I have a lot of audio editing experience, but don't know much about video quality settings, I'm replacing the audio of a video with new audio which I've speed corrected, and made various other edits, and have it perfectly in sync with the old audio it's replacing, but don't know which programme or which encoders / settings to use for highest possible quality...
I want to export the new video/audio at the highest possible quality (as close to lossless as possible). PAL format.
Is there a lossless way of doing this?
This project is using an extremely rare video, so once I'm done syncing the new audio to it I want the end result to be as lossless and high quality as possible!
Thanks in advance.
+ Reply to Thread
Results 1 to 24 of 24
Last edited by efc1978; 1st Jun 2015 at 02:33.
For best quality don't mess with the video at all and simply replace the audio, PCM of course.
Originally I had hoped to do exactly that, but couldn't find a way and ended up thinking it wasn't possible..., it would be brilliant if I could!
Can you please tell me a way I could just add the new WAV PCM audio files into the video without changing the video?
I think I have done this correctly, I followed 2 steps on this page here, using PGCDemux and MuxMan:
I downloaded PGCDemux and separated the video from audio (using the exact same settings as seen in the PGCDemux picture), and exported a .m2v.
I then loaded the .m2v into Muxman and simply added my new WAV PCM audio track, and exported (there didn't seen to be any quality/settings adjustments I needed to make in MuxMan?).
So other than tick/untick some boxes to make the settings exactly the same as seen in the PGCDemux picture at the link I gave I have not altered anything else on either programme....
Is this all I needed to do to keep the original video lossless?
Thanks so much!
Last edited by efc1978; 1st Jun 2015 at 01:16.
As I have 6 different .VOB's from the same video I wish to replace the audio in individually (so 6 different times), how do I combine this all back into 1 DVD?, I can't figure it out, as I am doing each separately each output VIDEO_TS folder has the following:
How do I instead make this a single DVD?, as my little knowledge on video tells me a whole DVD should only have just one VIDEO_TS.BUP and just one VIDEO_TS file, not one of each for all 6 .VOB's I'm editing...
I have tried loading each of the 6 files and adding the 6 separate audio tracks into MuxMan at once, but once exported I keep only getting the first of the six...
Geez, you edited the audio, once for each original VOB file? If so, that was totally unnecessary.
Anyway, In the audio section of Muxman is an 'ADD' button. It's there you can add in your individual audio files so that the complete audio is authored for DVD along with the M2V. I'll assume the M2V you got from PGCDemux was for the complete video.
For future reference, extract the complete video and audio before going to work editing the audio. Also, unless the DVD audio was WAV audio to begin with there's a chance it won't author because of your edited WAV audio having a much higher bitrate than the source audio (AC3, MP2?). If that's true, then reencode your WAV files to AC3 audio. There's also a chance that by adding your individual WAV files to Muxman you'll wind up with out-of-synch audio after authoring. Never know till you try.
Is this all I needed to do to keep the original video lossless?
OK I have it sorted now, I wasn't loading all into "Import File List", I was loading each file individually repeating steps on the main interface!
Now I'm getting it to all combine as 1 DVD, so this should be all lossless video quality?, as I used the same settings I see in the pics at the link I put a couple of posts ago....
Thanks for the help!
The reason I edited audio individually is that my DVD is a combination of songs from a live concert combined from multiple DVD's, each of which contained different songs of the same concert, mostly in the wrong order they were performed live, so originally I used "DVD Shrink" to cut and then put in correct order of performance each song individually, and then finally combined the different DVD sources into 1 DVD, as I'm pretty sure DVD Shrink's "No Compression" setting does that losslessy.
Then I edited together the new audio, from a cassette tape, which I had to speed correct for each individual song, so that's why I edited the audio for each individual song....
Last edited by efc1978; 1st Jun 2015 at 03:16.
The reason I edited audio individually is that my DVD is a combination of songs from a live concert combined from multiple DVD's, each of which contained different songs of the same concert, mostly in the wrong order they were performed live
By the way, if it did have chapters for each song, when you used PGCDemux to extract the complete video you should have gotten a 'Celltimes.txt' file at the same time. It contains the list of chapters by frame number. You can add that when authoring with Muxman in File->Import Chapter and get the same chapters as before. Without doing that you get none. Maybe you already figured that out. Or maybe the songs are all separate titles rather than separate cells.
