Avisynth source loading?

[yugurya]

Which source loading function should be used for better frameaccuracy and quality(without audio/video sync problems)? FFVideoSource/FFAudioSource or DirectShowSource? When i use FFVideoSource/FFAudioSource with Virtualdub i get this errormessage: 'ffaudiosource: seeking is severaly broken' when trying to save the video. I have the newest plugins installed, so what could be the reason for this? And how can you use the Layer function together with ImageSource(), and FFVideoSource/FFAudioSource(arguments in AudioDub() ) without errors? When i use VirtualDubMod(mpeg input supported) the output video has no audio sync problems, but with an .avs input in the latest version of Virtualdub(same sync problems happen also with the .avs+Avidemux combination output), the output file video/audio is out of sync is several areas and also there are 'fast moving scenes'. So how can i define the avisynth script the best to cause for the best output?

[manono]

MPEG2Source for MPGs, AVISource for AVIs, FFVideoSource for the rest.

[Gavino]

how can you use the Layer function together with ImageSource(), and FFVideoSource/FFAudioSource(arguments in AudioDub() ) without errors?

What errors are you getting? Post your script.
One point to note is that Layer() accepts only RGB32 or YUY2 input (not YV12), so you may need to convert your source to one of those formats before using it.

[yugurya]

When i use Overlay, i get this error after a minute of successful saving: 'Avisynth read error: FFAudioSource: Seeking is severely broken' :
A = FFAudioSource("video.mpg").ConvertToRGB
V = FFVideoSource("video.mpg").ConvertToRGB
C = AudioDub(V, A)
p = ImageSource("photo.png")
Overlay(C, p)

When i use Layer, i get this error when importing the .avs to Virtualdub: 'Avisynth open failure: Layer image formats dont match' :
A = FFAudioSource("video.mpg").ConvertToRGB
V = FFVideoSource("video.mpg").ConvertToRGB
C = AudioDub(V, A)
p = ImageSource("photo.png")
Layer(C, p, "add")

Look at the tutorial at this url: http://www.avidemux.org/smf/index.php?topic=4397.0
FFmpegSource2(FFAudioSource/FFVideoSource) is recommended instead of MpegSource. So, what combination to use, FFmpegSource2+Layer or something else?

[Gavino]

When i use Layer, i get this error when importing the .avs to Virtualdub: 'Avisynth open failure: Layer image formats dont match' :
A = FFAudioSource("video.mpg").ConvertToRGB
V = FFVideoSource("video.mpg").ConvertToRGB
C = AudioDub(V, A)
p = ImageSource("photo.png")
Layer(C, p, "add")

As I said above, it requires RGB32 input, so use ConvertToRGB32().
You also need pixel_type="RGB32" in your ImageSource call.

Look at the tutorial at this url: http://www.avidemux.org/smf/index.php?topic=4397.0
FFmpegSource2(FFAudioSource/FFVideoSource) is recommended instead of MpegSource.

Reading that link, it says "For MPEG-2 and VOB files "DGDecode" (DGIndex) is the preferred source filter." Actually, the source filter is called MPEG2Source (as recommended by manono), DGDecode.dll is the name of the DLL containing it.

[manono]

Look at the tutorial at this url: http://www.avidemux.org/smf/index.php?topic=4397.0
FFmpegSource2(FFAudioSource/FFVideoSource) is recommended instead of MpegSource.

That's not the way I read it, not when it says:

For MPEG-2 and VOB files "DGDecode" (DGIndex) is the preferred source filter.

The DGDecode.dll works together with MPEG2Source.