Avisynth-Tweak color-question

[SerbianBoss]

I capture VHS in DV avi files using canopus. So when make avisynth script i need to correct colors. Because my source is dv avi interlaced and make dvd with CCE(want to stay interlaced) where to put Tweak filter in this interlaced script. Is better to put before or after ConvertToYuY2 or elsewhere?

setmtmode(2)
AVISource("C:\Documents and Settings\test\Desktop\test.avi")
ConvertToYuY2(interlaced=true)
SeparateFields()
odd=SelectOdd.Convolution3D (1, 32, 128, 16, 64, 10, 0)
evn=SelectEven.Convolution3D (1, 32, 128, 16, 64, 10, 0)
Interleave(evn,odd)
Weave()
crop(8,4,-8,-12)
AddBorders(8,8,8,8)

[SerbianBoss]

Any comment?

[vhelp]

Hi SerbianBoss

I would do any color adjustments first, before any other color space conversions.

As long as the color() or RGB() filter (or whatever color filter(s) are used) are
designed algorithemically correct, and then apply your convertToyuy2() or ..yv12()
afterwards.

The general idea is to do something like this at first oportunity. That would be
just after the AVISource() function. The reason for this is because any conversion
applications will Remove or Reduce image detail. And you want to give your
conversion routine(s) every bit of image detail to work with from the start.

-vhelp 4393

[SerbianBoss]

So its better to put after avisource instead after convertoyuy2()

Because this is interlaced material(dv avi) does is maybe better to put tweak filter insade separate fields/weave?

[vhelp]

So its better to put after avisource instead after convertoyuy2()

Because this is interlaced material(dv avi) does is maybe better to put tweak filter insade separate fields/weave?

In this case, no. You can prob put in sep/weav fields or not. Doesn't matter.

Part of the purpose of the sep/weav field (filter) method is to gain better ore more
optimized image processing, where each field get filtered (ie, degrain; temporal; spacial;
NR [noise reduction] processing) Sometimes it helps the image to come out better in terms
of NR or whatever the filter was designed to turn the output into. So, adjusting color
levels in your case is not realy needed for this type of order of sep/weav routine.

I mainly use VirtualDub because my work is mostly in RGB color space, plus my mpeg
encode of choice is TMPGenc, which processes filters in RGB color space.

DV-- When I capture in the DV color space format, [ie, my ADVC-100] I use the
cedocida v0.1.7 codec, and I set it up for TMPGenc handling by using the following setup:

Enabled output formats: [x] RGB
YUV to RGB conversion: (o) [0..255] --> [0..255]
*YV12 chroma sampling: (o) MPEG 2 interlaced

If I don't set it up this way, when I encode the dv source inside TMPGenc, it will mess up
the color levels slightly .. may not be as noticable for most people, but for me and my
acute eyesight, I see the difference is off .. so, I use the above setup on my DV source
and for TMPGenc usage and the color levels (after DV->MPEG) are as good as the source,
imho.

Analog Capturing-- Because my (new) computer set is working better than usual,
and my analog capture card seems to be working properly (no more line noise and/or hum
bars (audio induced noise) so far, keep fingers crossed) I like to capture in uncompressed
color space for the most reproduction in image detail. DV is great (specially in my home
and unusual circumstances with my noisy (vhs/cabletv) sources) but it goes so far in the
image detail area. When the source is too noisy, you can spot macroblocks or pixelation.
This is usually seen in poor vhs recordings from analog cabletv. And this is harder on the
mpeg encoding. And if you used a low bitrate (ie, to fit 2 hrs on a signle layer [4.3g] disc)
with this type of source, the finished result will reveal pixelation. I'm too picky and in most
cases, I can easily spot them even without looking for them -- and this annoys me.

So, with noisy sources (as above) I've been putting up with DV all this time. But now that
I have my Analog Capturing (capture card) for uncompressed capturing settled, I can now
finally enjoy the full (ok, almost full) capturing and processing and encoding to mpeg projects,
though still keeping my fingers crossed with regards to the line noise and hum bar (audio
hum induced noise) issues in my (new) computer setup, ohh, and using my older and finally
used, Pinnacle Studio AD/DV analog capture card with VirtualDub.

The reason why I went a bit off was to help you to understand some of the issues with
noise in captured sources and how I dealt with them, plus I gave you some personal
insight in mine and how I fixed them.

Good luck in your capture projects.

-vhelp 4394