I'm having audio syncing issues when converting an mkv file to mpeg 2 using AVStoDVD.
I have checked the necessary options and preferences not to add an audio delay however my result still has an audio delay.
Source mkv file - https://mega.nz/#!1c11gDSA!Jzfz6WodWuWlBphAtR5nyOf9TCcc9DVo8c2_Rj3MR4A
Mpeg2 encode - https://mega.nz/#!1F1GAYQZ!NtPNl-FOfpSc2teVqcLIoJKc6xBoj5SpWkTI4v4uvAk
Log file - see below.
According to the mpeg2 encode, the first 3 seconds are dramatically faster than the first 3 seconds of the source file.
I can't understand why there is a delay in audio. I've tried several things to resolve the issue - perhaps you can take a look and advise me how to resolve this issue please?
+ Reply to Thread
Results 1 to 17 of 17
Could be some "bad" DirectShow filters loading. Go settings->AviSynth and select "ffmpegsource" for both video and audio source filters.
@MrC - Thanks for pointing that out. I realised I didn't have the LAV filters set to default but I've changed that now. I will do some test encodes to see how things turn out.
Last edited by Imy; 15th Jan 2018 at 08:01.
The problem is resolved. I'm now getting encodes with perfect syncing. Thank you.
On a side note I have another question. I'm trying to join 2 titles using the 'Join Source Titles (AVS File)' option. This works if the files have the same parameters however I have 2 MKV files with different resolutions 1080p and 720p. If it's possible to overcome this please can you give me some code for an AVS file to overcome this problem?
As a workaround I'm using ConvertxtoDVD which handles the scenario well however I would prefer to use AVStoDVD.
nice to read you have fixed About joining source titles with different properties, you can try following steps, that require some AviSynth coding. Assuming you have 2 titles with following AviSynth scripts:
Import("D:\AVStoDVD\Lib\A2DSource.avsi") Video = A2DVideoSource("D:\Movies\Title1.avi", CacheFolder="D:\Movies\Temp", FrameRate=25) Audio = A2DAudioSource("D:\Movies\Title1.avi", CacheFolder="D:\Movies\Temp") Video = Video.ConvertToYV12() Video = Video.Lanczos4Resize(720,576) Audio = Audio.SSRC(48000) AudioDub(Video, Audio)
Import("D:\AVStoDVD\Lib\A2DSource.avsi") Video = A2DVideoSource("D:\Movies\Title2.avi", CacheFolder="D:\Movies\Temp", FrameRate=25) Audio = A2DAudioSource("D:\Movies\Title2.avi", CacheFolder="D:\Movies\Temp") Video = Video.ConvertToYV12() Video = Video.Lanczos4Resize(720,576) Audio = Audio.SSRC(48000) AudioDub(Video, Audio)
Import("D:\AVStoDVD\Lib\A2DSource.avsi") V1 = A2DVideoSource("D:\Movies\Title1.avi", CacheFolder="D:\Movies\Temp", FrameRate=25) A1 = A2DAudioSource("D:\Movies\Title1.avi", CacheFolder="D:\Movies\Temp") V1 = V1.ConvertToYV12() V1 = V1.Lanczos4Resize(720,576) A1 = A1.SSRC(48000) V2 = A2DVideoSource("D:\Movies\Title2.avi", CacheFolder="D:\Movies\Temp", FrameRate=25) A2 = A2DAudioSource("D:\Movies\Title2.avi", CacheFolder="D:\Movies\Temp") V2 = V2.ConvertToYV12() V2 = V2.Lanczos4Resize(720,576) A2 = A2.SSRC(48000) Video = V1+V2 Audio = A1+A2 AudioDub(Video, Audio)
Press OK and run the project. Take note that the output file calculation routine will always consider only Title 1, so the final output size will be larger than the estimated output size.
1. Does AvstoDVD support DVD menus with chapter pages?
2. I'm trying to downscale an NTSC 1080i source to 480p (NTSC DVD). What is the most appropriate downsize filter I should use to prevent jagged lines? Is there calculation that helps decide?
3. To de-interlace my source I uncheck the 'auto avisynth' > Enhance > De-Interlace(10). My source is a concert performance with vew few fast moving scenes. Can you advise what the best number would be to use please?
my suggestion is to keep the output interlaced, i.e. 480i. Let the modern TV/monitor hardware to do the conversion job, if needed. If you want to force de-interlacing, then you have to run your own testing to discover which are the best parameters. Best parameters are too much dependant on source footage and personal flavour.
I had wrongly assumed that interlaced material on a modern day flatscreen tv is a bad idea. On the other hand my Samsung TV display interlaced footage perfectly so maybe you are correct in that Interlaced material should remain interlaced unless very its very necessary to de-interlace.
If I provide a clip of the footage I am applying a resize filter to - Can you help recommend a suitable filter?
Don't know if it was mentioned, but resizing interlaced content isn't good idea. You have to first deinterlace it and resize it after this. So you end with progressive content. (when you change number of lines) But it is obvious, sorry to disturb you.
Here is the link to the footage I wish to resize to 576p. Any advice on the most appropriate resize filter would be greatly appreciated.
The footage is 50 fps, but every frame is duplicated, giving 25 fps unique interlaced frames 50 unique fields.
You could certainly deinterlace to 50p if this is what you wish; for dvd I would leave it interlaced and create 720x576
in an mpeg2 encoder. For other needs create a 50p output resolution your choice. To create the proper 16/9 aspect ratio
you could leave the height as 576 and set the width to 720 and make it anamorphioc so that it gets stretched during playback on the
computer. Or non-anamorphic at 1024x576. Note anamporphic may not work on other devices, PC only.
Encoded in Virtualdub filtermod with this script:
Video = FFVideoSource("C:\path\Glastonbury 1080i PAL(1).mkv", track=-1, seekmode=0,fpsnum=25, fpsden=1)
audio=directshowsource("C:\path\Glastonbury 1080i PAL(1).mkv",video=no)
Video = Video.blackmanresize(1024,576)
Video = Video.ConvertToYV12()
Audio = Audio.DelayAudio(-0.35)
Last edited by davexnet; 20th Jan 2018 at 20:49.
truly interlaced in last step or just set flag? To do it really interlaced seems to me be redundant and weird when already progressive.