hi Kwag,

The reason why I just use tok to predict the CQ and not totally do all the process is because I need to enable Deinterlace(Double) filter on TMPGenc. Since you were able to give me the config file to automatically set motion precision and detect scene change, is it also possible to automatically set TMPGenc Deinterlace(Double) filter.

By the way, I just did a sample clip with the above step and finally I was able to encode at higher resolution (480 lines) without the jumpy and combing problem.

So if there is a way for tok to enable the deinterlace(double) also thinking of the using Noise reduction as well(but probably not a good idea as this will take me about 24 hours to encode a 55min movie)...then I can just use TOK for everything....

is there a way?

Thanks once again.
Dredj

Have you tried using bob() or any of these: http://www.avisynth.org/index.php?page=DeInterlacing in your script instead of using TMPEg's internal deinterlacer

-kwag

HI Kwag.

Okay looks like i'm getting somewhere....I also found how to enale deinetrlace(double) on tmpgenc using from the video.en1 file, but i didn't do that instead I use some other deinterlacing filter....

I currently use SmootDeinterlace() and it looks like this is what I need, the output clip is not jumpy. Now before I finally encode the whole video, I have one more verification that I would like to ask:

It is okay if I do this in my filter script:

AviSource("G:\Capture\Barney\inconcert.avi")
ConvertToRGB32()
SmoothDeinterlace()
ConvertToYV12()
undot()
Limiter()
asharp(2, 4)
GripCrop(528, 480, overscan=2, source_anamorphic=false)
GripSize(resizer="BicubicResize")
STMedianFilter(8, 32, 0, 0 )
MergeChroma(blur(MaxTreshold))
MergeLuma(blur(0.2))


Is it okay to do COnvertToRGB32() first and then SmoothDeinterlace and then COnvertToYV12() ? The reason why I did this is because smoothdeinterlace only supports RGB32 or YUY2, since the capture from DV is RGB24, I have to convert it first to RGB32, in order for the deinterlace to work. I didn't choose YUY2 because it turns half of the pciture to pink color. and lastly I have to convert to YV12 because the rest of the filter does not support RGB32.

If the script is okay then I'll guess this is it.....finally....finger cross...


As always....thanks....
dredj

Okay I just want to add some updates:

1. When using SmoothDeinterlace at 24% of encoding, it took 15 hours...so I cancelled it, also the CQ is only at 64

2. This time using tok also and enable TMPGenc built in Deinterlace(Double), at 26% of encoding, it took only 4 hours....and CQ is at 72.

Don't want to conclude yet...I will try some more testing before I decide which one should I use.

If the result of my first kvcd is very nice, I'm planning to write a guide for converting VHS to KVDx3, if that's okay with u...but the problem is I don't know how to be able to post it here since I will be putting some drawings on it.

Anyway that will be my next goal...for now more testing....

Regards,
Dredj