EDL for film rescan

[Jeremy Rush]

Follow up…

I exported an EDL sorted by “Source” and compared the timecode to the timeline burn in. The order of the edits on the EDL is completely arbitrary, as far as I can tell. it is not sorted by Labroll or timecode. And the edit events on the EDL are from all over the timeline.

I have no idea what to do.

-Jeremy-

[Doug Beal]

Is this cut at 24 fps? (23.976, 23.98)
What frame rate did the scans go to and was there an ALE or flex file generated from this scan?

sorting by source should show groupings of reel numbers ascending TC and master in outs all over the place
sorting by master should show source code and reel numbers all over the place but the master TC ascending from edit 001 through to the end

a flex file or ALE should match film frame to BITC dub TC accurate
Is there BITC on your locked edit?

Doug Beal
Editor / Engineer
Rock Creative Images
Nashville TN

[Jeremy Rush]

Hey Doug,

Thanks for the response.

I do not have a Flex file or an ALE from the original film scan. The cut is at 23.98fps. But the timecode of the EDL matching the burn-in is not the issue.

The issue is that sorting the EDL by “source” does not produce groupings of reel numbers with ascending TC.

Here is the first ten lines of the EDL sorted by “source”:

TITLE: NO ROAD_SORTING_SOURCE
FCM: NON-DROP FRAME
*FCM: DIGITAL TV 24P

001 L8546015 V C 01:01:10:22 01:01:15:00 01:00:11:11 01:00:15:13

002 L8546017 V C 03:08:52:12 03:08:57:15 01:00:06:08 01:00:11:11

003 L8546015 V C 01:04:34:17 01:04:37:03 01:00:18:19 01:00:21:05

004 L8546017 V C 03:09:17:22 03:09:21:04 01:00:15:13 01:00:18:19

005 L8546015 V C 01:05:32:06 01:05:34:09 01:00:21:05 01:00:23:08

006 L8546015 V C 01:06:57:04 01:07:00:12 01:00:23:08 01:00:26:16

007 L8546009 V C 02:24:27:21 02:24:39:09 01:00:32:11 01:00:43:23

008 L8546015 V C 01:02:26:14 01:02:32:09 01:00:26:16 01:00:32:11

009 L8546009 V C 02:29:05:00 02:29:09:20 01:00:47:01 01:00:51:21

010 L8546009 V C 02:31:18:17 02:31:21:19 01:00:43:23 01:00:47:01

Going through this list I can identify the clips in my timeline and match edits with the timecode on the burn-in, but there is no discernible order to the EDL. You can see that it is not sorted by reel number. And it is not a cut-for-cut match to the timeline. These clips are from random places throughout the locked edit. You can also see that the edit event numbers are in chronological ascending order, but they do not match the chronological edit events in the locked timeline (which a “master” sorted EDL should do.)

My understanding is that a “source” sorted EDL will produce a list of edit events in order from the lowest reel number, with timecode ascending for that reel number, through he highest reel number, so that the telecine lab can pull the selects reel by reel. And that a “source” sorted EDL will NOT have ascending numbers for the edit events.

Here is a sample of a good “source” sorted EDL. (It was provided as a sample to me by the lab.)

TITLE: ARCHANGEL_DATACINELOCK_FLAT
FCM: NON-DROP FRAME

235 K425 V C 09:10:44:03 09:10:46:18 01:11:55:18 01:11:58:09
237 K425 V C 09:10:49:11 09:10:51:07 01:12:00:04 01:12:02:00
239 K425 V C 09:10:56:14 09:11:00:22 01:12:05:04 01:12:09:12
233 K426 V C 10:01:08:11 10:01:11:04 01:11:50:16 01:11:53:09
241 K426 V C 10:01:58:19 10:02:01:21 01:12:11:21 01:12:14:23
291 K427 V C 11:00:06:18 11:00:16:05 01:14:59:10 01:15:08:21
293 K427 V C 11:01:27:14 11:01:28:15 01:15:10:14 01:15:11:15
295 K427 V C 11:01:40:05 11:01:41:12 01:15:14:23 01:15:16:06

Here are the first ten lines from my “master” sorted EDL, which DOES match the cut-by-cut chronology and TC burn-in on the timeline.