Thank you very much to all for your advice and help!
When PGCDemux splits the VOB's ripped off my DVD into video and audio, the audio file is a system signed, 16bit, 48000, 1536kbps .mpa file, however I know that DVD audio can't really be .mpa, is there a reason this is happening?
Previously when experimenting with other programmes I saved the audio from the same VOB's using AVIDemux which saved the audio as system signed, 16bit, 48000, 1536kbps .WAV, so I'm not sure why AVIDemux saved as .WAV but PGCDemux saved as .mpa (which seems odd as DVD audio can't be .mpa).
Though I'm pretty certain any difference in sound quality between the 1536kbps .mpa file made from PGCDemux and the 1536kbps .WAV file made from AVIDemux will be inaudible, unless there's some kind of exception for .mpa? (I really don't know much about .mpa).
PGCDemux didn't change it in any way. AviDemux may have converted it. You can check the VOB in MediaInfo or GSpot to confirm (or disprove) what I just wrote.
I've split my DVD .VOB's using PGCMux to separate the audio and video, the split out video is .m2v, and the audio .mpa.
I have now finished editing the 'new' audio I plan to use, when I first edited the audio I had it as WAV PCM 16bit 48000 1536kbps (as I wanted to keep everything lossless), and when I tried to add the new audio back to my video by combining it with the .m2v file in Muxman I got the error message : "multiplex operation failed - probably caused by excessive bitrate", so I then converted those WAV's in Audacity to 386kbps .MP2's and am still getting the same error...
What could be the problem and how do I fix it?
Thanks in advance for any help!
That's odd and I don't know for sure what's wrong. One would think if it was a 'compliant' DVD to begin with, by demuxing and remuxing it would still be a compliant DVD. Maybe it was originally authored by an authoring program not as strict as is Muxman. Muxman is very particular about the max bitrates.
Anyway, a couple of things you might try. One is using IFOEdit to do the muxing. Up at the top go DVD Author->Author new DVD. I think you can add additional audio tracks (never tested it). IFOEdit is much more 'forgiving' of bitrate spikes. It will almost certainly complete the authoring, but might give out with a warning when done. Or you could try any other authoring program with which you might be familiar. You could also test the compliancy of the original DVD by reauthoring the original M2V and complete MPA audio to see if Muxman completes without an error. My guess is it won't.
The other thing, as I don't think the problem is the audio since you reencoded it at a lower bitrate, is to try and get the video differently. Maybe open all the original VOBs together in DGIndex and go File->Save Project and Demux Video. But that probably won't fix it.
I haven't tried any of the suggestions from the last post yet as I did not work on this project over the weekend or today (Monday)...
I will have time again from today....
OK I have remembered something which seems a bit odd...
As I told you in my last post I'm getting the error about bitrate when trying to add the audio back to the video..., however I now seem to rember that it was previously working, as I'm pretty sure when I tested it out some days before it was working, I could be mistaken but I really do think I remember it working - supported by the fact that earlier in this thread I posted asking how to combine a bunch of VOB's and as I stated it was only exporting the first VOB out of the whole bunch (as I was incorrectly loading each VOB separately into the scene list not by correctly using add), and I do remember loading all 6 VOB's a few times and every time it would export just the first VOB, so then as you can see in my previous posts I posted here to ask why it was exporting just the first VOB, so yeah I'm pretty much certain it previously was working!!!!
Are there any settings I could have changed which would be causing it to not work?, I can't see anything which seems would have any effect.....
Oh just as I am writing this I remember the other day when it worked I had a few other programmes downloaded onto my computer which were suggested on the MuxMan download page for fixing out of synch audio, which I have now deleted, maybe I should not have deleted them?, but I didn't think they were necessary for spitting the audio and video from a VOB and combining new audio to the video, as I didn't open those programmes when it worked, the programmes I have since deleted were Media Player Classic, Vobblanker and DelayCut.
Maybe vobblanker assisted in helping it previously work even though I didn't directly open that programme????
Are any of those necessary to have for it to work?, could that be the problem?, I think pretty unlikely, but there's got to be a reason for it previously working..... **come to think of it I have re-edited the the audio files also, but didn't do anything different to the ones that worked....
Last edited by efc1978; 8th Jun 2015 at 08:52.
Muxman might not be working? If so, Muxman doesn't take VOB input.
