logo Sign In

THX 1138 "preservations" + the 'THX 1138 Italian Cut' project (Released) — Page 62

Author
Time

It looks exactly the same as the one I have.  I need to photograph the box and cassette one of these days for the archives. :)

Forum Moderator

Where were you in '77?

Author
Time
 (Edited)

@ SilverWook & ww12345

Thanks guys! To quote another hotshot ... "sometimes I amaze even myself".  :)  But don't count-out those faded prints. They have much more in them that we just don't see.  ;)

My "select samples" were carefully grabbed. The clips have not yet been brightness stabilized. Therefore, the color correction keeps drifting in and out of the optimum settings of the work frame, from where came the video processing numbers. Once stabilized, the clips should play uniformly well corrected.

There also is the non-uniformity of frame illumination. It grows darker towards center-frame. I'm thinking if an inverted, blank title frame is used as a luminance offset, that should correct the intensity across the picture area. (I'm looking into it now.)

 

[Just a mention about eBay links. To  prevent eBay tracking (for whatever their nefarious purpose), remove everything from the "?" onward. The first half of the link, alone, will bring up the item-page just fine -- but without the original reference hash-info.]

The US VHS (pre US laserdisc release by about 8 years and pre JP laserdisc by only 3) has "©1983" for "package design and summary" and no extended SEN shots (edited out by GL as per my aforementioned versions theory). I thought this was the first release of Lucas' new cut, after his Star Wars success.
The eBay seller's picture of the Betamax box  (inside flap) shows "©1982" for it's packaging -- a year before the VHS one. Would this still be Lucas' 2nd director's cut (without those extra SEN shots)?

 

Author
Time

Horrible pan and scan aside, I couldn't spot any differences from the version we're most familiar with. The exact same transfer was used for the early full frame Japanese LD, which I sent msycamore a capture of sometime back.

Forum Moderator

Where were you in '77?

Author
Time

Spaced Ranger said:If you had used gamma and mid-tones, OT forum members would have fair-chance at restoring the original colors. But you used low/high adjustments and crushed a quarter of the Red channel to achieve your effect.

Thanks, Spaced Ranger.  Your post on FotR confirms what I have long suspected: it's impossible to color-correct the EE BDs satisfactorily.

Is anyone still seeding the Italian cut of THX-1138?

Author
Time

Chewtobacca said:

Is anyone still seeding the Italian cut of THX-1138?

One seeder at the moment.

 

Author
Time

Chewtobacca said:

... what I have long suspected: it's impossible to color-correct the EE BDs satisfactorily.

Thanks for bringing that up!  :)  After I threw that well-deserved ripe tomato at PJ, I wondered if something could be a satisfactory solution.

Color correction cannot fix crushed color. I checked all 3 color channels and only Red was crushed (indicated in it's bright color), Green & Blue were not.

snicker's demonstration of injecting other channel(s) detail into the damaged channel area (see his comments in http://originaltrilogy.com/forum/topic.cfm/New-colormatching-script/topic/15002/) is a real solution. It necessarily requires significant manual adjustment on a per shot basis, though.

Color correction alone can be a mostly-good fix. Red is crushed in it's 1-70 luminance range -- yeah, I know, ouch! But if this now 0 area is raised up to 8 or more (to be balanced where it would show the most, for example, on skin tone), it helps fill in that color gap.

So the top picture (TFOTR theatrical edition) was damaged into the center picture (TFOTR extended edition). After applying this color correction, the resulting bottom picture looks pretty good. However, originally crushed areas (mountain shadow, darker sky) will remain somewhat red-anemic.

<-- histogram fix

Histogram  low input  high input  low output  high output  gamma
RED             0              192            8                 255              0.9
GREEN         0              220            8                 255              0.8
BLUE            0              204           16                255              0.75

 

Author
Time

Thanks for the demonstration. :-)  The result is better than expected, but I'd still prefer to stick to the theatrical BD of FotR, despite its shortcomings.  Unfortunately, I could never locate the HDTV broadcast itself.

stretch009 said: One seeder at the moment.

Thanks.  It's downloading very slowly.

 

Author
Time
 (Edited)

So are you saying if one were to make a Despecialized version, they could use the 1991 laserdisc as a reference guide since it has the extra five or so minutes that were originally cut out?

