eac3to - audio conversion tool

[rickardk]

Just did a new attempt with Born on the 4th of July.
2.15 gave a couple of frame gap/overlaps
2.16 gave less frame gap/overlaps problems
2.13 gave a smooth playable file with audio in synch (watched a couple of minutes at the start of the movie, att the middle and a few minutes before the end)

[madshi]

Originally Posted by rickardk

Just did a new attempt with Born on the 4th of July.
2.15 gave a couple of frame gap/overlaps
2.16 gave less frame gap/overlaps problems
2.13 gave a smooth playable file with audio in synch (watched a couple of minutes at the start of the movie, att the middle and a few minutes before the end)

Could I get a sample, please, which covers some of the gaps/overlaps? Thanks!

[Chumbo]

I was too slow to grab 2.15 before 2.16 came out. Would anyone be kind enough to provide a link to 2.15 via a site like rapidshare, mytempdir or other file hosting place please? Many thanks.

[rickardk]

Originally Posted by madshi

Could I get a sample, please, which covers some of the gaps/overlaps? Thanks!

Yes I will make a sample of Sum of all fears too.
I never done a sample of an evo. How is it done?

[nautilus7]

Originally Posted by Chumbo

I was too slow to grab 2.15 before 2.16 came out. Would anyone be kind enough to provide a link to 2.15 via a site like rapidshare, mytempdir or other file hosting place please? Many thanks.

Why do you want v2.15?

[madshi]

Originally Posted by rickardk

Yes I will make a sample of Sum of all fears too.
I never done a sample of an evo. How is it done?

Easiest way is to let EvoDemux rebuild and then abort rebuilding without deleting the file. A hexeditor would do, too.

[madshi]

Originally Posted by nautilus7

EDIT: I changed the video renderer in mpc from haali to vmr9 and now it behaves the same way like wpm11.

That's strange. I'm getting garbled output even with VMR9. This is the very same effect I also had with "The Searchers". However, when I tried to rip audio from the Million Dollar sample, I got a message "the track is dirty" from eac3to. Is it possible that the rip is corrupt?

The garbled output appears on my PC with the MS VC-1 decoder and VMR9. The Sonic VC-1 decoder doesn't even show any picture. It just freezes. Had exactly the same symptoms with "The Searchers". That makes now 2 HD DVD VC-1 movies which my PC cannot handle at all. Fortunately all other VC-1 movies worked just fine for me until now. Not sure where the problems with those 2 movies are coming from. Maybe the VC-1 stream is incorrect somehow and PowerDVD's VC-1 decoder just knows a way to work around that somehow?

[dburckh]

I'm try to get a pulled down raw VC-1 stream out of an EVO with eac3to. I tried the -demux option and it said "no format was specified for the demuxer..." I did manage to get an mkv to work with 2.15, which is not what I want. Any pointers? Is this fixed in 2.16?

Why:

My ultimate goal is to get a VC-1/AC3/TS (or M2TS) stream out of an HD DVD. There seems to be a lot better support for this with the commercial players and hardware units. If I can get the raw streams, I think I can mux it to TS with ffmpeg. I'll write a muxer if I have to.

Thanks for your help and your awesome tool.

[nautilus7]

Originally Posted by nautilus7

Well, with whole movie the problem is only seeking (vmr9). But with the sample playback is also a problem.

You missed this...

Originally Posted by madshi

That's strange. I'm getting garbled output even with VMR9. This is the very same effect I also had with "The Searchers". However, when I tried to rip audio from the Million Dollar sample, I got a message "the track is dirty" from eac3to. Is it possible that the rip is corrupt?

I tried something... I demuxed the vc-1 stream with evodemux and then muxed to mkv with eac3to. Result is playable (not the very begging though) but not seek able.

I also demuxed the audio with evodemux and run it through delaycut. The frames (e-ac3) 33 to 41 are corrupted.

Maybe the first seconds of the evo (original) is damaged/corrupted and they screw the whole file.

[Kumo]

how does the eac3to "speedup/slodown" feature work?does it stretch the intermediate wavs mono files(temporary encoded) or does it change the sample rate and re-sample to 44/48khz?is the result better than stretching the wavs mono files with an audio editor(like audition)?

[nautilus7]

Originally Posted by dburckh

I'm try to get a pulled down raw VC-1 stream out of an EVO with eac3to. I tried the -demux option and it said "no format was specified for the demuxer..." I did manage to get an mkv to work with 2.15, which is not what I want. Any pointers? Is this fixed in 2.16?

Why:

My ultimate goal is to get a VC-1/AC3/TS (or M2TS) stream out of an HD DVD. There seems to be a lot better support for this with the commercial players and hardware units. If I can get the raw streams, I think I can mux it to TS with ffmpeg. I'll write a muxer if I have to.

Thanks for your help and your awesome tool.

