DVX audio problems - That Synching Feeling

John Bertram

New member
Using: FCP 5.0.2 on a G5 dual 1.8 under OS 10.4.2
Experience level: tons of analog editing; still getting familiar with NLE
Project material: mostly 24p DV shot on DVX-100 & DVX-100a

Issue:
Am having a very frustrating problem with out-of-sync footage from a DVX100 and Final Cut Pro 5.0.2.


After learning that this was a "known issue", I located, installed and began to use the "Panasonic DVX100 Audio Sync Tool" available from the Extras folder on my FCP install disk, which adds an easy-to-use Offset Audio Sync command to my Tools menu (the default setting for which, at +2 frames, is usually just right to deal with this bizarre business of the DVX100 sometimes delivering audio that's 2 frames ahead of the video).

But here's the new problem:

If I look at that now-in-sync clip in the Viewer, and simply add an In Point (as a sync mark for the creation of a mutliclip, or just to begin using the clip in a Sequence timeline), the audio offset suddenly disappears, and I have to re-apply it from the Tools menu. But then if I move or re-set the In Point, the audio offset disappears again -- and again I have to re-apply it!

This is getting crazy: so now whenever I set an In Point for a clip whose footage was originated on the DVX100 (as opposed to the 100a, which was our "B cam"), I have to first re-apply the Audio Offset before I add that clip to a sequence? And re-apply it again every time I happen to change that clip's In Point?

I notice in the Apple Read Me file for this plugin, it says: "When you modify clips using the Offset Audio Sync command, the corresponding media files are unaffected. Only the clips in your current project are adjusted."

But what if I WANT the media files to be adjusted, so that the corrected sync remains locked in?

ARRGHHHH.

So now, the questions become:

- Why is this Offset Audio Sync so tentative that merely assigning or changing an In Point "Undoes" the command?

- Could there be something in how I'm applying it (though I don't see what) that's causing this?

And if the answers to the above are: "Who knows?" and "I doubt it", then do I need to adopt the following workflow for each of the 7 or 8 "A cam" DV tapes still remaining to be captured:

1) Capture everything useful from the tape (either by logging first and Batch Capturing, or by Capturing Now and then using "DV Start/Stop Detect", "Make Subclip", and logging in the Browser);
2) drag all those new clips into a blank Timeline;
3) unlink the audio tracks from the video;
4) shift all the audio tracks two frames to the right;
5) relink the audio tracks with the video;
6) use the Blade All tool to even off the beginnings and ends of each clip;
7) then drag all the now in-sync clips back over to the Browser, where I make these newly-added clips into Master Clips, and finally delete the original out-of-sync Master Clips altogether?

A test I just did on one clip treated this way confirms that it holds on to its new sync no matter how many In Points I proceed to enter and change in the Viewer -- which makes sense, since the above workflow bypasses this fragile "Offset Audio Synch" tool altogether.


So, despite my initial relief at discovering the "Panasonic DVX100 Audio Sync Tool" plugin, it now looks like I'll just have to ignore it and create whole new, sync-adjusted clips via the Timeline.

Unless one of the experienced denizens of this list can offer a simpler solution!

At this point, the floor is open for all nominations.
 
I've not had a problem in 4.5. Log and then batch capture (avoid Capture now)
Select clips in the BIN.
Apply the sync offset.
Now move the clips into the timeline and edit normally. Sync should remain.
If you want the media to be in sync for other projects or as stand alone files you'll need to export the clips.

If that's not working make sure you have the latest updates for everything. Try creating a test pproject with a couple of clips. It's possible your project is corrupt and it's no remembering the offset.
Save often.
 
Scott:

Thanks for the reply, but it looks like a simple solution is still eluding me...


[I've not had a problem in 4.5. Log and then batch capture (avoid Capture now)
Select clips in the BIN. Apply the sync offset. Now move the clips into the timeline and edit normally. Sync should remain.]

