[No Longer Available] Evil Dead II - Open Matte (Upscale Project #2)

[PDB]

Evil Dead II - Open Matte (Upscale Project #2 v1.3)

See project origin here: https://fanrestore.com/thread-1071.html

Video:
The open matte DVD was upscaled with a combination of super-resolution and nnedi with filters to reduce macro-blocking and edge enhancement/halos. The resulting image was further processed and then re-grained with a real grain best matching the BD version. The video was then color corrected with a 35mm LUT and color matched to some 35mm frames and the 35mm trailer scan from book of the dead(www.bookofthedead.ws)

Audio:
1. Mono PCM 2.0 from the wide screen laserdisc

2. Dolby Digital 5.1 from the open matte DVD.

3. Dolby Digital 2.0 commentary from the open matte DVD.

Pics:




Thanks:
Doombot: for testing
MrBrown: For the Japanese PCM soundtrack

[MrBrown]

The video really looks impressive, considering it is an upscale.

[TServo2049]

Holy crap, that looks a hell of a lot like the 35mm print I saw a few years back. I don't know where you got those 35mm frames, but it looks dead on, down to the final scene having a greenish tint and being nearly sepia in how little color it has.

[spoRv]

Great project! I watched it few days ago on laserdisc (the bloody red one) and colors seems to be the same.

Just a note: isn't better to do the color correction as first step? This is what I do with my regrade projects; crop black bars, color correction, enhancement, upscale, grain plate, add black bars again.

[MrBrown]

Great project! I watched it few days ago on laserdisc (the bloody red one) and colors seems to be the same.

Just a note: isn't better to do the color correction as first step? This is what I do with my regrade projects; crop black bars, color correction, enhancement, upscale, grain plate, add black bars again.

I would say that it makes sense to do first upscaling, so you have the original source to work with.
If you color correct first you may kill some (minor)details from the original source that get a different result in upscaling before and after correction.

Also if you regrain or enhance before color correction, the grain may get some more natural feel, because it may be a bit more in the color tones as the video, that might give a bit different feeling, like layed on, when regraining after color correction.

Enhancement after upscaling: because the computer has some more pixels to use fore enhancement, so the result might be better.

Lets say, we take a 16*16 pixel field, and upscale it to 32*32 pixel (so you have 4 times the area)...
It may be that a algorithm to enhance change one pixel.. and after upscaling 4 pixel are affected,
while the same enhance algorithm may be a bit more precice after the upscaling was done and just changes 1 pixel extremely, two mixel slightly, and 1 pixel not at all, compared to the 4 upscaled pixels in the other case....

(just some try to approach the steps logical, and theopretical, I have never testet, or done anything like that :o I MAY be wrong. )

[spoRv]

Following the logic, you are right MrBrown... but, following my experience, and the tests I made (and continuing to make every day), logic is not always to be followed with video restoration...

Color correction: if you make it at the end, you will end up to color correct added grain as well; even if i must admit difference could be barely noticeable in most cases, it should be there. Note that grain color is neutral, until you regrade it as well; that's why I add it only at the end.

Enhancements after upscaling: apart the fact it will take up to five (PAL) or six (NTSC) times to compute, I noted that results, that should be better, in theory, is ofter (really often) worst than doing the enhancements before upscale.

Again, just my own experience, so your mileage can vary. A proper thread where other project makers could talk about this topic could be interesting.

[PDB]

The video really looks impressive, considering it is an upscale.

Thank you sir and thanks for the help with the extra track and previewing it.

Also thanks to Doombot. This video was pretty crappy. The video was 2-ish GB of Mpeg-2 data (minus the soundtracks), so it was block city. But besides that there was a lot of noise, not grain but video noise. Luckily, Doombot watched about 20 versions till I got it as good as I could. I'm very much a fan of avisynth plugin: QTGMC. That is a terribly powerful tool and was able to minimize the noise.

Holy crap, that looks a hell of a lot like the 35mm print I saw a few years back. I don't know where you got those 35mm frames, but it looks dead on, down to the final scene having a greenish tint and being nearly sepia in how little color it has.

I'm happy to hear that Servo!

Following the logic, you are right MrBrown... but, following my experience, and the tests I made (and continuing to make every day), logic is not always to be followed with video restoration...
......

Again, just my own experience, so your mileage can vary. A proper thread where other project makers could talk about this topic could be interesting.

I've tried it both ways when I was coming up with a workflow and find that color correction at the end yields more HD and dynamic looking colors.

I think it comes down to part of what MrBrown says, more information, more data yields better results. Or it could be changes in colorspace from SD to HD or uncompressed. It just looks more HD to me in this case.

This is coming to the organ in a little bit to join BTTF. Got to finish Army of Darkness....

[The Aluminum Falcon]

Very excited to see this! Never seen EDII theatrically, but this looks like a nice watch, easier on the eyes than the untouched open matte DVD.

[zoidberg]

This looks pretty sweet. Evil Dead I I is pretty much one of my favourite films. This would make an excellent companion piece to the full frame Evil Dead. Can't wait to see this!

[PDB]

Thanks guys. Unfortunately, its still an upscale but hopefully a watchable one. In the end I don't think we will ever see a HD open matte copy of ED2 so this might be the best alternative for now.