This one was a real pain in the @ss to do.
Finally, the only acceptable quality was to use the lowres pal-plus
template at 120min/cd, CQ60.

Then I read in this forum the author Kwag recommends the (hires)
kvcdX3 templates, even at 120min/cd! After a little testing I'm
still not convinced. This is a real non-action movie but I can't get the
quality above the low-res template. (my CQ-value around 50)
This is my .avs:

---------------------
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\MPE G2DEC2.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\Gri pFit_preview.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\blo ckbuster.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\leg alclip.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\sam pler.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\flu xsmooth.dll")
LoadPlugin("H:\Warez\Dvdsoft\Encoders\Mpeg2dec\dus tv5.dll")

Mpeg2Source("E:\Video\Sweetnov.d2v")
LegalClip()
GripCrop( width=544, height=576, overscan=1 )
GripSize()
FluxSmooth()
SpaceDust()
Blockbuster(method="noise", variance=.4, seed=1)
GripBorders()
LegalClip()

Sampler(length=24)
## MPEG size = ((Total frames/MovieTimeInMinutes)/24) * MPEG sample file size * .98 ##
-----------------------------
Do I have to use the mpeg-1 or mpeg-2 template at 120 min/cd?
Are my filter-settings correct?

Please help.
Regards,

Try KVCDx3 again, but change your script to this:

Get the temporalcleanerold.dll over at doom9.
Now run your sample session to find your new higher CQ value

-kwag

how about this instead and takes less time:

Mpeg2Source("E:\Video\Sweetnov.d2v")
LegalClip()
GripCrop( width=544, height=576, overscan=1 )
GripSize()
SpaceDust()
NoMoSmooth(30,2,6,2,6)
TemporalCleaner()
GripBorders()
LegalClip()

I've been messing around with nomosmooth after spacedust and getting good results...

Yes, that's good too . But I suggested PixieDust() and TemporalCleaner() so that he could raise more the CQ value. With Space+Nomo or Space+Flux, the file size is much larger.

-kwag

The few tests I did with space+nomo then temporalcleaner were equal to pixie then temporalcleaner...

so I am saying to use all three

SpaceDust()
NoMoSmooth(30,2,6,2,6)
TemporalCleaner()


instead of just

PixieDust()
TemporalCleaner()

Then you can drop TemporalCleaner completely, because NoMo is a spatial/temporal filter, while TemporalCleaner is only a temporal filter
So for heavy noise sources, use PixieDust or FaeryDust, and for fairly clean sources use SpaceDust.

-kwag

FaeryDust uses temporal filtering only. It should preserve the maximum
detail. (Not really true yet with strong settings but will be
Recommended limit settings 1-5

SpaceDust is just the spatial filter from PixieDust. It is for those
who want the April version speed from Dust now It can improve the
visual quality quite a bit. However unlike PixieDust, It can only reduce
noise - not remove it. So the compression gains (and compressed quality)
will be less.

PixieDust is essentially FaeryDust + SpaceDust (with some twists). It is
probably the best to use for medium noise and those who shoot for
1-CD encodes. Recommended limit settings 3-8


But going by what you just said you wouldn't need to add temporalcleaner to pixiedust since pixie has temporal and spatial...

so what I did is almost the same thing...

pixiedust() spatial+temporal
TemporalCleaner() temporal

---------------------------------------------------

SpaceDust() spatial
NoMoSmooth(30,2,6,2,6) spatial+temporal
TemporalCleaner() temporal

I just have 2 spatials and 2 temporals
where you put 1 spatial and 2 temporals

Try it...

I think you'll like the results

Thank you for your fast replies,

One more question. When do I have to use the kvcdX3
mpeg1 or the mpeg2 template?

Do low-action movies or high action movies make any
difference whether you choose mpeg1 or mpeg2?

SpaceDust+NoMo+Temporal is too strong . I tried that before, and the result viewed on a HDVT is like if you have a constant fog on the screen. One spatial and one temporal should be enough, unless the material is very noisy

-kwag

Kwag,

I agree, but for a 1CD encode I couldn't tell a difference on a regular television using pixie(very long) then temporalcleaner vs space+nomo(much shorter) then temporalcleaner. Since I don't have an HDTV I couldn't say about the fog.

For me...
I only use spacedust and temporalcleaner and backup usually to 2 CD's since I rather have quality than trying to cram 2 hour and 3 hour movies onto 1 CD

Hi KingTuk,

Ok, I agree that for the "ultimate" compression ( squeezing the last drop ), it is a good choice
I'm currently doing a test run with Snollygoster's new "AviSynthesizer" using ALL filters turned on

-kwag

Why are you using such a high resolution? If you decrease the resolution could you not up the CQ?

Yes, but the lower the resolution, the more details you loose including sharpness. That's why my favorite resolution is KVCDx3 528x480, because it's closer to DVD resolution, but still far less pixels to encode. So it's the best compromise

-kwag

Ok, so if I use the higher res, can I drop Blockbuster? I've been doing CVD w/Flux and Blockbuster(noise) with great results.
Thanks

Yes. You don't need blockbuster at >= 528x

-kwag

Ok, great. I'm starting to use TMPEG more now, since it is faster doing a CQ encode than CCE is doing a 3pass encode but just need a bit of advice. Is the Grip resizer the best to use and what filters would I use to increase compressability the most ,DCTfilter, Mergcroma .....? without slowing the process down to much and not decreasing videoquality? Plus if I was to go to Avisynth 2.5.1 instead of 2.0.7 would I get a speed increase with TMPEG?
Thanks

Ok, after some testing, looks like the lower res is the way to go for a one cd encode. I'm testing with Dragon Fighter - 90min - NTSC interlaced. When using the KVCD3 >528x480 template I would need to use a CQ value of 53 which is unacceptable quality but when using 352X480 I can use a CQ value of 70 which looks good If I was going with a two cd encode then I would use KVCD3 because then the encoder would have enough bitrate to play with.
Oh, yeah, about the AVISYNTH version, I'm using 2.08 for now because I run into all sorts of problems with Moviestacker, ToK and DVD2SVCD with version 2.51

P.S. the encode took 16hrs w/CCE 3 pass and took 7hrs with TMPEG on a AMD 1000. That's using Bicubic, de-interlaceing, Fluxsmooth and Blockbuster. The TMPEG encode looks better.

would you agree with the above statements?

100 %%%%
I should have been more specific

-kwag