Have now tried capturing the DVX100 material (our "A cam") both ways -- Capture Now and more recently the full Log and Batch Capture route, but the end result is the same: I apply the Sync Offset to selected clips in the Bin, but the moment I attempt to add (or change) an In point in the Viewer, the Offset disappears.



[If you want the media to be in sync for other projects or as stand alone files you'll need to export the clips.]

Just tried that as a test. Now have to decide (and would appreciate any input or suggestions) which route is more efficient and/or desirable:

1) Exporting the Sync Offset clips, trashing the original media files and replacing them with the new identically-named exports, then having to reconnect all my Browser clips to those exported QT movie files (which all seemed to work fine in the test I just did);
OR
2) dragging the plain, freshly-captured clips from the Browser right into an empty Timeline, manually resetting the sync and relinking the tracks, then dragging them all back into the Browser (which turns them into new clips), after which I delete the original (out-of-sync) clips, and the new ones magically become Masters, still connected to the original Media files, but able to retain their new sync alignment no matter what In points I enter or change in the Viewer.

Workflow #1 gives me media files where the video and audio are truly and permanently in sync, BUT in my test just now, an 85 MB QT Movie became a 79MB Final Cut Pro Movie File, and a 63 MB QT Movie became a 57 MB Final Cut Pro Movie File. I made sure the Export was to a plain QuickTime Movie, using the same settings, and with no frame recompression (as per the Manual's recommendations). So what's with this 8-9% reduction in file size -- I thought the Export was supposed to come out as basically the identical QT file.

And if I use these new Exported files as my project's official Media files, am I risking any possible reduction in picture quality (like when you re-open and re-save a jpg file)? And if not, where did those extra MBs go when I did the Export with no recompression selected?



[If that's not working make sure you have the latest updates for everything. Try creating a test project with a couple of clips. It's possible your project is corrupt and it's not remembering the offset.]

On your suggestion, I tried that as well. Unfortunately the freshly-captured clips in the brand new project behaved exactly the same way.


[Save often.]

Always do. Even tried doing a Save, Quit, and Re-Launch immediately after applying the Offset Audio Sync command (to see if it would somehow hold better), but no luck.


So again, I'm hoping someone can suggest the best way for me to proceed:

- leave the captured Media files alone, and just fix the synch a reel's worth of clips at a time in the Timeline, creating new Master clips by then dragging everything back into the Browser and deleting the original clips;
OR
- just apply the Offset Audio Sync tool, and Export those sync-corrected clips into new QT media files, which become the permanent replacements for the original media files (and to which the Browser clips get reconnected)

But as I asked above, are those new, slightly smaller QT movie files really the same quality as the originally-captured ones?


(Thanks for reading -- my brain hurts!)
 
Not sure why they'd be different except if you did a pulldown removal the exported/saved clips woudl be at 23.98 fps, not 29.97 so that could certainly reduce the size. Try to do an manual export withDV codec with max quality settings as a comparison.

If you haven't already edited a number of scenes I'd just consider creating a new project and import the new clips.
 
Haven't done any Timeline/Sequence editing per se as yet, so anything goes at this point.

Defintely didn't select any pulldown removals for the export. Just chose Export as QuickTime Movie, Current Settings, include Audio & Video, Make Movie Self-Contained, and did NOT select "Recompress All Frames".

Would like to go the route of having properly synched media files (as opposed to sync-adjusted clips), but want to make sure that I'm not losing any quality by doing the Export/Reconnect thing.

What I really need is some way to offset the audio sync (for a given reel of tape) by a specific amount during the Capture process. That would be the ideal solution.
 
Did you do any pulldown removal on capture? (24pa?)
You can check quality difference by layering the two clisp and turning visibility on/off of the top one to toggle between them.
You can also set the comp mode to difference, followed by a level filter then it will show you the true difference between images.
You can't do audio offset during capture as far as I'm aware. For some reason your sync's not working. Check the Apple discussion boards since I suspect it's working for most people.
 
