[Andy Johnson-laird]

I can’t speak to the specific issues that you have been having, but I did have a look at the FCP XML specification (I found it via Google) because I couldn’t understand how to set markers colored other than red in CatDV — and I discovered that the FCP XML specification doesn’t allow for colored markers other than red ones.

I have a sneaking suspicion that the FCP XML specification (and FCP’s XML implementation) is a subset of the capabilities of the actual FCP software. That’s, of course, something way out of the control of CatDV.

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

Hey Kevin:
Thanks for that long answer. Much appreciated.

I’ve interlineated some comments to your reply.

[Kevin Duggan] ”
Well the best tip I can give you is to create a new catalog and drag and drop all of the previous content into a new GLOBAL catalog.”

That makes excellent sense. I’ve not tried dragging and dropping list view items to new catalogs, but now you say it, it’s obvious, init? 🙂

[Kevin Duggan] “CatDv can easily deal with thousands of clips in a catalog , it will maybe take a second or two to build the window but from on its all gravy as everything is loaded into RAM and its super fast. If you do this you will then realise that you have two friends in finding your stuff. One is what I call GLOBAL tagging.”

That also makes sense.

[Snip of details of global tagging.]

[Kevin Duggan] “Start to think like normal editing folk.”

That’s really hard for me. I only know abnormal editing folk.

[Kevin Duggan] “So for instance if you tag all your interviews with name of the interviewee and a tag INTERVIEW .”

Yeah…actually what I do is create special subjective encoding tags as geeks sometimes call them. By using a prefix of, say, * you can then easily refer to either *interview for the tags and not get hits when the verbatim logging includes the word “interview” because the talent actually said it.

I also put a number after the subjective tags, e.g. *funny 5 (where the number 1 – 5 means “not bad” to “really good.”

That way I can search for all *funny 5 moments in the verbatim logs.

[Kevin Duggan] “Bryson came to mind because this is a stock technique in reality shows of which he has done many.. He might be tempted to share some of his knowledge but he is a kindly but timerous animal and you may have to be patient , to see if he will appear..”

Do I have to close my eyes and make a wish? 🙂

Thanks again for the global catalog suggestion. I thought I was going to have to use the dodge of printing out to PDF, then using Acrobat’s advanced searching options (which does feature keyword proximity and stemming — pretty powerful searching features).

Thanks again for taking the time to respond.
Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

Hey Bryson/Rolf:
Thanks very much. I *knew* that there had to be a way. Never thought to look there, though. (Sound of hand smacking forehead! Duh!)

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

[Rolf Howarth] “CatDV already has naming schemes that will include leading zeros, the problem is they surround the name with parentheses (to indicate it’s an auto-generated name you might want to change to something more meaningful). What we really need is to introduce additional naming scheme options.”

You’re absolutely correct, Rolf. And, as you say, the problem then is that the left paren breaks FCP’s sort order and separates subclips from the master clips in the browser.

[Rolf Howarth] “By the way, I would claim the problem lies with the FCP browser window not sorting things properly. CatDV knows that 9 comes before 10 🙂
“

A valid claim So, I’ve sent an email to Steve Jobs. I’m sure he will arrange to fix FCP in a timely fashion.
So much so that I’m going to hold my breath until its fixed….

(thud)
🙂

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

[bryson jones] “Variable Expansion – in subclip names. hmm…
“

Well, actually it is more changing the format modifier currently in use.

The odds are (given that CatDV is written in Java), that there is a “%d” format modifier for the integer sequence number that is already being suffixed to the clip name.

To fix the problem I’m talking about requires that to be changed to “%02d” to make the sequence number appear as 01, 02, 03, etc. to allow for 99 subclips created from one master clip, or “%03d” to get 001, 002, etc. if 99 seems too small.

So, if we whine persuasively, we can maybe get Rolf or Kevin to insert the two characters “0” and “2” in the source code while they’re fixing the “null” string problem.

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

The problem with the approach you suggest is that I’ll then end up with a bunch of misnamed subclips and no apparent easy way of automatically renaming them when 8.1.6 comes along. (The back story is I’m just starting to catalog a project with 1,000 clips, each of which will have several subclips…)

An alternative approach is to click and drag through the clip Name: in the Summary window (not including the .mov part), do a Cmd-C to put it on the clip board, then, whenever I insert at Clip Marker, click and drag through the automatically inserted fragment of the clip name (but not the sequence number) and do a Cmd-V to paste in the correct clip name.

Then, when you hit Apply, the subclips get the correct names.

It’s a bit tedious, but it works.

