AntcuFaalb

RELEASE PLAN

This should keep everyone happy.

Release 1.0: All 5 Raw Captures; Huffyuv-compressed AVI; Audio uploaded separately; Usenet.
Release 1.1a: Median1(c1, c2, c3); MPEG-2; Audio sync'd; Usenet and Myspleen.
Release 1.1b: Merge(Median1(c1.EF, c2.EF, c3.EF), Median1(c1.OF, c2.OF, c3.OF)); MPEG-2; Audio sync'd; Usenet and Myspleen.

I will include several high-quality images of the tape itself. Furthermore, I will include high-quality images of my setup, including all relevant settings involved.

Redistribution of PS78 outside of Usenet or Myspleen should be bit-for-bit copies -- nothing more.

Release 1.2, if it ever happens, will be less painful to watch. Furthermore, it will be a proper DVD with menus, chapters, extras, and more.

reave said:

Very slowly, if you couldn't tell. It's the next project I'm finishing up, so hopefully it will be done soon(ish)

 That's great news! I was hoping this wouldn't be a goner!

Am I missing something here? Prior to the release of the 97SE, GL claimed that the films were modified to match his original “vision” for them, so do the 2004 DVD and 2011 Blu-ray releases not match his original “vision”?

Would it be worthwhile to have a list of active and dead projects? Maybe another sticky'd post...

TServo2049 said:

Laserman is alive, he posted on Mike Verta's forum a while back.

"He's alive! The Laserman's alive! He's in the Old West, but he's alive!"

You_Too said:

Yeah, the aspect ratio looks better. Maybe it should be just a slight bit taller?

Anyway, are you sure there is no way to keep the original resolution? I mean, each field is half resolution in the original too, but together they make up the full image.

Otherwise, how about upscaling all the captures to the double size and then downscaling it again to 720x480? At least double the height?

It's an interesting trade-off. Using Merge to average the even and odd MedianOfThree'd fields really reduces noise in the image, as you can see. The downside, of course, is the loss in resolution. I'll try to see if I can find a way to have my cake and eat it too... I hope!

Moth3r said:

It's moderate dot crawl made worse by over-sharpening. ;-)

Does your LD player have an s-video out?

You_Too said:

That last sample pic does look pretty good.

Could you post it in full resolution?

That is full-resolution. It's the result of averaging the MedianOfThree of the even fields of each capture with the MedianOfThree of the odd fields of each capture. Since fields are half-resolution, 720x240 is all we're going to get. The upside is that the aspect ratio is now correct.

Median1(c1, c2, c3)

StackVertical(Median1(c1.EF, c2.EF, c3.EF), Median1(c1.OF, c2.OF, c3.OF))

Merge(Median1(c1.EF, c2.EF, c3.EF), Median1(c1.OF, c2.OF, c3.OF))

Note: EF and OF are defined for the sake of brevity as...

function EF(clip c) { return c.SeparateFields().SelectEven() }
function OF(clip c) { return c.SeparateFields().SelectOdd() }

g-force said:

might want to compare a TOOT or median1 with median2 on a short section, just to make sure that median-of-five is really needed before you kill yourself over the 5th capture.

-G

Great idea! Will do. I'll post the results here in a little bit.

Hmmm...

Moth3r said:

It's moderate dot crawl made worse by over-sharpening. ;-)

Thanks for the update :-) I love your release, BTW. Did you ever do an NTSC conversion of it? I remember reading that somewhere...

Harmy said:

Rendering is in progress again and going very slowly (0.75fps), but as long as everything's fine in the encode, I don't really mind.

I'm still waiting for v2 of hairy_hen's 5.1 mix. If it's not ready before I have to start uploading to meet the 28 August release date (I'll have to start uploading on the 24th or 25th at the latest!) I'll use the older one and people will have to mux the new one in on their own.

On a related note, just wanted to report which audio tracks I have been able to collect so far and ask you guys if you knew about any other NTSC GOUT synced tracks I could use. So here goes; I have and will be using:

h_h's 70mm 5.1 (v1 for now)

1977 Stereo (probably will use the analogue capture)

1977 Mono (Belbucus VHS preservation)

1993 Stereo (LD capture)

Czech Stereo (the same VHS capture as v1.0 but with improved sync)

German Stereo (Improved version from Laserschwert with missing lines returned)

French Stereo (GOUT?)

Italian Stereo (GOUT)

Spanish Stereo (GOUT)

Japanese Stereo (GOUT?)

Hungarian Stereo (1993)

Audio descriptive commentary Stereo

 

Have you seen schorman13's recent work? He ripped digital '77, '80, and '83 stereo mixes from an LD set.

You_Too said:

Wouldn't 4 captures be enough? Or does it have to be 5 for that median thing to work?

It has to be five for g-force's Median2 function to work. It's fine. I'm going to stick with it until I've solved the problem.

stretch009 said:

msycamore said:

stretch009 said:

How does Lee's compare to Moth3r's?

 http://moth3r.heliohost.org/ld-screens.htm

 

Thanks, and from that I would say I definitely prefer Moth3r's.  It appears sharper.

Is Moth3r's capture suffering from terrible dot-crawl or is it just the screenshot?

R2D2 said:

Hi AntcuFaalb! My guess is a TooT() with two average captures and one good captures will lead to an average quality result, but i am open for being wrong on this. :)

You're probably right, but I'm just experimenting anyway :-)

Capture #5's being a pain-in-the-neck. Argh!

Are there any other JSC captures floating around?

More Good News: It's the same story for Captures #2-4. Furthermore, the numbers of duplicated frames are making sense; that is, when you subtract the number of duplicate frames from the total number of frames for each capture, you end up with each capture having the same total number of frames. Yay!

reave said:

Update: I have had zero time to work on this, so no update. :|

I am still going to finish this. 

 How's it going?!

Good News: For Capture #1 (the only one I've tested so far), MultiDecimate identifies the same duplicate frames as those I've manually identified!

You_Too said:

I've explained it earlier in the thread but I understand if you don't wanna go back and search for it.

First I analyzed the whole movie scene by scene, writing down any frame that had glue marks, big dirt marks, stripes, flashes or light bleeding from frames before and after a frame, for example. (There were hundreds) And now we've also added scenes with color errors to the list.

DJ has extracted them from the render and I'm fixing them all in photoshop. Then DJ will take care of reinserting the fixed frames.

EDIT: @none: Thanks for adding that. :)

I'm sorry for being lazy! This is a big thread and "site:originaltrilogy.com xxxx" Google searches only go so far... Impressive work!

You_Too said:

@AntcuFaalb: You mean with incorrect aspect ratio and such things?

Since our renders for all three films are already done, and we've worked so much on fixing bad frames already, we can't go back and change any crop settings or anything now, so if there's scenes with some slightly different aspect, we'll have to live with it.

Anyway, each film was cropped to remove black borders, any edge discoloration and stripes on edges.

Here's how that frame looks in our version:

It looks great! Which bad frames did you fix?

toho-scope said:

a handful are on the spleen and seeded but the best of the early caps was Moth3rs PAL set IMHO

I see. I'm interested in experimenting with averaging >= 5 of the best (and unfiltered) NTSC DC captures. There are so many available, so why not?

none said:

^ Post 19.

Wow, I'm silly. ^_^ Thanks for pointing out that post.