Author
Time
 (Edited)

Well, according to this (the "theory" part) http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/665567/#TopicPost665567

                   

  Lucas'             ver.1            ver.1.WB          ver.1             ver.2               ver.3
THX1138          1970              1971              1978             1983               2004

where:
  version 1 1970 = Lucas' original (un-released)
  version 1.WB 1971 = Warner Bros' re-edit (shorter) (released theater)
  version 1 1978 = Lucas' original (theater, TV [16mm])
  version 2 1983 = Lucas' next edit (- SEN shots) (TV, home video [LD Beta VHS etc.])
  version 3 2004 = Lucas' next edit (+ CG) (TV, home video [DVD])

So, only 35mm & 16mm film for theater & TV, printed from 1978 to 1982, would be of Lucas' original movie (original length, original SEN shots) -- his 1st Director's Cut (as he lately, redundantly calls it).   :)

 

After watching those SEN shots, I can see why GL would remove them. The voice dubbing on them is poor -- it sounds like what it is, women voice actors dubbing for the children (standard practice). Also, when the 2nd child asks SEN his name, SEN looks at the 1st child and asks if he has a name. But the 1st child whispers in the 2nd child ear not to tell his name. Clearly, SEN's counter-question should have been directed at the 2nd child, who asked the original question. Cut those shots and the problems are fixed.

Author
Time
 (Edited)

This gets me ... every time ..

http://forum.blu-ray.com/showpost.php?p=7472786&postcount=303
WyldeMF said:

In any event, there's no reason Lucas can't leave his damn films alone aside from fixing lighting, matt lines, color issues or what not. Instead he still gets those obvious things wrong, but manages to add cartoonish stuff absolutely no one but himself wants to see in these classic films.

.. as demonstrated by this prime example:

  <-- 2003 TV (film ver.2)

In this shot, the film was threaded too tight inside the camera. This caused intermittent drag during exposure, resulting in vertical blurring (the direction of film travel in camera). It is most noticeable in brightly exposed areas -- the overhead lights, in this case. Lucas probably couldn't go back for a re-shoot, so he added sound effects to mask it as a deliberate visual effect. But anyone who's ever watched movies knows this was a mistake (like a "mic in picture") and an obviously absurd "fix".

But, in 2004, GL threw real money into a new edit, complete with computer animated monkey-creatures ... except he never fixed this mistake! Instead, he puts additional actors into the originally empty hallway and brightens them when the blurred exposure shows on the lights!

  <-- 2004 DVD (film ver.3)

This is the filmmaking genius George Lucas

 

Author
Time

 

Saw a different Japanese THX Laserdisc on eBay. This one is is numbered: 08JL-61162.

The all-white, Japanese copy that I have is numbered: NJWL-11162

What's the difference between these two? Is it simply a different cover but same print?

Author
Time
 (Edited)

http://www.lddb.com/search.php?search=thx+1138&sort=title

The LaserDisc Database lists 2 JP laserdiscs for THX 1138. The 08JL-61162 (1986) is P&S (fullframed pan & scan) and the NJWL-11162 (1995) is WS (letterboxed widescreen).

The quality issue, as well as the censorship issue, were raised on the 1st post page of this thread. (I don't think anyone established anything conclusive thereafter.)

Author
Time

Spaced Ranger said:

http://www.lddb.com/search.php?search=thx+1138&sort=title

The LaserDisc Database lists 2 JP laserdiscs for THX 1138. The 08JL-61162 (1986) is P&S (fullframed pan & scan) and the NJWL-11162 (1995) is WS (letterboxed widescreen).

The quality issue, as well as the censorship issue, were raised on the 1st post of this thread. (I don't think anyone established anything conclusive thereafter.)

Thanks for the info—appreciate it.

Author
Time
 (Edited)

Previously, I mentioned the 16mm print (which I believe is THX 1138 ver.1) has a non-uniform brightness across what should be a dark, flat background during the opening titles:

                 background brightness variation                            uniform self-correction (explained below)
 

Normally, one would think it to be a center-spotlight effect and there are Avisynth filters to handle such specific situations. This, however, is different. There is variation across the frame, but it's mostly darker in the vertical center and lighter towards the vertical sides (optical printer light leakage?). The best solution is to use itself as it's own correction. Here's how I worked that out.

First, for this to be effective, the film needs to be brightness stabilized (which it is not). My demonstration here is only on select frames, to prove a concept. As this is a laserdisc preservation thread, I probably won't take it beyond this. But the steps I outline to do this in a paint program are reproducible in Avisynth.

Second, I'm thinking that this technique should be applied independently to each color layer. As this is a faded print, it should not be assumed that all the layers are equally affected by this anomaly. For this demonstration, I applied the correction as a single brightness correction. Even still, it is an improvement -- most noticeable in the flat title graphics, but less so in the live film that is naturally non-uniform.

