Im I Doing Something Wrong For Final Output.

[Master Yoda]

Just wonder what im doing incorrect or if theres something i can do to speed up my encoding to my final output.I have captured some footage and want to encode to mpeg2 using cce.Now i know what im doing on the cce part but for some reason when trying to encode my capture cce runs real slow... This is my script...

Avisource("my_family.avi")
Addborders(12,0,12,0)
Trim(184,73305)


Do i need to frameserve in vd and then have the source in the avs script as the vdr file instead?

As im new to capturing and encoding the capture to mpeg2 is it also normal to be needing to use 7000 as the avg to get good quality.?

[Boulder]

What codec did you use for the AVI video?

You should denoise the capture, it will then require less bitrate for good output. You might also have interlaced video, which requires more bitrate than progressive. See the capture guide for more details on both issues.

[Master Yoda]

Captured using huffyuv, i want to leave it interlaced as i will be watching the dvd on a tv.Even though its interlaced 7000avg still seems high.Dont think i need to denoise as i captured it using my tv card but the source was the s-video lead which was connected to the s-video output on the sky digital box. Although after encoding to mpeg2 i did notice a slight flicker/beating on the top part of the image, a little bit from the top.

[Boulder]

You probably do want to denoise, you can never have a perfect, noiseless capture. That + interlaced footage is why you need a high avg bitrate. There's a recent thread at the Avisynth usage forum regarding filtering interlaced sources while keeping them interlaced. Regarding the flickering, did you use the correct settings for interlaced encoding in CCE?

If you used "Predict median" to compress in HuffYUV, you'll get slow decoding. Predict gradient is somewhat faster.

[Master Yoda]

Yeah i did use Predict median, i will have a look at the thread in the avisynth board as i think i will encode the mpeg2 again as the picture didnt seem to be as good as it could be even though in the cce under bit allocation under the q.scale the back ground was green and the q.factor was under 5.

Something just didnt seem right, some specklyness.

Do i need to frame server like the guide said for encoding a capture to mpeg2, but then again i would'nt be able to youse and filters etc.

Is this the title of the thread you talk about ' Speed and compression of different denoisers '

[Boulder]

Originally posted by Master Yoda
[B]
Do i need to frame server like the guide said for encoding a capture to mpeg2, but then again i would'nt be able to youse and filters etc.

If you use Avisynth to frameserve, you can use the all the AVS filters you like, of course.


Is this the title of the thread you talk about ' Speed and compression of different denoisers '

Nope, but that's a good one too. My personal favourites for denoising analogue captures are RemoveGrain, RemoveDirt and TemporalSoften. There are dedicated threads for RemoveGrain and RemoveDirt at the Avisynth usage forum.

The thread I mentioned is here : http://forum.doom9.org/showthread.php?s=&threadid=86394 . See scharfis_brain's first post, it contains the proper method for treating interlaced material.