On Monday 11 June 2007 22:14, Aaron Newcomb wrote: > So, what does this mean really? If Cinelerra decodes and re-encodes > what do you lose in the process? And what about 4:4:4 as was mentioned > earlier?
Mpeg2 is lossy compression. That means that you will have artefacts in the frames. Cinelerra decompresses each frame and sends it through the render pipeline. At the end, it sends the frames to an mpeg2 encoder. At this point, the encoder can only assume that its input is the "real" picture, including the artefacts. Therefore, it tries to encode the artefacts, too. Since artefacts are usually almost noise, extraordinarily many bits are consumed to encode the artefacts and fewer bits remain to encode the "real" (wanted) image information. Consequence: even more artefacts. YUV 4:4:4 is roughly a superset of 4:2:2, so in this transformation no information is lost. A bit of averaging may happen (I'm not sure) when the conversion back to 4:2:2 takes place, but this is something you want for mpeg2 input because it may reduce the input artefacts a bit. -- Hannes PS: Care to not top-post if you expect an answer from me, please? > > On 6/11/07, Johannes Sixt <[EMAIL PROTECTED]> wrote: > > On Monday 11 June 2007 17:39, Terje J. Hanssen wrote: > > > No response on this topic so far. Therefore I try again with two > > > follow-up questions: > > > > > > 1) Is Cinelerra capable to preserve mpeg2 with 4:2:2 color space > > > throughout the editing process? > > > > No. Cinelerra decodes and re-encodes. But it has YUV color space. > > > > -- Hannes > > > > _______________________________________________ > > Cinelerra mailing list > > [email protected] > > https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra _______________________________________________ Cinelerra mailing list [email protected] https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
