PhraseFind ignores ‘Search: Current Bin”

[Frederic Lumiere]

So moving the mxfs created by the AutoSync to the ‘1’ folder didn’t force PhraseFind to index them. Somehow, PF is still ignoring them. So when Autosync is done combining a video and audio clip it creates a subclip and trashes the reference to the masters in MC? I say this because the resulting icon looks like a subclip (different from the other master clips). I wonder why PF is ignoring these…

To make sure, I brought a new AMA file in, transcoded it to DNxHD and right away, I did see the new .pat files created by PhraseFind in the PhoneticData folder and the new clip is indexed properly. So PF definitely has a problem indexing MXFs created by MixDown/AutoSync.

Frederic Lumiere
Producer
Lumiere Media

[Frederic Lumiere]

Something’s up with these subclips… MC won’t let me enter timecode in them to navigate the clip.

Someone else was experiencing a similar issue with group clips:

https://community.avid.com/forums/p/106457/616291.aspx#616291

[Frederic Lumiere]

I also tried to do a new Audio MixDown from the final subclip to see if PhraseFind would index it and when I set an IN & OUT point in the clip and choose to Audio MixDown, I get the error: “Exception: Mark in or Mark Out incorrect. But it lets be place that edit in the sequence if I try…

Frederic Lumiere
Producer
Lumiere Media

[Frederic Lumiere]

[Michael Phillips] “There is a trick to make the the sequence timecode match the source timecode and that is by selecting the source clips and doing “AutoSequence”. That will lay all clips into the timeline where source and record timecode match. If shooting time of day, you will have black gaps where there is no material.
“

This is a great trick Michael! Wow – very cool. Somehow when I was resetting the sequence timecode manually it was slipping over time compared to the clip’s timecode. I assume because of dup TC, missing, etc.. this trick takes care of all these issues. Very nice! Thanks.

[Michael Phillips]

You’re welcome. One of those little known features. That one has been in the product since about 1997 and was designed for audio layback to MOS tape masters when syncing was done in Media Composer but still has value in some file-based workflows.

Anyway, I did look into the audio files being in a folder that is not “1”. I tool all my audio from BWF import and moved them to a folder “7”. I trashed the search .pat files as it didn’t feel a need to re-index. Once that was done, it went through the indexing process. Will a full green index indicator I did a search that just worked yesterday when all the files were in “1”.

Nothing was found.

Sounds like a bug to me. I will fire off an email to engineering QA.

Michael

Michael Phillips

[Frederic Lumiere]

Hey Michael,

So I’m going back to what you originally suggested. Make a QT Reference of the sequence rather than doing a MixDown. The main reason being that the audio Mixdown only give me 1 stereo or 1 mono audio track option. Some of the clips have 3 tracks that I want to preserve.

What I need is the equivalent to producing a “self contained” clip with 1 track of video and as many tracks of audio that are present in the sequence.

Could you take me through the steps of getting that done?

Thanks Michael!

Frederic

Frederic Lumiere
Producer
Lumiere Media

[Frederic Lumiere]

Looks like I may have figured out how to make 1 video clip with all audio tracks which combines all the DNxHD takes from an interview and searchable by PhraseFind without re-transcoding the whole thing.

Here’s how I did it:

1. Select all the clips from a bin in order and choose Bin/AutoSequence

2. Open that sequence, place an IN and OUT point to select it all and choose Special/Video Mixdown

3. Delete the video track from the sequence and File/Export QuickTime Reference

4. Open the QT in QT pro and save as self-contained

5. Bring the Self-contained audio (multiple tracks) with AMA

6. Transcode the audio AMA to Avid codec

7. Select the new audio clip and the Video Mixdown from the bin and choose Bin/AutoSync

The final subclip has video synced with audio and is searchable by PhraseFind but somehow lost its timecode!

Almost there…

Frederic Lumiere
Producer
Lumiere Media

[Frederic Lumiere]

Success! A much simpler way yo accomplish it. Here’s the summary:

1. I used AutoSequence to make a sequence with the same timecodes as the clips. Works great! Now I have one sequence of the whole interview.

2. PhraseFind won’t index sequences so I Mixeddown the Video and each Audio track individually and used AutoSync to make a subclip of the video and synced audio tracks.

3. I verified that the Mixed down audio tracks were indeed being indexed by PhraseFind in the PhoneticData folder which was encouraging.

The subclip still didn’t show in the PhraseFind results when it should’ve so I moved the subclip to a different bin, saved that bin and it looks like it updated the PhraseFind database and now that subclip shows in the PhraseFind results.

The Video and Audio both have the same TC but I still can’t navigate the clip by typing timecode… I’ll live with it for now.

Frederic Lumiere
Producer
Lumiere Media