RU.08 said:
As we've discussed, you really need a DVD that has considerable aliasing to benefit from it.
if it really needs aliasing, why not try upscaling each field. I'll bet dr dre's script would be great at deinterlacing 60i material
Here's the problem, the script used to make the myspleen reencodes.
a=ffvideoSource("G:\TEMP\THX1138_16mm\THX1138r1a.mov")
b=FFVideoSource("G:\TEMP\THX1138_16mm\THX1138r1b.mov")
c=ffvideoSource("G:\TEMP\THX1138_16mm\THX1138r1c.mov")
d=ffvideoSource("G:\TEMP\THX1138_16mm\THX1138r2a.mov")
e=ffvideoSource("G:\TEMP\THX1138_16mm\THX1138r2b.mov")
trim(a, 1721, 30599) ++ trim(b, 104,19433) ++ trim(c, 418,11770) ++ trim(d, 312,50058) ++ trim(e, 62,13728)
Crop(0, 180, -0, -180)
Spline36Resize(1280, 720)
The first trim takes out everything before the WB logo
What are you guys talking about? The black and white 16mm starts with buck Rodgers. Unless there's now multiple edits of that too
Did you tweak any other settings or add any kind of compression? because I'm seeing subtle luma differences in frame 7218 comparing your new sample with srv10 in photoshop. Though looking just at the chroma in the rgb channels, the new sample seems more detailed than both srv10 and spline64. I really hope you post your script in the end, sounds full of clever solutions.
I kinda thought that too, but was too lazy to do a real test.
http://screenshotcomparison.com/comparison/128400
look at the orange light to the left of 3po's head.
I was wondering how chroma is being handled. A few small lights in your comparisons seem less colorful than Spline64, for instance the red lines on the wall in frame 4691. I would think you could just use the raw super resolution for chroma without artifacts being noticeable, but if not spline64 would be good enough
Looking at the screenshots, seems to me like some of the heavily damaged frames are just being dropped and replaced with an interpolation of the surrounding frames. Is this cleanup software being used on all your scans? Anyway, It still looks amazing and I'm sure a lot of people would love to have these.
I'd say just give up on super resolution for the vertical resolution.
Here's a quick script I whipped up.
http://screenshotcomparison.com/comparison/126389
a=FFmpegSource2("Star Wars 10 min sample.mkv")
a.VerticalReduceBy2().Addborders(0, 1, 0, 1).MCTemporalDenoisePP(settings="very high", aa=true).eedi2(pp=2).crop(0, 0, 0, -4).merge(a, 0.1)
followed by virtual dub filters MSU smart sharpen(small overall and borders) and msharpen.. avisynth probably has better sharpening filters, but I'm far from an expert
_,,,^..^,,,_ said:
Don't know if this is due to compression artifacts (even if I don't think so, because 20Mbps should be enough to avoid them), but it seems that the superresolution algorhythm you are using is not that effective... a simple avisynth bicubic upscale, using the compressed MKV SD video you provided, sharpened a bit, seems very similar...
http://screenshotcomparison.com/comparison/117604
what do you think?
here's a better comparison.
http://screenshotcomparison.com/comparison/117784
Since the artifacts look more like interlacing problems then blown up interlacing problems, eedi takes care of it easy enough. I also added a sharpening filter to try and match yours. Now the difference is incredibly subtle, I'd guess it's temporally finding noise and dithering it into smaller noise?.. though without a raw frame its hard to say, fake fine grain + recompression might look the same
edit: ignore any differences in sharpening
Since nobody else has, I thought I'd share some thoughts and opinions on this release.
First, after reading the description a couple of times, I was very excited to see how this turned out. It sounded like the sort of playing around I'd be doing if I had that plugin or the harddrive space. Adding information from a faded print to a fanedit sounded crazy and probably as destructive as helpful, but the results are amazing!
The two issues I've had with the colors in harmy's ESB 2.0 were a lack of overall consistency(especially with saturation), and a large number of shots with overly blue highlights. This version does wonders in making the overall colors look more like film than a patchwork of digital tinkering. It also fixes alot of the blue highlight issues especially on hoth where it feels more like a color boost than something entirely artificial. It fails, however, on more extreme examples like many of the cockpit shots on the millennium falcon (just look at leia's coat when they discuss the "Lando system"). Judging by how bad those shots look next to the gout footage sounding them, I've been assuming they're just place holders untill the whole scene can be replaced with 35mm scans.
I haven't downloaded the full bluray version, but if it's as artifact free as the dvd-9 version seems, you may have saved Harmy alot of work on his v2.5 release. There could be some black crushing going on... but it feels natural enough to me.
"And a picture with the GOUT reference in the corner (the GOUT reference has increased saturation but I pulled red saturation down by the same amount I increased the master saturation, so red saturation remains the same as in raw GOUT but all other colors are increased)" -Harmy
can you expand on that? Are all gout clips affected? I'm amazed at how film like the work print looks(especially without a proper color reference), but the one issue I had is with an inconsistent hue, especially in the first half in hoth and mostly noticeable in bright red lights(and the sometimes purple blues). Bright red lights sometimes jump from pink, red, to occasionally orange... I don't know if it's the selectively colored bright lights of the blu ray or the analog nature of gout and the SE source to blame, but even a fraction more consistency in the red channel would really sell me on the overall colors.
Sorry that this is my first post, but I thought you might want to know it's extremely easy to find and download what I assume are the videos you sent harmy. I also found a dickpic that I hope was meant for Winnie cooper.