-
PhraseFind ignores ‘Search: Current Bin”
Does PhraseFind search in the whole project regardless of what is selected in the search dropbox? I selected ‘Current Bin’ but it gives me results for the whole project.
Also, it would be awesome if phrase find could only search a sequence… to know where someone says something in a sequence.
Frederic Lumiere
Producer
Lumiere Media -
28 Replies
-
PhraseFind always searches its entire database and does not yet search sequences. You can do a bin filtering after the fact as an additional step, but I agree it would be nice to have the filter at the front end of the search process.
Michael
Michael Phillips
-
Thanks Michael,
Actually I just discovered that the filter (not the Search Current Bin) does work at the front end of the search. Since my naming convention is pretty good, I can search for the name of the character in addition to the PhraseFind search.
Do you know if there is a way to output a sequence as a ‘self contained’ DnxHd clip without re-transcoding? I think I’d like to output the 3 hour interviews as one clip for finding purposes…
Frederic Lumiere
Producer
Lumiere Media -
Which filter are you referring to? I’m pretty sure the initial search is an either/or – PhraseFind or bin based metadata. But maybe something has changed, I will need to look again. I haven’t had occasion lately to use it. Search has always been PhraseFind for all results, then use the additional filters to drill down to metadata in a specific column within a bin and then sort by bin, etc.
As far as making sequences into sources, there are a couple ways, but I would have to ask from what your sequences were made? I am going to assume you synced in the timeline? If so, take each event (if sequence is made of multiple events) and create sub-sequences where each sequences only contains V1 and audio tracks and are of all the same duration. Highlight these sequences then select AutoSync from the Bin Menu and they will turn into subclips.
Or… export your rendered sequence (if needed) as a QT reference, and bring it back into your project as a new source. AMA linked files will get indexed by PhraseFind but if sequence consisted of several clips, all bets are off for source metadata, EDLs, etc. if used as source in a new sequence. All you will have if filename of QT and whatever timecode the original sequence was at.
Michael
Michael Phillips
-
[Michael Phillips] “Which filter are you referring to?
Any really. I select Name in the ‘Column’ drop box – enter a specific name and the results will return only hits with that name. Things probably have changed…
[Michael Phillips] “As far as making sequences into sources…export your rendered sequence (if needed) as a QT reference, and bring it back into your project as a new source….all bets are off for source metadata, EDLs, etc. if used as source in a new sequence. All you will have if filename of QT and whatever timecode the original sequence was at. “
I bring in footage already synced audio-wise with DualEyes, so the sequence is multiple clips comprising the whole interview. All the clips are now DnxHD175 X (the result of transcoding AMA’s H.264).
So when I create a QT reference, it exports the audio as AIFF and creates a 1Mb QT Reference file, which looks fine. When I bring it back in MC, it looks like it re-encodes the whole thing rather than fast importing like I would expect since it’s already DnxHD 175 X. Is there a way to bring it back without re-encoding?
Thanks for all your help Michael BTW!
-
I guess doing a video mix down is the way to go then?
Frederic Lumiere
Producer
Lumiere Media -
I don’t think I mentioned a video mixdown… that is a third option, but also will loose all links to any metadata if used as source for editing. In order of preference IMHO:
1. Ensure sequence has only one V track (V1) and whatever number of A tracks associated with it. Highlight sequence in bin, select AutoSync and you will now have a source subclip that retains all properties from its sources. You can do this on a duplicate of a sequence as the original sequence will disappear as it becomes a subclip.
2. Export each single event sequence as a QT reference, “Fast Import” or AMA link to it. Only metadata now will be whatever the sequence timecode was and filename of QT (which will be the sequence name)
3. Mixdown. No metadata.
Also, there have been no changes to the “Find” or PhraseFind. Basically you enter a search term, get a set of results, then apply the same search term again and select column of choice to use as a filter. A better approach would be to have the filters up front eliminating the need to re-enter the search term again for those who are well organized. That is what I am referring to. You get the same results, just fewer steps and a bit more intuitive.
Michael
Michael Phillips
-
1. It won’t let me AutoSync because I have more than 1 clip in the sequence. I have dozens of takes for the interview.
2. When I bring back the QT reference I exported of the sequence with dozens of clips, isn’t doing “Fast Import” even though the clips were DNxHD.
3. My only solution is Mixdown since it doesn’t re-transcode the clips but simply copies the data.
Filtering before I do the PhraseFind search works great for me. I only get results based on the phrase criteria + the filters. But then again, based on the Found & after filtering numbers at the bottom right of the window, it looks like it applied the search to the whole database anyway, even though I don’t see all the hits.
-
You need to mark each clip in the timeline (Mark Clip) and subclip that out to a bin so each sequence is only one event. From there you will be able to AutoSync as a new subclip source. Should be fairly quick to do. Park in the middle, select tracks, mark clip, option click in middle of record monitor and drag to bin. A new sequence is created.
Michael
Michael Phillips
-
Also, I see what you are seeing with PhraseFind, that once a filter has been applied, all subsequent PhraseFind searches will be filtered. I am referring to the first time a search is initiated, that it takes an extra step in re-entering the search term that was initially set in many cases to get to subfiltering stage. For example, If in my BWF metadata I got a character name Abby in the TRK4 column, I can search the word “Abby” and get all instances of the word regardless of column, then I need to re-enter “Abby” again the subfilter and select TRK4 to get just those results rather than being able to apply TRK4 right off the bat.
I also just discovered in testing this, that in the previous example, “Find” won’t find “Abby” in TRK4 column on the .sync subclip created regardless of filter or not. Which is not a good thing if I was looking for those clips…
Michael Phillips
-
I’m a little confused Michael. Are you suggesting that I should subclip each clip and then join them all together?
Reply to this Discussion! Login or Sign Up