Regarding arrangement of plugins..

[zanaitoryoushi]

..................................................

[poisondeathray]

Yes order matters. They are applied sequentially

Genarally if you have a noisy source, applying sharpening before denoising will enhance/sharpen the noise as well as the details you intended to sharpen . This will yield in larger filesizes

There is no "correct order" for all cases. It depends on what you're trying to do specifically.

Comment out "#" lines in the script to see sequentially what each filter is actually doing

Learn to use Avspmod and preview your scripts . You can use multiple tabs and different versions of scripts to compare (push number keys to quickly toggle between them)

Finally, "quality" is VERY subjective. If this is still your youtube source, many people might think you're probably overfiltering the source . Some people like to preserve noise and grain. Some people like it squeeky clean. Some in between etc....

[zanaitoryoushi]

...................................

[poisondeathray]

edit: also, i guess since i'm practicing on converting Blu ray clips that are available on youtube, unfilter should come first then denoise then smooth..just saying, because so far, all sources i've read about using this plugins aim to enhance a low quality source, but since i'm doing mostly bluray clips right now, if i apply their theory in reverse i think my results will be better, any opinions on that? i think it's plausible somehow...

Well they aren't really blu-ray clips - anything on youtube will be quite low quality compared to the actual blu-ray

It depends what's wrong with the clip specifically, and what you think needs to be adjusted (many people will think differently on what needs to be adjusted or if anything at all)

[zanaitoryoushi]

............................

[poisondeathray]

I think they are blu-ray preview clips, not taken from the actual blu-ray (which isn't released yet) . Besides youtube re-encodes them to about 5-6Mb/s, the actual blu-ray will usually be around 25-35Mb/s . The real clip will have fine grain and detail that the youtube clip wouldn't preserve

Shouldn't you be polishing ShellHead's armor instead of encoding clips ?

[zanaitoryoushi]

Shouldn't you be polishing ShellHead's armor instead of encoding clips ?

LOL i like that..can't stop laughing when i read that, haha lol

edit: thanks a lot for the help bro, take care...going to polish that armor now, haha
see you next time...

[Cornucopia]

There are always unique situations which require changing the "formula" for how to do these things, but in general, for restoration type work, a good rule of thumb is (and this applies to audio & video):

Fix framerate/samplerate problems first - get your timebase right
Maximize existing dynamic range early in the chain
Next, remove noise & artifacts (surgery in stages is better than bruteforce)
Later, enhance/sharpen

Don't be afraid to use parallel processes or sidechains. Sidechains which use different levels of noise/resolution thresholds are particularly effective.

So:
STRAIGHTEN - REMOVE - REMOVE - REMOVE - ADD - ADD...

Scott

[zanaitoryoushi]

.............................

[Cornucopia]

On the surface, I'd say only use Tweak AND Smoothlevels if just one doesn't have the full controls necessary (they kind of overlap/duplicate their work). Also, it would make sense to me to put dfttest before deen. But this is all WITHOUT having seen any of the before/after video. Plus, I only use AVISynth occasionally, so maybe I'm not the best reference. Otherwise, the order looks fine.

These things are best approached from Doctor's point of view. Diagnose, Prescribe, Treat. With each obstacle being at least one task (though some task treatments influence others).

Using waveform monitors/vectorscopes really does help also, as then you are getting objective, statistical data to back up your decisions. They help clarify what's really going on. Plus, as suggested above, solo (isolate) each process to see what it's specifically doing irrespective of the others.

Also, just reading up and better understanding video signal, digital sampling, psychovisual/psychacoustic data compression, and color theories will help you to better categorize the kinds of elements and processes needing to be involved.

Scott

[zanaitoryoushi]

On the surface, I'd say only use Tweak AND Smoothlevels if just one doesn't have the full controls necessary (they kind of overlap/duplicate their work). Also, it would make sense to me to put dfttest before deen. But this is all WITHOUT having seen any of the before/after video. Plus, I only use AVISynth occasionally, so maybe I'm not the best reference. Otherwise, the order looks fine.

These things are best approached from Doctor's point of view. Diagnose, Prescribe, Treat. With each obstacle being at least one task (though some task treatments influence others).

Using waveform monitors/vectorscopes really does help also, as then you are getting objective, statistical data to back up your decisions. They help clarify what's really going on. Plus, as suggested above, solo (isolate) each process to see what it's specifically doing irrespective of the others.

Also, just reading up and better understanding video signal, digital sampling, psychovisual/psychacoustic data compression, and color theories will help you to better categorize the kinds of elements and processes needing to be involved.

Scott

Hi, thank you so much for the insight (just logged back in here, been real sick for a couple of days)

anyway, isolating plugins and scripts is really a good idea to get the feel of how they will function (or if they will behave the way you like them too)

[Cornucopia]

Hope you're feeling better.

Scott

[zanaitoryoushi]

yep, feeling much better now, thank you