-
Qt – FCP playback & render relations? (issues with avc)
What is the relation between Quicktime (Pro), FCP Viewer, FCP Canvas (=sequence, rendered), and exporting using Quicktime conversion, or Compressor?
Since FCP uses the Quicktime libraries to render, I also assumed that it uses most of its engine. But I’ve come across so many differences in functionality between the parties above (differences in playback, in need to render in order to view it, in exporting..) that I’m totally confused now about my assumptions – which means I’m right at the magical moment of learning something :-).
Apparently things are handled quite differently, and differently per codec. Is there anyone who can enlight me a bit?
____
Below a recent (unsolved) example issue concerning the rendering & conversion of an avc1 codec (which uses the opensource libavcodec library). TchSoecs: 2x 2,8 GHz quad-core, 4GB 800 MHz RAM, OS 10.5.8, FCP 6.0.6.1) I was presented with a series of to-dvd-recorded material, transcoded in avc1 avi files. I installed the (free) avc1 codec, and made .mov’s out of the orignal material, in order to clean up and export in FCP.
2) Quicktime (Pro) 7 plays it fine. (Finder does not show a preview though, only sound). VLC plays it too (it has avc1 installed by default).
3) In FCP 6.0.6
– Viewer shows, but with quite some difficulty (low, irregular playback framerate (aka stutter), flashing cross-icon in right upper corner, white ‘processing’ arrow). sometimes it just stops by itself, too tired of doing something. My assumption was always that the viewer is entirely handled by Quicktime, like an embedded Qt player in FCP. Guess I was wrong.– Although sequence settings are set to match clip settings, everything (audio too) needs rendering. Which takes upt to 15min for a 30s. excerpt… After rendering, a lot of artifacts entered the videostream that were not there in other playbacks, nor in the original material (squares, frames mashed up in time). Looks like I can’t trust FCP rendering by itself. My second assumption has been, that rendering was handled by a ‘specialized’ FCP engine, that applied all installed Quicktime codecs, and those installed for FCP. Wrong again.
4) Hence I tried exporting the finished sequence using quicktime conversion, keeping the ‘same’ codec settings (h264 for avc1) and size. After 9+ hours of rendering I found it crashed and undone. I’ll try Compressor now.
5) If that fails, we’ll go back in time (as 2006 posts on this forum suggest) and use Streamclip or likewise to do the import from DVD again, this time straight into .mov.
Sytse
-
1 Reply
-
Thanks, Dave!
I’ve resolved my issue meanwhile (reimported all), but the main question (top of post) remains. I read between the lines of your post (and that of many others here) that there is an enormous lot of industry-specific context knowledge hidden behind the codecs list that drops down when ‘export settings’ is pressed. Which made me wonder 10y ago, and still. That’s why the web is good.
[Dave LaRonde] “H.264 is not among the standard set of codecs that’s best in FCP. It’s extremely processor-intensive and difficult to edit.” I noticed that, and hardly ever use it for editing; I was under the assumption though, that it would at least be handled smoothly as a delivery codec, aka replay and export. At least by Apple themselves.
[Dave LaRonde] “In any single edit timeline it works best using one single solitary video resolution, one single solitary frame rate, and one single solitary codec.” True for the industry-side that grows on well established soil of film, video & tv. Less true -or at least continuously under attack- for the yet dirty side of digital video exchange.
I wish, wholeheartedly, that things were more clear and singular for transcoding recording, editing, delivery, and intermediate formats (and let’s not get started on the Metadata issue). But you find the web exploding of codecs, transcoders, and forums full of questions about it. For my clients (at the moment its Universities) every image that moves on a screen is video, and I stand often alone and abhorred watching what trashed qualities they willfully accept or deliver- as long as some deadline is matched. Me thinking things like: Why would I even try explaining what could really be done using the EX-1 and 7D they own, if they happily accept anything YouTubeified the same day? But let’s not get too drifted away.
So my question is in fact quite practical. There’s a list of codecs that Quicktime uses for itself, and a different one that FCP uses. Very easy to see, just compare what’s in export, or sequence settings in both programs. My question was aimed at FCP’s specific use of these (sub)lists. Like in: how come i can see some clip fluently in the viewer, but not in the canvas, although settings are the same? Where is the line between what’s used (and useful) of Quicktime, and FCP? I’m sure this forum hosts some reference on that.
(Please don’t tell me to start using iMovie 🙂 )
Reply to this Discussion! Login or Sign Up