 

Using the THX  1138 title graphic as my sample, this process is applied in the paint program's Layers -- using frames that were approximately the same background-brightnesses, to compensate for lack of brightness stabilization:

This frame needs to be processed by a "clean version of itself". Fortunately, there were  complete "blank" frames  in this scrolling title sequence. If there were none, a clean frame would need to be constructed from an assembly of blank areas of different frames.

As the picture is grainy, I didn't want that grain to apply itself into the correction. Several grabbed blank frames, at matching overall brightness, are blended together by temporal smoothing. The randomness of the grain in each works to cancel out the others (yes, it really works!). The more frames, the more randomness, the better the random grain self-cancellation. As each blank background is added for temporal smoothing, it's count from the bottom becomes it's applied percentage of opacity -- according to this technique:
  * base layer -- 1st layer from bottom (1 of 1)  -- set to 100% opacity
  * 2nd layer   -- 2nd layer from bottom (1 of 2) -- set to   50% opacity
  * 3rd layer    -- 3rd layer from bottom (1 of 3)  -- set to   33% opacity
  * 4th layer    -- 4th layer from bottom (1 of 4)  -- set to   25% opacity
and so on. They are grouped together here for organization:

To remove the color and work strictly in luminance, the merged blanks are desaturated (0% color):

As this is supposed to become a flat background, the reverse of the brightness variation will be averaged itself to produce flat brightness. This is accomplished by inverse:

Of course, inverse works across the entire dark-to-light spectrum and our darkness has become inverted to lightness. With levels, the lightness can be brought down back to it's original darkness baseline (but still inverted). By using a Histogram function, one can see the range-spike of the picture both before and after the inversion. In this case, the range before was 32-54 and the range after is 181-213. Therefore, the high range is brought low in levels with:
  * Input low    = 181
  * Input high   = 213
  * Output low  =   32
  * Output high =   54
and is so applied:

That is the correction for any frame to which it is applied (using the paint program Screen overlay), and it grouped together for organization. You can see, however, I had to include an additional adjustment at the very top of Layers. The result was overall lighter than is should've been. To compensate, a simple adjustment using levels, from 36-255 to 0-219, slides the spectrum down by 36.

 

All this, applied to the original title frame, produces an excellent result:

To test the effectiveness of this process, I also applied the correction to a blank background. The before and after range-spikes where expanded to full spectrum to easily observe their density distributions:

                    density distribution before                                               density distribution after
  

Success! It actually works!
Here is the title frame, self-corrected from the 16mm's own non-uniform brightness, shown full-sized, before & after:

 

Author
Time

Spaced_Ranger, I love it when you post - it's like a pro magazine article or something. You never fail to impress!

Author
Time

ww12345 said:

Spaced_Ranger, I love it when you post - it's like a pro magazine article or something. You never fail to impress!

Agreed!

A picture is worth a thousand words. Post 102 is worth more.

I’m late to the party, but I think this is the best song. Enjoy!

—Teams Jetrell Fo 1, Jetrell Fo 2, and Jetrell Fo 3

Author
Time

What program is that? Looks great, looks exactly what i need to remaster some old VHS tapes i have. Then again i have Vegas Pro...

Author
Time
 (Edited)

Thanks guys!  And little wonder it takes me so long to put up a simple post.  :D

 

I wanted to show the density distribution process but didn't have the graphics ready ... until now ...

Spaced Ranger said:

To test the effectiveness of this process, I also applied the correction to a blank background. The before and after range-spikes where expanded to full spectrum to easily observe their density distributions:

                       density distribution before                                                  density distribution after
  

The grey graphing is the actual luminance of the backgrounds and the red is from the narrow spike's min-max spread out across the full spectrum.
Paint program processing could've been better (for a flatter result) but it shows the concept working well enough. Avisynth, for the actual video processing, may do the internal math better or have better functions.

 

@ DarthAss - Most of these are performed on screen captures in a paint program (any one will do as most have the basic functions I use). Applying them to the actual video can be with Avisynth, a free, non-linear video/audio editor using scripts to direct it's processing -- http://avisynth.nl/index.php/Main_Page -- that many of us use. But other video editors, like your mentioned Vegas Pro, can do the same things.

 

Author
Time

This was the other section where I thought to show what it was doing. So, again, quoting, and this time adding the analytical graphics (note that the before & after windows are the same; Histogram is being used only to display the spectrum) ...

