I am looking for a way to edit an AVC encoded .ts file (1080i) by replacing and adding some of the parts with segments from a MPEG-2 video. After that, I would like to perform some cutting in MeGUI and apply a few filters in AviSynth, deinterlace, and finally export everything to a mkv (x264) file.
What I am having difficulties with is the editing part. I would like to know how to combine AVC and MPEG-2 files and use the result as the input to the AviSynth script. Are there any tools out there that allow this? Prefferably able to export/create to AviSynth script.
Any points and tips will be much appreciated.
Thank you in advance.
+ Reply to Thread
Results 1 to 6 of 6
You can do the editing and joining in the avs script - it makes sense to do it that way if you're only doing simple edtis since you're using avisynth anyways . AVISynth doesn't care what compression is used in the source file; everything is decoded to uncompressed (all operations are done in the uncompressed domain)
Thank you for the reply.
Is there any simple way to make precise cuts in AviSynth? I am trying to get my head around how to make the cuts more efficient. I know that in MeGUI it is possible to cut ONE file at the time (either AVC or MPEG, depending on which one that is currently loaded). What I would like is some kind of multitrack editor in which I can make cuts/joins for both videos, and export/generate cut segments (as MeGUI does, but for a single video at the time) as AviSynth.
It seems a tedious task to manually play both videos and try to make accurate cuts that way. I hope I managed to explain my problem.
You can load multiple videos in 1 script.
The closest thing to an "editor" for avisynth is avspmod . Push F5 to preview.
If you're doing any editing, you should be using a frame accurate source filter - that's where the "precise" comes into play. If you have an Nvidia card, I would suggest DGDecNV (not free), but a free non GPU alternative that handles AVC in TS fairly well is L-smash . For MPEG2 sources, DGDecode is still very reliable
You can use Trim() to specify start and end frames. Aligned splice to join segments - it's really not that difficult, unless you need to "hear" audio timing simultaneously to sync your edits. In that case, you should use a NLE
e.g let's say video A is AVC TS, video B is MPEG2 TS. You want frames 0-1000 from video A, frames 3000-4000 from video B, frames 2000-3000 from video A, all spliced together
A = LWLibavVideoSource("videoA.ts") B = MPEG2Source("videoB.d2v") A.Trim(0,1000) ++ B.Trim(3000-4000) ++ A.Trim(2000,3000)
AVS Cutter under the Tools menu? Is that what you're using? You can add as many cuts to a script as you like. The source video type is irrelevant. If you want the audio to match you have to save the cuts file for loading into the audio section or using with the Audio Cutter.
What I sometimes do is add the required number of Trims to each script, then make as many copies of each as required so I can delete all but one Trim() and each copy of the script is then encoding a unique segment. You can encode them individually and append the encoded video together, or you can combine the individual scripts into a single script. something like:
Import("D:\script 1.avs")++Import(D:\script 2.avs")++Import(D:\script 3.avs") etc etc.
Doing it that way might make matching up the audio a little more tricky. You might need to create the copies of each script before adding a Trim() to them, as that way you can save a matching cuts file for each script. Or you can create a separate script yourself to encode the audio correctly if need be.
Edit. I just read poisondeathray's post. Combining the Trims into a single script that way may in fact be easier. It depends how much cutting/combining you have to do and how familiar your are with avspmod etc.
I didn't include audio in my example above, but it's trivial to add it into the script with AudioDub() . Trim() would cut both audio & video in that case . If you need an example or more help just ask