There is one major caveat with using this technique: if you do the following:

1. Open up the Verbatim Logger.
2. Log a clip, inserting verbatim transcript, clip markers, and event markers.
3. Change the names as above.
4. Hit Apply.
5. Realize you’ve screwed up some clip names and want to correct the Verbatim Logger information….
6. And so you hit Cmd-Z (undo)…

Then you (a) see the subclips all go away (which is what you want), but (b) all of the text, subclip markers, and event markers also disappear (“this may be not quite what you want!”).

Workaround: Be sure to hit Save Text before you hit Apply!

Another problem with clip markers is that CatDV inserts a sequence number without any leading zeroes. So, in FCP when subclips are sorted by name you get this effect:

subclip 1
subclip 10
subclip 11
subclip 12
subclip 13
subclip 14
subclip 2
subclip 3

Is there are a way of getting one or two leading zeroes on the subclip sequence number automatically assigned by CatDV, please?

If there were, the sort sequence problem in FCP would go away.

Again, the workaround is to manually change the subclip name in the Verbatim Logger text entry to include a leading zero before you hit Apply — basically manually setting the subclip name to exactly what you want before it gets created by the Apply.

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

I agree Rolf. That’s a good mental model — the “Apply” button really means “Create” — at least that’s how I started thinking about it.

I’m now officially unconfused (at least about this). 🙂

Thanks for the rapid response.

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

Sorry…tried to Edit this posting as follows but just missed the 30 minute deadline.

The problem is that the SETTINGs for the Verbatim Logger were set to “Always save as text.”

If this setting is chosen, you can never create subclips from within the Verbatim Logger. You have to hit the Apply button to create subclips, not Save Text.

This is what is known in the software business as a “feechur” and it confused me. I thought creating a subclip created a subclip regardless, but, alas, all is not what it appears to be.

The “SUBCLIP” button really means “Create subclip unless you were dumb enough to set the Always Save As Text Feature in which case insert a subclip marker, but don’t actually create a subclip, merely lull the End Loser into thinking he’s created a subclip.”

You may now safely ignore this posting (and my original one, too).

Andy.
(Exits stage left and crawls back into his box.)

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

A well earned vacation, I’d say, Rolf.

I see the problem now: I was trying to ingest HDV format. I didn’t realize that LCP would only handle DV — as you say the HDV/GOP compression is substantially different from the independent frames of HD.

The meta problem I was trying to solve was to find a way of ingesting a tape that contains both timecode and date/time of recording and be able to extract that information. iMovie can name HDV clips according to date/time, but discards timecode. FCP does the inverse — it preserves the time code but discards the date/time of recording.

Sigh. I was trying to do this because I use an FS-CV DTE hard disk recorder from Focus Enhancements that normally names the clips based on the date/time when you start recording. But one of the three units I was using went rogue and was drifting like mad and was emitting date/time filenames back in 2003. Fortunately I was also recording to miniDV, but then the problem was how to get the date/time when the clips were recorded back into the filename.

Anyway, I solved the problem using some bash/gawk scripts I wrote on the Mac and a little command line interface program called aditc that, for a given .mov file, extracts the timecode and spits it out to stdout. I don’t know who wrote aditc but I owe them several beers.

(Oh…and the deck/cameras were connected directly to the Macs — the problem is/was the HDV issue.)

Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon

[Andy Johnson-laird]

That’s exactly what the extend marker function does. It creates a tail to the marker to mark an end point. When you make that into a subclip, the subclip runs from the head of the marker to the end of the tail and not to the next marker.

Aha! Now that’s something else about which I was completely ignorant, Tom. (There’s an ever increasing number of things about which I’m completely ignorant, these days).

Thanks for that. I’ll experiment with markers some more. What I’ll need to try is something like this where I extend the Marker 1 to form a tail that actually overlaps the start of Marker 2 (see below where “TT” is tail and “===” is the marker extension:

………. M1====TT
…………………….M2====TT
==========================
Not sure whether FCP will allow an extended marker to overlap with the start of another marker. Nevertheless, you’ve given me some ideas to try — for which thanks!

Again remember to keep markers very loose.

Understood. The issue, as you can doubtless see, is where the are two subclips, one right after the other in the master clip.

Am I right in thinking you’re shooting continuously without stopping and starting the camera?

For the long form work I do, yes. The Canon FS-CV allows for about six hours of recording. I typically run the cameras non-stop except when there’s a break of more than five minutes.

For the short form work, I’ll stop more frequently, usually so that we can tweak the set, etc.

Thanks for the ideas.
Regards
Andy

Forensic Software/Internet/Image Analyst
Portland, Oregon