Are any of those necessary to have for it to work?
Another thing you might try, just a as a test, is to author the M2V alone. A DVD doesn't need audio to be a DVD. Just to confirm whether or not there are big bitrate spikes that prevent Muxman from authoring it. If it also errors out, the program used to author the DVD originally didn't produce DVD-compliant output. I'm inclined to think that's the case anyway and you'll have to use a more 'lenient' authoring program yourself, one such as IFOEdit. It won't error out unless there are a lot of bitrate spikes.
Worked it out....
Last edited by efc1978; 10th Jun 2015 at 19:31.
Just a couple of questions about some results I got doing a test/experiment to see how much PGCDemux and Muxman changed the video and audio files....
So I have 6 different .VOB's from the same video (each is actually 1 song), using PGCDemux I separated the audio and video from each of the 6 .VOB's, so I had a .m2v and .mpa file for each of the 6 .VOB's, I then combined each .m2v and .mpa back together using Muxman so I once again had 6.VOB's, and then for a second time used PGCDemux to once again split the video and audio.... (So I had split video and audio out of a .VOB, then combined the output files back together again and then once again split them apart again).
What I found was that the final .m2v files for all 6 VOB's were only 3 or 4 bytes different compared to the original .m2v eg. 147, 224, 344 bytes and 147, 224, 348 bytes. As this should be lossless I'm wondering what the change in file size would be?, I know it's incredibly minuscule in size, but just trying to understand, as it' supposed to be lossless I expected the 2 to be identical in size..., what has changed here?
Secondly, something I am actually a little concerned about as it could have an effect on how the audio syncs with the video....
Using the same experiment/test process as described above, I found when comparing the .mpa audio files 2 out of the 6 final .mpa's had the ends slightly chopped off, the .mpa from the 4th .VOB was 0.024 secs shorter than it originally was and .mpa of the 5th .VOB was 0.048 seconds shorter than it originally was, extremely tiny amounts but seeing as I use the time lengths of the original .mpa's to sync my NEW audio speed to this makes me wonder why this has happened?, and why only happened to 2 of the 6 files I experimented with?
Last edited by efc1978; 10th Jun 2015 at 04:45.
Easy answer - 'don't know, don't care'. Why beat your head against a wall, or waste time, over trivialities? Maybe someone else will humor you with your anal retentiveness and try and help. Good luck, and I'm retiring from this thread.
Like I said, I use the original .mpa audio to exactly speed/pitch match the NEW audio file I edit into the video, for example in the NEW audio file I make the snare drum start at the same time to 0.01 secs as the original audio .mpa, which properly syncs it, now as I have found that PGCDemux cuts of various lengths of the original audio file (as I have seen it cut off at least up to .048 secs) this most certainly isn't trivial as you put it, as watching a drumstick hit a snare, or pick a string, which is .05 secs out of time because the file has been truncated makes it look visibly out of sync compared to something which is .05 secs closer to being in sync.....
The method I'm using is a perfectly accurate and easy way to sync audio/music to the picture, if you think it's trivial you just don't get it...
No need to be so rude, you sound old and grumpy, do you have bad ears/eyes?, maybe that's why you can't see if your audio is .05 out of sync! hahahah
I'm retiring from your forum, bite me pal!
Last edited by efc1978; 10th Jun 2015 at 19:45.
Delete my account please.
Sequence headers and metadata can change when you demux / remux video. Not just with those programs, but in general with other video formats, other containers as well. ie. it's common to get a very minor filesize difference of a few bytes.
"Lossless" in this type of A/V content is defined by decoded image or audio is bit for bit identical. Not necessarily those other things like metadata, sequence headers etc... Absolutely everything is only preserved when you use archival software like winrar, 7zip etc... Decoded image or decoded audio isn't affected when you demux - at least it shouldn't be. The way to verify that is decode to uncompressed audio or video and test with various methods like PSNR etc... it should be bit for bit identical to source.
As for the other problem of cutting audio off...not sure I don't use PGCDemux much. First, how are you determining that? What programs ? You can also check with DGIndex to demux and check (is there a problem with your playback/analysis method, muxing or demuxing, or something else? )
It's incomprehensible that two individuals end a post this way. Both have a low tolerance.
The original post will probably go elsewhere for his answer. Way to go!!!!!!