It's true that I also use the last CVS binary (that is not the official delivered one). The official version is 12-april, the last one is 15-05 (or something close).

In YV12 color space, MergeChroma(Blur()) and MergeLuma(Blur()) are nearly timeless ! (something like 5 sec more for a total encoding time of 10 minutes).

If you're looking for the latest binary, see sh0dan's signature :wink:

For the new script update, do you recommad to totally remove STMfilter? Amazing work guys :wink:

//Wolfi

I wouldn't drop it just yet :!:
Maybe lowering the values more will still give us a little file size saving with some spatial filtering :idea:
I'm doing a couple of tests right now...

-kwag

Done :!:

Changes are STMedianFilter(8, 32, 0, 0, 8, 32)
I've updated the current script. Please test it! I can't see ANY visual difference with or without STM in my monitor :!:, but the file size is still smaller with the use of STM and the new values.

Edit: @muaddib
Please take note for your next update to moviestacker, and move the :arrow: "unfilter" line "after" the LegalClip line :!:

-kwag

I do not agree with you. There is a 2.5 version that was given here :

http://www.kvcd.net/forum/viewtopic.php?t=3544
(latest post, done by latexxx).

I can also send it to you if you want.

That exactly what I suggested to muaddib for his 2.5 version of MoviStacker :-)

You can also go to http://www.avisynth.org/~warpenterprises/ and see if there is a 2.5 version of the plugin you need.

Dang..I should really start punishing my brain :twisted: Thanks for the correction.

I guess I'm moving on to AviSynth 2.5x in the very near future... :wink:
So what do you recommend using instead of TemporalSmoother in AviSynth2.5x Boulder :?:

Maybe a light TemporalSoften would do?

I did some samples with and without STMedianFilter and I totaly agree...
I watched the samples on TV and I couldn't see any difference but
STMedianFilter brought down my Fullsample size by ~600kB (which would
be ~36 MBs for the full movie...quite a lot I think) :wink:

...so my STMedianFilter isn't burried yet :wink:

@jorel: great job, thanks for your hard work :D

reards, ***mfb***

hey I'm with you on this!! :D I only use Fluxsmooth or TemporalSmoother too :lol:

ren

8O Wow, great new script guys... I added an additional FluxSmooth(5,5) line between STM and TemporalSmoother...
kwag, jorel you should try that. FileSize drops again, and the picture looks the same as without it at the first test...
will you guys do some tests too? it's hard to tell for me... !!

I agree. I've been using it for a while now and haven't had any trouble with the latest build. It's quite stable now, IMO. And with loadpluginex.dll, there is really no reason to stick with 2.08 anymore.

If all else fails, you can always load the old 2.0 mpeg2dec.dll and use it with a colorspace conversion, if necessary. That's what I'm doing currently. The colorspace conversion is a little awkward, but I haven't noticed that much of a hit from it.

Yes, I've been using undot on dirty DVD sources with great success. I like to put it pretty high up in the filter chain, like right after the resize. It takes care of a lot of grain and noise in preparation for more powerful noise reducers like pixiedust(limit=3).

I can't speak for Boulder :lol: but I have not yet found a 2.5 temporal filter that matches temporalsmoother. I've tried atc, temporalcleaner, and temporalsoften, and all leave a lot of noise that temporalsmoother cleans right up.

I'll personally start using AviSynth 2.5x when it's tightly integrated with MovieStacker. Right now, it's not, as most of the filters are 2.0x :roll:
Right now, MovieStacker works seamlessly with 2.0 :!:, but not with 2.5x

@muaddib,
Is it time to cut over to 2.5x in MovieStacker :idea:

-kwag

[quote]Better of why not add 2.5 to the allready existing 2.0 with the Avi switch plugin, :D more work for Muad. :wink:

If it works as smooth as MovieStacker/AviSynth 2.0x, hey!, go for that :D

-kwag

so are you guys suggesting to go to Avisynth 2.5 now? after all the bad things I've heard? or am i not hearing correctly?

You can use TemporalSmoother as it is included in Mpeg2Dec, and a 2.5 version of this filter exist.
(but just use Mpeg2Dec for that, let use Mpeg2Dec3 for reading your Mpeg2 source, its far better and faster)

I remember Kwag saying "the lack of Gripfit for AVS2.5 is the point that stick me to 2.0". And I thought "He's right" (I always find Kwag right :-)).

The day I saw the 2.5 version of GripFit, I run on AVS2.5 :-) :-)

Yes, but IIRC TemporalSmoother only works in YUY2. If you convert to YUY2, there's really no need to use AVS2.5x as you'll also lose the speed advantage. There is a quality loss each time you do a colorspace conversion and God knows how many times TMPGEnc converts :evil: They really need to add direct YV12 support since MPEG-1 and 2 are YV12!