RePal Artifacts

[Piper]

Is there a good way of dealing with dot crawl type artifacts introduced by functions like RePal? I'm assuming this is a well known side effect of scripts like this. The effect is random dot crawl (is that a contradiction?), on thin horizontal lines. The bottom row of pixels is always affected by this. My guess would be that it's an undesireable result of alternating field orders?

I've tried DeDot, DeSpot, Guavacomb etc, but I'm not happy (yet) with the results. Perhaps somone's tweaked some Guavacomb settings to help with this kind of artifacting?

I can post some screenshots/avi's if needed. The source is analogue NTSC.

Thanks

[Piper]

Well it seems after long last I can finally answer my own question. The recent Doctor Who Christmas Invasion episode provided another opportunity to refine my script.

The effect I originally described is greatly influenced by what bobber is used and may not be a product of RePal at all. After many comparisons with various bobbers, I stumbled across tryweave in TDeint. Using it improves stability and greatly reduces the artifacting on horizontal lines.

TDeint(order=1,field=1,mode=1,type=2,tryweave=true)
RePal(quality=3)
AssumeFPS(25,1,true)
SSRC(48000)

Glad to finally put that one to rest! :)

[Boulder]

Btw, did you know that Restore24 also contains a "RePAL" mode? The point in using it is that you don't need AssumeFPS as R24 will output 25fps natively.

[Piper]

Yes, I have tried restore24, however at the time RePal produced better pattern matching & smoother motion on this particular source. It's because of this that I continue to use RePal rather than restore24 though I freely admit I haven't actively followed restore24's progress since about the time I originally posted to this thread.

EDIT: Clarification