alignment to Broadcast Wav Start Time'

More here:

http://www.pro-tools-expert.com/home...t-of-sync.html

Which has a link to Frank's DUC post (not sure if you've seen this):

http://duc.avid.com/showthread.php?t=345993&page=6


The film on at the moment was recored in Germany and Australia. The Australian recordings all phase nicely, but the German recordings are up to a frame out-of-sync when I expand from AAF or EDL (via EdiLoad conform ref tracks). The German sound sheets don't say what the recorder is (guess I could look at the metadata for clues) and the picture assistant says this audio was loaded into MC before he came on so he isn't sure what the pref was set to during load.

So not sure if this MC pref actually resolves this issue.

Hence for this job I'm using a combination of clips from the AAF (which has incorrect channel names) and conformed tracks.

BTW, regarding the EdiLoad conform, there were 3 events that didn't find any location audio, but not bad for a 2 hour film.


Regards,

Mark

[mfranken]

Originally Posted by mfranken

The German sound sheets don't say what the recorder is (guess I could look at the metadata for clues)

Found they were recorded with a Nagra VI

[Ray JB]

Yes I've seen the offset from zaxcom deva recorders and also the Nagra, is it fixed on later zaxcom machines like the Nomad etc? The Aaton and SD recorders at least close off on the full frame which helps us lots.

I might try a small AVID test where I stamp standard bwav bext info into the file headers, get them imported into MC and then get them to sync those clips and make me an AAF and see what sort of mess we have?

[mfranken]

Originally Posted by Ray JB

I might try a small AVID test where I stamp standard bwav bext info into the file headers, get them imported into MC and then get them to sync those clips and make me an AAF and see what sort of mess we have?

To me if you have access to some of these recorders it would be better to do some short recordings with each machine and load these into MC - with this pref checked and un-checked (if available) to see if it fixes this problem and to see which recorders do what.