I don't know whether we are talking about the same error, but here's what i get:

C:\Tools>eac3to file.evo -demux
EVO, 1 video track, 1:16:47
1: VC-1, 1080p24 /1.001
Extracting primary video track...
The VC-1 pulldown remover didn't receive the format information.

C:\Tools>eac3to file.evo 1: file.vc1
EVO, 1 video track, 1:16:47
1: VC-1, 1080p24 /1.001
Extracting primary video track...
The VC-1 pulldown remover didn't receive the format information.

[nautilus7]

Originally Posted by Kumo

how does the eac3to "speedup/slodown" feature work?does it stretch the intermediate wavs mono files(temporary encoded) or does it change the sample rate and re-sample to 44/48khz?is the result better than stretching the wavs mono files with an audio editor(like audition)?

Re-samples back to original sample rate. Quality is very good as it takes a loooong time to complete.

[dburckh]

Originally Posted by nautilus7

I don't know whether we are talking about the same error, but here's what i get:

C:\Tools>eac3to file.evo -demux
EVO, 1 video track, 1:16:47
1: VC-1, 1080p24 /1.001
Extracting primary video track...
The VC-1 pulldown remover didn't receive the format information.

C:\Tools>eac3to file.evo 1: file.vc1
EVO, 1 video track, 1:16:47
1: VC-1, 1080p24 /1.001
Extracting primary video track...
The VC-1 pulldown remover didn't receive the format information.

That's the one. Sorry for the ambiguity, I'm at work and didn't have the actual error in front of me.

[madshi]

Originally Posted by dburckh

I'm try to get a pulled down raw VC-1 stream out of an EVO with eac3to. I tried the -demux option and it said "no format was specified for the demuxer..."

Hmmmm... Seems to be a bug. Will fix that in the next build. For now try adding the "-keepPulldown" option. I think demuxing should work then.

Originally Posted by dburckh

I'll write a muxer if I have to.

A well working TS muxer would be quite lovely...

[madshi]

Originally Posted by nautilus7

You missed this...

You're right, sorry.

Originally Posted by nautilus7

I tried something... I demuxed the vc-1 stream with evodemux and then muxed to mkv with eac3to. Result is playable (not the very begging though) but not seek able.

I also demuxed the audio with evodemux and run it through delaycut. The frames (e-ac3) 33 to 41 are corrupted.

Maybe the first seconds of the evo (original) is damaged/corrupted and they screw the whole file.

I do think that the first seconds of the EVO file are damaged. But I'm not sure why seeking doesn't work with the rest of the movie. I've checked the VC-1 stream in the sample you uploaded for me and there seem to be enough key frames in there to make seeking work just fine.

Have you tried remuxing only the 2nd EVO to MKV? What happens then?

[madshi]

Originally Posted by Kumo

how does the eac3to "speedup/slodown" feature work?does it stretch the intermediate wavs mono files(temporary encoded) or does it change the sample rate and re-sample to 44/48khz?is the result better than stretching the wavs mono files with an audio editor(like audition)?

Stretching the wavs with an audio editor probably results in something quite similar to what eac3to is doing. But a lot depends on which algorithms are used exactly. There are good and bad ones. I'm not an expert on which algorithms are the best, but those eac3to is using (based on r8brain) were tested to belong to the better ones.

[nautilus7]

Originally Posted by madshi

I do think that the first seconds of the EVO file are damaged. But I'm not sure why seeking doesn't work with the rest of the movie. I've checked the VC-1 stream in the sample you uploaded for me and there seem to be enough key frames in there to make seeking work just fine.

Have you tried remuxing only the 2nd EVO to MKV? What happens then?

I will try the 2nd evo right now.

(god, video/audio editing is trouble. But we wouldn't be here if we didn't like it. )


EDIT: 2nd evo (original & rebuilt with evodemux) results in 100% playable/seekable mkv with both haali and vmr9 renderers.

[nautilus7]

madshi, i cleaned the million dollar baby audio track and tried to process it with eac3to. The libav decoder crashed and probably this caused eac3to to crash too. I believe that shouldn't happen. Bug report

[madshi]

Originally Posted by nautilus7

madshi, i cleaned the million dollar baby audio track and tried to process it with eac3to. The libav decoder crashed and probably this caused eac3to to crash too. I believe that shouldn't happen. Bug report

Ah well, maybe eac3to could catch the crash. But it's really a crash in the Nero decoder. It's not much of a difference if eac3to catches the crash or not. The processing won't work, anyway. Cleaning up didn't help, obviously. Did you use "fix" option? That's dangerous. Better use "silence" or "skip".

[mmoore99]

What does Ratatouille video have to do with eac3to? Generally I have always recommend against demuxing video and then remuxing it (which is currently the only way to mux Ratatouille to MKV). At least for h264 this can be problematic. I believe it's a better approach to use the Haali Matroska Muxer which gets along without fully demuxing the video first.

Would you provide an example of how to do this?