|
|
Is there any tool where you can see that there are actually 2 fields present for a progressive file? In the case you mention, the resolution of each field would be half and the data of one frame would be spread in two fields. Spatially correct, but temporary incorrect. By setting the Main header as 30p they could fool many decoders but not them all because you have to process the two fields anyway (odd first/even first?). If I playback my SX1 FHD originals in MPC-HC by using ATI's hardware decoder or ffmpeg soft one, I get correct statistics of 30fps being played by EVR. If I step frame by frame, I see the image change every step, so I guess that if it is true that they are 30p ala 60i, ffmpeg and ati hardware decoders are smart enough and capable of be instructed to render P content and on the fly reassemble the two fields and render the P frames.
The question is if Cyberlilnk's decoder inside PD8 does the same.
If I render one file only( from my FHD 30p) originals, the result is Ok, but as soon as I make combinations, namely more pieces and multitrims, then the result is still P but with Interlaced artifacts as the pic I posted shows in the thread that is referenced here.
I hope there is someone with enough expertise and the soft tools to point us in the right direction.
If the originals are indeed 30p with two fields (60i) the easy solution would be to get a tool to preprocess the originals in order to merge the two fields in a real Prog frame. That way there is no chance that CL's decoder could make any mistake.
Saludos!!!
|
|
|
Hi all, I was busy the whole week and didnt have time to test the changes in the profile until today.
Thanks Dafydd and Ad for your support, but unfotunately, the problem is somewhere else. I Encoded the same file again, but this time without any transitions nor effects, just the source files. My camera is a Canon sx1 IS, fullHD 30p files in a mov container. I assumed that the PD output files were 60i because of the interlace artifacts, but the mediainfo shows something else. Attached the file you will find. The thing is, the main header seems ok, but the artifacts are still there. Im not sure if with avc you can do the same thing as with mpeg2, where you can have a main header with certain params and another header per frame or gop with another AR for example. I haven't found a tool to analyze this yet but it'd be interesting to check if the frames match the header info. If I load the produced file in tsMuxeR it recognizes it as 720i, so there is something fishy going on.
Is there a way to force (via registry ie) the encoder to make 30p and not 29,97(i/p) ? I could send you the source and production files so you can check them and maybe we can see where it goes wrong.
Thank you guys.
|
|
|
Hi, Im trying to generate a mt2s file from the .mov files my canon camera generates. When I include them in the project they are correctly detected as 1920*1080 @ 30p Stereo Raw audio. I try to produce a 720p video from it by choosing a H246 AVC file with a custom profile (ntsc) by choosing 1280*720p.
When I check the generated file, the resolution seems ok but the file is 60i. and not the original 30p of my files. You even see the comb teeth issue common in interlaced content. Where can I disable this? I want the output files to have the same frame format as the input ones, namely 30p.
Thanks.
|
|
|
