Problems with video, differing weird frame rates reported.

[jagabo]

There is only a fraction of filters that are available in native 64-bit.

Yes, that is why I still use 32 bit video tools. And there's no big advantage to using 64 bit anyway.

[poisondeathray]

There is only a fraction of filters that are available in native 64-bit.

Yes, that is why I still use 32 bit video tools. And there's no big advantage to using 64 bit anyway.

I would have agreed with you 2 years ago, but there are big advantages for other video programs like 3D rendering, Adobe apps (They only run 64-bit)..., NLEs, memory intensive programs like compositing apps with big textures. Even x264 64bit is consistently 5-10% faster now test on a variety of conditions, sources when used directly or piped (Even without a memory bottleneck and huge amounts of free RAM)

[jagabo]

There is only a fraction of filters that are available in native 64-bit.

Yes, that is why I still use 32 bit video tools. And there's no big advantage to using 64 bit anyway.

I would have agreed with you 2 years ago, but there are big advantages for other video programs like 3D rendering, Adobe apps (They only run 64-bit)..., NLEs, memory intensive programs like compositing apps with big textures. Even x264 64bit is consistently 5-10% faster now test on a variety of conditions, sources when used directly or piped (Even without a memory bottleneck and huge amounts of free RAM)

I was speaking about AviSynth (and related tools around it) in particular.

[poisondeathray]

oops I though you meant video tools in general

But there are cases where moving avisynth completely to 64bit would be beneficial. Right now , running a few avisynth filters on HD footage you commonly run into memory bottleneck issues.

[poisondeathray]

still no luck with lavf?

Another way you could do this that would be independent of your directshow filter configuration is to use x264 to encode to lossless intermediate. If you download a binary from komisar , jeeb or taro, they lav built in and you can force it by using --demuxer lavf (I think jeeb builds the ones on x264.nl, or you can use this http://komisar.gin.by/ )

x264k --demuxer lavf --crf 0 --preset veryfast --fps 30000/1001 --tff -o lossless.mkv split-006.mkv

Then open "lossless.mkv" with ffms2() and apply your sd reinterlace script. You shouldn't need another lossless intermediate go round, this is ok to input into HCEnc

Load_Stdcall_Plugin("C:\Program Files (x86)\AviSynth 2.5\plugins\yadif.dll")
FFVideoSource("lossless.mkv")
AssumeTFF()
AssumeFPS(30000,1001)
Yadif(mode=1, order=1)
Spline64Resize(720,480)
SeparateFields()
SelectEvery(4,0,3)
Weave()
ColorMatrix(mode="Rec.709->Rec.601", clamp=0)
ConvertToYUY2(interlaced=true)

[Killer3737]

Ok I just recieved a sample of the original file and I'd surmise the MKV was borked. The TS sample is fine, I ran it through DGAVCIndex without a problem which reports that it's MBAFF, probably why my output wasn't right. Although saying that Mediainfo doesn't report it to be MBAFF. I know from a previous thread I made that AVCSource doesn't support MBAFF though so back to DSS?

Here is the sample if anyone else wants to take a look - http://dl.dropbox.com/u/63805455/Foo%20Fighters%20-%20Live%20At%20Lollapalooza%20Brasi...ts.12.ts.12.ts

[poisondeathray]

DGAVCIndex has problems with PAFF streams. It will decode with macroblocking and artifacts. It's no longer developed or supported for a few years now. Use whatever works