TITLE: NO ROAD_SORTING_MASTER
FCM: NON-DROP FRAME
*FCM: DIGITAL TV 24P

001 L8546017 V C 03:09:46:21 03:09:50:08 01:00:00:00 01:00:03:11

002 L8546017 V C 03:27:01:18 03:27:04:15 01:00:03:11 01:00:06:08

003 L8546017 V C 03:08:52:12 03:08:57:15 01:00:06:08 01:00:11:11

004 L8546015 V C 01:01:10:22 01:01:15:00 01:00:11:11 01:00:15:13

005 L8546017 V C 03:09:17:22 03:09:21:04 01:00:15:13 01:00:18:19

006 L8546015 V C 01:04:34:17 01:04:37:03 01:00:18:19 01:00:21:05

007 L8546015 V C 01:05:32:06 01:05:34:09 01:00:21:05 01:00:23:08

008 L8546015 V C 01:06:57:04 01:07:00:12 01:00:23:08 01:00:26:16

009 L8546015 V C 01:02:26:14 01:02:32:09 01:00:26:16 01:00:32:11

010 L8546009 V C 02:24:27:21 02:24:39:09 01:00:32:11 01:00:43:23

I have no idea what is going on with the “source” sorted EDL or how the sorting is arranged. It seems to me to be completely random. I have no idea what to do or how to even begin to approach a solution to this issue.

Any ides?

Thanks,

-Jeremy-

[Jeremy Rush]

Hello,

I received a forum response to this same thread that I made over on the apple.com forums. Reik (creator or letterbox 2.0) provided me with a solution. Apparently my reel names are too long. FCP CMX3600 EDL files can only handle a 6 character reel name. I will make the appropriate changes today and make sure this works.

Thanks,

-Jeremy-

[Doug Beal]

Glad you found an answer

this is not for the feint of heart but an edl is very much like a text file arranged by tabs
I often use Text Wrangler for resolving some issues due to it’s find and replace capabilities.
the cool thing is if you make the changes to the edl, import in a new project and relink ignoring the reel# doesn’t match, you should end up with your edit. Then export source. a lot of work.

depends on the situation if you don’t have a lot of reels, changing the reel number can be the way to go.
I’d suggest a print out your EDL and make notations about your new naming conventions.
Easier than remembering after a phone call or a beer or
we keep naming conventions strict in house here cause we use EDLs a lot and if it doesn’t work it’s a lot of non billable work to fix.

Doug Beal
Editor / Engineer
Rock Creative Images
Nashville TN

[Jeremy Rush]

Hey Doug,

Thanks for the info.

I have not actually implemented a solution to my reel # naming issue. Do you have any recommendations?

I exported the EDL, with the 8 character reel names that were out of order, and then used Excel to sort the EDL by ascending reel names and timecode. I have no idea if that was a good idea or not.

I sent this new EDL to the lab, to see if it was formatted properly.

Can you tell me what you’re thoughts are? I shot on super 16mm film, scanned to DNxHD MPX files, did the edit in FCP. Now I’m looking to get either DPX or Tiff files scanned from the lab to conform the negative at home, on FCP.

Any opinion on DPX vs. Tiff?

Then I’ll be adding VFX, sound, etc. I’ll be color grading at Technicolor.

I’d like to then lay to D5 tape for archiving and exhibition (as well as other formats depending on film festival preferences.)

Any advice on this process/workflow would be appreciated. I’m trying to learn as much about postproduction as possible on this project. So far it’s been a painful, and rewarding journey.

Thanks,

-Jeremy-

[Doug Beal]

I would get with technicolor as to what they would like to see. They will also have recommendations for
workflow.
AJA has a dpx to quicktime converter to download on their site.
after converting to QT you’ll want to throw it in compressor and generate a pro res or similar in order to play it (10 bit RGB is too much data for almost all but high end arrays) xfer to prores maintains color fidelity quite well. I tale dpx’s from our scratch system to QT using AJA’s converter regularily.

Doug Beal
Editor / Engineer
Rock Creative Images
Nashville TN