AviSynth and GH1 1080p25 50i wrapper removal

F

frear

Member
Hello, this should be the easy route but.. when I import transcoded files into they are recognized as 1080i.

- just rewrapping from mts to mp4 and importing = shows 1080p so it's proved that gh1 source is progressive (if I had any doubts..).

Okay, so.. here's the avisynth syntax to remove the interlaced wrapper:
directshowsource(name.mts).ConvertToYUY2().AssumeFrameBased().ComplementParity()

Then transcoding to Lagarith or UT in VirtualDub, none recognized as progressive.

I also tried with something like AssumeTFF.SeparateFilelds.SelectEvery(2, 0,1).Wave()

But still I can't get it seen as progressive.

?? :Drogar-BlackEye(DBG

I know the 25p are a reality, but I still would like to have the choice of properly editing w/ LP reliable footage.

If that was correct, I'll assume there's a limitation in the NLE or codex, and live with that. :)

Thanks.
 
I honestly read and tested all sorts of methods before my unhackable gh17 became hacked a few weeks ago after Vitaliy got the gh17 ptool out.

All my tormenting waste of time trying to figure out all that insane transcoding crap is over as I can just use the blackout powell native patch and I don't have to even know what transcoding even is anymore. Seriously why torture yourself, there is no end to it if you try to go about getting consistent progressive files that way rather then shooting native in the first place. I really don't think you will ever find the answer you are looking for when trying to mess around with transcoding crap, it is an endless path to nowhere and I am glad I will never step back on that road again.
 
> AssumeF rameBased
> AssumeTFF.SeparateFilelds.SelectEvery(2, 0,1).Wave()

No gap between the F and a of AssumeFrameBased ... and it would be SeparateFields .... and Weave() not Wave(). But I assume you had those right in the script itself or else Avisynth would have thrown an error?

I'm not sure what you mean by "seen as progressive" since Virtualdub doesn't really identify interlaced footage. In Vdub use the ViewFields 1.2 filter to inspect footage to determine if its progressive or interlaced: http://www.geocities.com/siwalters_uk/viewfields.html

I'm not sure what's up with @Observe - the avisynth/vdub combo can do a great job of converting genuinely interlaced footage to progressive, so I would have thought converting PsF material would be a breeze.
 
GrahamH said:
> AssumeF rameBased
> AssumeTFF.SeparateFilelds.SelectEvery(2, 0,1).Wave()

No gap between the F and a of AssumeFrameBased ... and it would be SeparateFields .... and Weave() not Wave(). But I assume you had those right in the script itself or else Avisynth would have thrown an error?

I'm not sure what you mean by "seen as progressive" since Virtualdub doesn't really identify interlaced footage. In Vdub use the ViewFields 1.2 filter to inspect footage to determine if its progressive or interlaced: http://www.geocities.com/siwalters_uk/viewfields.html

I'm not sure what's up with @Observe - the avisynth/vdub combo can do a great job of converting genuinely interlaced footage to progressive, so I would have thought converting PsF material would be a breeze.
Click to expand...

Yes Graham, I'm sorry I wrote it without copypasting but the script is correct.

VB transcodes just fine, but when I import the avi in the editing program it recognizes it as 1080i. Not p.
Which to my eyes shouldn't be possible.

If you tell me that the routine is correct (since I'm not the avisynth ace..) then it must be a problem with the NLE. Or some indexing stuff with codex/avi wrapper.
 
Last edited:
I have the same problem. I thought that the problem was present only on NTSC cameras. However, when checking my final footage in more details, I was shocked:

http://www.dvxuser.com/V6/showthread.php?252435-Edius-conversion

What did you mean by "- just rewrapping from mts to mp4 and importing = shows 1080p so it's proved that gh1 source is progressive (if I had any doubts..).". Which program have you used to do that?
 
GrahamH said:
Is this any help? http://www.dvxuser.com/V6/showthread.php?250107-How-to-handle-25p-footage-in-an-50i-wrapper
Click to expand...

Unfortunately not. It doesn't help even if using any combination of progressive/interlaced project and/or clip properties. When I export to Canopus lossless codec (or some other Canopus codec) I get artefacts.

GrahamH said:
Edit: or perhaps this (note the comment about how to avoiding comb artifacts in Win7): http://www.dvxuser.com/V6/showthrea...via-AviSynth&p=2275266&viewfull=1#post2275266
Click to expand...

I am using XP, so it should not be related to the mentioned problems.
 
crunchy said:
What did you mean by "- just rewrapping from mts to mp4 and importing = shows 1080p so it's proved that gh1 source is progressive (if I had any doubts..).". Which program have you used to do that?
Click to expand...

Just grab ffmbc (or ffmpeg, but it has problems with mts) and do:
ffmbc -i input.mts -vcodec copy -acodec pcm_s16le output.mov

Same syntax for ffmpeg.

Then import the .mov it should do the trick if edius can deal with the content (lightworks does not).
 
Last edited:
You can download a precompiled binary of ffmpeg for windows here
http://ffmpeg.zeranoe.com/

But both ffmpeg and ffmbc detect the PAL GH1 25p footage as interlaced. See the other thread for details. The content is progressive, but it has been encoded interlaced, so ffmpeg "thinks" it's interlaced

EDIT: sorry some ffmpeg builds think it's interlaced (my current ffmbc build thinks it's interlaced), but others think it's progressive but get the frame rate incorrect. Go figure.

this is what ffmpeg thinks
Stream #0.0[0x1011]: Video: h264 (High), yuv420p, 1920x1080 [PAR 1:1 DAR 16:
9], 50 fps, 50 tbr, 90k tbn, 50 tbc

this is what ffmbc thinks
Stream #0.0[0x1011](und): Video: h264, yuv420p, 1920x1080i tff [PAR 1:1 DAR
16:9], tbc 2/50, fps 25.000

Both are incorrect. 50Fps progressive vs. 25i
 
Last edited:
crunchy said:
Thanks for the answer! Does this program exist for windows?
Click to expand...

It is for Windows, you must compile it (tech stuff) or download it from someone Who did it.
This might work:
http://www.dma.ufg.ac.at/assets/23954/intern/ffmbc_compiled_by_S_20110607v06rc7.zip

I don't know why it sometimes works and some other times it does not.
But giving it a try might be worthy. That's the only suggestion I can give, to try. :)
 
You must log in or register to reply here.
Back
Top