-
transcoding one video file with multiple settings
Hi everyone!
I’ll get straight to my question:
Is there a way (or program) to transcode one single video file with multiple settings?
what I mean is: I would for example like to have different GOP structures for different parts of the video. (Let’s say only i-frames for the beginning transition from black in a video and for the rest of the film a GOP structure of 1i-; 5p-; 2b-frames or something alike.)
well, I’m not even sure if any of this is technically possible but I believe I have read about transcoding with multiple settings in compressor by marking different regions of a timeline in FCP before exporting it to compressor. I admit I might be hallucinating because I haven’t found a single word about this ever since. 😉
The reason I ask is that I usually get very bad transcoding results with transitions (especially from or to black) or with other parts of my vids – surprisingly when very slow and subtle movements are being shown and typically when there is some kind of gradient in the background (i.e. slowly moving shadows with a large umbra) whereas all the other parts of the video are ok.
It would be great if anybody could share his experience with improving the results of transcodings in such problematic regions of a video or if anybody knows about ways to encode files with multiple settings (if possible).
greetings and thanks from vienna, austria
manuel zauner
c/o studio trizepsPS: we typically use SONY EX-3 and Canon 5D markII, sometimes a phantom if that is of any help
PPS: right know I prefer using Episode as transcoding software.
-
3 Replies
-
Well if you’re compressing for DVD Video (MPEG2) compression markers can force I frames which can be used during transitions or other fast action.
A good 2 Pass VBR setting will allocate more bits to difficult sections.
In the distant past I remember talk that when Microsoft Expression Encoder was under development, it might have the feature to re-encode sections with different settings.
-
compression markers forcing i-frames.. now I remember – that’s what I’ve read about… 🙂 thanks craig. I will try that.
the funny thing is that I usually get good results with fast action but mainly get artefacts in the slow movements of subtle gradients.
about 2-pass vs. 1-pass encoding I’m a bit uncertain. there are so many different things I’ve read about this. (I’m sure I could find some threads in this forum, I didn’t look into that to be honest though.)
in the last tutorial I’ve read (don’t ask me where) I was told that 2-pass encoding favors encoding speed for quality by being as precise as possible concerning a predefined file size, whereas in 1-pass encoding I should be able to choose between bitrate as a criterium or constant quality (without regards to file size) or as a third option: “constant quantizer” (choosing quality by complexity). all of that is nice in theory… I haven’t found an encoding program that does allow to choose by exactly these options. I tend to believe that the CQ – method is the one used by most encoders today by default. maybe the tutorial was really old too.
oh well.. so much to still learn… ‘sigh’ 🙂
thanks however fo the tip. I’m looking forward to make some tests with forced i-frames.
manuel.
-
[Manuel Zauner] “in the last tutorial I’ve read (don’t ask me where) I was told that 2-pass encoding favors encoding speed for quality by being as precise as possible concerning a predefined file size”
Different codec have different settings and with H.264, with so many implementation, it really depends. It doesn’t sound like that tutorial is valid unless you’re misinterpreting what they were recommending.
Reply to this Discussion! Login or Sign Up