 

Spaced Ranger said:

To remove the color and work strictly in luminance, the merged blanks are desaturated (0% color):

 

continued:

As this is supposed to become a flat background, the reverse of the brightness variation will be averaged [to] itself to produce flat brightness. This is accomplished by inverse ... Of course, inverse works across the entire dark-to-light spectrum and our darkness has become inverted to lightness.

 

continued:

With levels, the lightness can be brought down back to it's original darkness baseline (but still inverted). By using a Histogram function, one can see the range-spike of the picture both before and after the inversion. In this case, the range before was 32-54 and the range after is 181-213. Therefore, the high range is brought low in levels with:

  * Input low    = 181
  * Input high   = 213
  * Output low  =   32
  * Output high =   54
and is so applied:

 

So the result of these Layer steps was that the background luminance-spike was mirrored in-place.

Note:  There shouldn't be manual reading and entering of numbers to make it work, as I did here. Better designed, it should be a state machine (https://en.wikipedia.org/wiki/Finite-state_machine). One only inputs any source and any source-blank, to have it blindly produce the proper correction. Avisynth should have functionality to handle this.

 

Author
Time

I've received the new cables I've been waiting on from svideo.com: (1) composite[male]->BNC[male] and (2) s-video Y-cable. I'm sick of using adapters.

I'll be able to switch rooms and start capturing soon.

A picture is worth a thousand words. Post 102 is worth more.

I’m late to the party, but I think this is the best song. Enjoy!

—Teams Jetrell Fo 1, Jetrell Fo 2, and Jetrell Fo 3

Author
Time
 (Edited)

What media-type capture are you remastering spaced ranger? Or, are you remastering, or just a simple clean up of artifacts, noise and/or other picture and audio nasties? Just asking because i'm quite interested! And i'm too tired to go and read back several pages! Might do tomorrow, i'm not that lazy, hahaha!

Author
Time

@ AntcuFaalb - And when things get going, looking forward to your continued procedure updates (either back in your post-holders, or just in-line here)!

@ DarthAss - For this thread's projects, I'm just wrapping up msycamore's laserdisc preservation work for him, which includes the curiosity of an older 16mm print's shots edited out of this newer version. My contribution (so far) is a video fix-up approach for that print. I demonstrate this R-G-B-channel-centric approach to projects in other threads so they may run with it there, too, if they find it useful.

 

Author
Time
 (Edited)

I've just finished a first pass at conforming the Blu-ray to the cut present on the Laserdisc (using the Angrysun edition). The color grade has, of course, been extremely altered and several scenes were rearranged, but other than THX's initial meeting with SEN and the factory sequences I didn't find much that was completely re-edited.

There's the movie-censorship article which cites most of the alterations so I'm only giving additional observations: 

-Two out of three of the SEN shots removed from the opening were found later on the Blu-ray with no missing or modified frames (the first extreme close-up is gone).

-The opening "grid" shot, which was modified with monitor/scanlines, is pulled from a clean shot before the trial, meaning it can be re-created.

-From the trial sequence up until the escape to the crowded mass the two edits are almost completely identical. *Almost* because I spotted missing frames on the Angrysun version (no idea how that happened), two missing frames from the Blu-ray, and only one digital alteration when THX and SEN walk off into the horizon (the food squares SEN drops don't fade to white on the Blu-ray).

-Most of the horizontally flipped shots were spotted except a couple during the white prison escape. I only noticed one digital alteration that wasn't spotted (unless it's Blu-ray only): the bird's-eye view shot of the officers at 1:04:06 (Angrysun timecode) has an added foreground that resembles looking down a manhole cover.

 

Quantified in minutes, starting after the opening credits

72:12:47 of the film exists intact on the Blu-ray

9:51:13 has been modified, including shots with missing frames

2:36:12 has been removed entirely

 

I've begun rotoscoping (a la Harmy) on some of the simpler shots and noticed the LD transfer doesn't fit the Blu-ray 1:1, which I'm pretty sure is because the LD was pulled from an anamorphic print while the Blu-ray/DVD was directly from a Techniscope element (if not the negative)-- your pesky white speck in the last third could be something that got sucked into the optical printer during the blow-up process. Is there a simple way to rectify anamorphic distortion?

(Software I am using is Adobe Premiere and After Effects CS4)

As some of these shots only modify a small part of the frame, I'd find the pan-and-scan versions to be useful to work with. Are there any seeders for the Italian cut (or any other pan-and-scan version)?

At least Ted Turner released the originals.