Mark Spano
Forum Replies Created
-
Reverse telecine only works with proper 2:3 pulldown. It will not work at all for what you are trying to do, which is to remove every fifth frame. For that, you need to do it manually, as there is no automatic solution. In the past, I’ve invoked FCP and MadRuby. Drop the offending clip into a matching FCP timeline, and with MadRuby you can have it learn a keyboard sequence to step forward four frames, mark in, mark out, ripple delete. Then MadRuby can fire this learned macro off as many times as you want (to save you the carpal effects). When done, export same as source, and then load into Cinema Tools and conform to 23.98. Now you have the original 23.98 clip with the offending pulldown removed.
-
Actually, no compressor program will get you back there. But it isn’t as hard as you think. If the cadence is like you say (A B C D D A B C D D) and it is consistent throughout, here’s what I have done in the past to rectify:
- Open FCP
- Choose 29.97 easy setup
- Create new project
- Create new sequence
- Import 29.97 (2:2:2:4) file
- Add file to sequence
- Remove every fifth frame (for this, I use a genius little macro tool called MadRuby which allows you to record a few keystrokes and then play them back on repeat until done. So you can record “right,right,right,right,i,o,shift-delete” which is a key sequence that will remove every fifth frame. Then you can set MadRuby to playback with repeat x number of times and it will edit out the duplicate frames for you.)
- Export newly edited sequence as QuickTime Movie, Current Settings, Self-Contained
- Open this QT in Cinema Tools
- Choose Conform…
- Conform the QT to 23.98 fps
Now you have the original 23.98p file without any loss in quality or duplicated frames.
-
[Michael Bekkers] “we want to output our desktop to the RGB cables. We don’t know if it is possible, but we do suppose it is.”
Yes, it is. What you want is to install the drivers for the KONA 3G but not the NDD version (NDD stands for no desktop display). Once you have the full 3G drivers, you can make the KONA 3G output become an additional desktop. In the Control tab of the control panel, under Default Output, you see now an option to choose Desktop. Now you have a new screen. In the Mac’s System Preferences for Display, you may also be able to choose whether this display is an additional display or to mirror the first display.
-
yes, guessing that your friendly post house is using FCP to lay off. FCP will take what is known as ‘discrete’ or ‘mono’ channels in a QuickTime source file and automatically pan them all center in a Stereo Output sequence.
What your post house should do is not pan at all, but change the sequence setting to 4-output Dual Mono. Then assign each channel in your QuickTime to the output channel it belongs. It looks like this in the Sequence:
-
[Rafael Amador] “Could be done in FC, but that will means to lose one generation (rendering), when you just need to read the Prores444 file as Upper instead of Progressive.”
No generation loss, since there’s no master yet. The export from FCP is creating the master.
[Rafael Amador] “And is not if Media Encoder “might” translate P in to PSF. Media Encoder will do it.”
OK – this was what I was referring to by saying “sure I could test it”. You tested it and it works – I have not since I haven’t had the reason to test. I was just offering a method I knew was accurate and foolproof, without speculating if another workflow would work.
-
Rafael,
I understand how PsF works. What I’m saying is why bother guessing that Media Encoder *might* translate P into PsF, when you already know that FCP will do it? What if Media Encoder decides to do some interpolation to get there instead of simple duplication? I could test this, for sure, but since you need the intermediate file anyway (coming out of FCP), why not just make that PsF (upper field) and eliminate the variables?
-
The “normal” way this type of thing works is this:
You edit with either source or proxy footage, maintaining reel names and source TC. Once locked, you just export an EDL and hand that plus your source files over to your colorist. Colorist imports source footage and then uses EDL to parse the source footage and recompile your cut. Colorist grades using full res and bit depth of source footage, exports a consolidated QT which you can then bring back into your edit, apply titles and graphics, and master.
Now here’s where this can get tricky. You’re using FCP. It’s notoriously bad at this kind of workflow if left alone to the software to manage everything. You have to take control of it. When importing the Alexa files, you need to make sure you give each ‘card’ footage a reel name if it does not already have it. This is so that the EDL at the end works. FCP only gives you a simple CMX 3600 EDL format. If you had Media Composer, you’d not worry about this since you can use the File_32 EDL format which allows the Alexa source file names to be used as reel names. Anyway, apply reel names to your footage, then edit. Keep your edit to V1 as much as possible. When you’re locked, export the EDL. You won’t get the error about clips not having tape names then, since you added them. Send the EDL and your source files to the colorist.
It is also likely that the colorist is using a system that can take in a FCP XML. This is preferable, because if you have more than one video track in your cut, the XML gives all the information for all tracks. I would supply the colorist with both EDL and XML – each takes a few seconds to export, so it’s not an issue.
The reason I advise this workflow is that so many times I get an exported QT from someone to grade, it is not the same format as the source footage. Especially if multiple sources are used (Alexa, Canon, RED, stock footage, stock photos, etc.). If you supply all your source footage, you’re guaranteeing that your colorist will have the best material to work from.
-
Are you sure Rafael? The Blu-ray spec I have says 25p is not able to be encoded on disc. Unless you’re saying that Media Encoder will automatically do the duplication from frames to fields, but then why bother chancing it when you know FCP will do that without issue?
-
[Thomas Casque] “But 25i ? I thought 25p equals 50i – since two half-images per frame ?!?”
Nomenclature – it’s confusing. I always try to write so that the number represents frame rate, and have the letter designate whether it’s interlaced or progressive.
25i = 25 frames per second = 50 fields per second.
50i = 50 frames per second = 100 fields per second = does not exist.
If the projectionist said 50i he probably meant 25 fps interlaced. Double check with him.
[Thomas Casque] “Do you think fcp will do just as well with the interlacing as the adobe media encoder ?”
Definitely use FCP to create the interlacing. FCP is good at the dumb tasks, so duplicating frames into fields is fine in FCP. Then your Media Encoder only has to worry about taking each field and encoding into H.264.
-
[Thomas Casque] “But another thing is, that I need to interlace my 25p project. Looks kind of simple at the first glance. Just hit interlaced and select upper or lower field order…”
I would first ask the projection if they can use 25 fps. Some projection can only use 24 or 23.976 fps.
If you want to make the Blu-ray with 25i, when you go to output your ProRes 444 from FCP, first duplicate your master sequence, and go to Sequence Settings, change field order there. It will not change any clips in your sequence, just duplicate each frame into two fields on output. The render will take a bit longer to export since it’s creating fields where there were none, but it’s not doing any motion interpolation, just duplication, so it should be easy. Definitely check this file (on a broadcast monitor) once you’ve output it. It should (for all intents) look identical to your 25p master.
[Thomas Casque] “How was that again with the field order and interlaced material… lower field for PAL and upper field for NTSC – or is this kind of irrelevant for Blu-Rays?”
That’s SD talk. In HD, all interlaced material is Upper Field First.