Scott--

No pulldown removals at any stage (at least not intentionally), since some of the material for this project was deliberately shot at the regular 60i for an intentional "video look", so I'll be editing everything on a standard 29.97 Timeline.

Meantime, tried your layering suggestion, and could not discern any difference between the two clips (representing the originally-captured media file vs. the "2nd generation" Exported version), at least not on my 14" NTSC monitor, or in my Canvas window with the image blown up to 800%.

Also tried the Composite Mode / Difference test -- the screen goes pure black, and if I'm reading the Levels filter correctly there's no difference between the images. (BTW, you get quite a cool effect when you have two layers of identical video in that Composite Mode / Difference setting, and you offset one of the layers by even just a single frame.)

Wasn't exactly sure how to properly do that "manual export withDV codec with max quality settings as a comparison" test you suggested. Which exact settings would I select?

In any case, I'm now leaning to creating properly synched, stand-alone media files -- but am still nervous about whether I'd be losing something by going that route, given the mystery of these 8-9% smaller files.

(And yes, I have been posting to other Apple/FCP discussion boards, and am still hopeful someone will have found a better workaround for this whole problem. Meantime though, thanks again for all your troubleshooting suggestions!)
 
Scottdvx100 said:
Try to do an manual export withDV codec with max quality settings as a comparison.


FYI: Just tried a couple more tests. Did an Export to QT Movie manually selecting the "DV NTSC 48 kHz" setting, and another Export manually selecting the "Custom..." setting (whose default was already at Max Quality). Both resulted in files exactly the same size as the Exports where I left it at "Current Settings" -- i.e., 8-9% smaller than the originally-captured QT media files to which the Browser clips were connected.

Interestingly though, when I "Re-Exported" a clip connected to one of those new, slightly smaller QT media files, the newly-created QT file was exactly the same size. So the "shrinkage" only appears to happen on the initial Export, when the originally-captured media file gets copied over (via a clip Export) to a new (sync-adjusted) QT file. Any subsequent Exports of a clip connected to that new sync-adjusted file (using the same default settings) seem to produce a file of exactly the same size.

I'd just still love to know what's causing that initial reduction, and if it's anything I need to be concerned about or even aware of.
 
FCP - QuickTime vs. QuickTime? (was DVX audio probs)

FCP - QuickTime vs. QuickTime? (was DVX audio probs)

John Bertram said:
[...] Both resulted in files exactly the same size as the Exports where I left it at "Current Settings" -- i.e., 8-9% smaller than the originally-captured QT media files to which the Browser clips were connected. [...]

Okay, here's the latest wrinkle, for anyone interested:

I just did a couple more of those clip Exports, wherein the clip connected to a QT Movie File (the media file captured by FCP) gets exported and saved as a Final Cut Pro Movie File (which FCP saves in QuickTime format, but with a special FCP/QT icon, instead of the plain QT icon which the captured media files display).

This time I tried using a larger clip, a 1:40 scene whose media file was 369.2MB. The exported version came in at 361.6MB, a "loss" of 7.6 MB. That's pretty much the same as the 6-or-so MB loss that the earlier, and much smaller, 85 and 63 MB media files experienced, but as a percentage it works out to only about 2%.

I then tried Exporting the clip for the smallest clip I had, which was 53MB. That export came in at 47.2MB, a "loss" of 5.8MB, which would technically be over 10% of the original's size.

So, does this mean that FCP (just for its own, inscrutable reasons) simply needs 6-8 MB more disk space, per file, to save and keep track of a captured QuickTime Movie File than it does to save and keep track of the same information as a Final Cut Pro Movie File (its own, more "native" flavor of QuickTime)? But the visual information is all the same (given the no-recompression Export process I've described)?

That's what I'm now suspecting -- and hoping -- 'cause it would allow me to proceed with a relatively easy way of creating new, sync-corrected media files where needed.

Anybody out there an authority on how the different flavors of QuickTime files get handled by FCP?
 
Back
Top