logo Sign In

THX 1138 "preservations" + the 'THX 1138 Italian Cut' project (Released) — Page 53

Author
Time

Have you looked at samples from moviestuff's newest unit by any chance?

http://www.moviestuff.tv/retro_8_telecine.html

There's supposed to be a 16mm version in the works, but, as per usual for moviestuff, it's not going to be able to take the larger reels feature films typically come on. :(

Forum Moderator

Where were you in '77?

Author
Time

Not many technical details there, but the operational description looks great:

"no sprockets, no advance claw, no capstan, no pinch roller, no belts, and no gears means virtually no maintenance. Nothing to lube, nothing to adjust. You don't even have to focus! ... Long life super bright LED produces zero hot spot, lasts thousands of hours and produces no heat on the gate or film. ... The Retro-8 features SafeScan zero image area contact at the gate and rubber guide rollers that handle the film only by the edges for maximum film safety. ... the Retro-8's Solid State Registration means a rock steady picture, generally accurate to within a pixel. The Retro-8 scans in full 24 bit color using industry standard square pixel orientation. HD output files are 16:9, 1080p with black bars on the left and right of the film frame."

The sample color-film frames are problematic as fringing there may be from the consumer-grade movie cameras. However, the B&W-film frames eliminate that possibility and do not show any fringing from the scanning process (as expected from a focus-less process):

A WorkPrinter trade-up might be worth the improved quality & convenience once a 16mm machine is available.

Author
Time

Recently came across this reference, from an obscure forum (aren't we all?) :) , to an 2005 eBay auction:

Anyone have one of these "THX-1138" (only in the U.S. is the title "THX 1138") VCD's (Video CD)? I don't recall a mention of this in this thread. The original poster didn't have any more information but wondered aloud that it should be widescreen. And, of course, eBay no longer displays that auction (and presently has none like it).

Author
Time

IIRC, someone mentioned having it several pages back. And it was fullscreen.

VCD would be on par with VHS in resolution anyway.

Forum Moderator

Where were you in '77?

Author
Time

Yes, Russ had a VCD copy which he posted a screencap of and it was Pan & Scan if I recall correctly.

I have also noticed that color fringing on Puggo's transfer you described on the last page, I'll see if I can use your method for it, thanks for the suggestion, appreciated! :)

 

 

We want you to be aware that we have no plans—now or in the future—to restore the earlier versions. 

Sincerely, Lynne Hale publicity@lucasfilm.com

Author
Time
 (Edited)

@ SilverWook & msycamore

Thanks for the info. Too bad the VCD isn't a more useful source. I guess it's back to waiting for a possible(?) 16mm scope print.

msycamore

Because of the slightly de-focused RED and BLUE separations, the chromatic aberration correction will always have some color fringing (unless serious processing is applied to those color-seps to fix the fuzzed edges). Interesting that they look allot like edge-enhancement halo's.

While not perfect, the result is yet another improvement. This example is from the early paint program color correction test -- the later color correction in Avisynth / VirtualDub will, of course, look better:

                   CHROMATIC ABERRATION w/CC                                                 CORRECTION w/CC

Author
Time

I had a lead on another print, but it was a rental. (Which is mind boggling there would even be a market for that now.) The owner does sell off prints, but I fear making inquiries at this point would only inflate the price.

Forum Moderator

Where were you in '77?

Author
Time

Hello, sorry to get a bit off-topic, but Spaced Ranger, can you really fix color fringing with home computer tools? I'm only curious, since I know it is an unfixed issue on some BD releases, registration issues, that is. Some BDs that have it off the top of my head, (Rope, Phantom of the Opera, etc.)

Anyway, to get back on topic, I do love all the work you guys are doing. I've never seen THX-1138 before, and I want to see your guys' superior version rather than Lucas's terrible edition first.

Author
Time

This particular "fix" is for chromatic aberration fringing -- see post on the previous page describing the particulars: http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/624565/#TopicPost624565. I wouldn't expect such a problem to be on a professional release.

However, if you are referring to effects of commercial publishers overusing their poorly implemented EE (Edge Enhancement), then, yes, there is a rudimentary Avisynth filter than can reduce that problem.

First, see post (also in this thread) for an explanation on the proper approach to "sharpen" an image, in Avisynth with the LimitedSharpen filter: http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/529702/#TopicPost529702.

For a possible reduction of EE fringing, see post (again, in this thread) for a demonstration of the DeHalo_alpha filter, also for Avisynth (I max'ed out the settings, for that demonstration, which created it's own problems): http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/552393/#TopicPost552393.

As always, filters are in constant development. Always research the Doom9 forums Avisynth sections for improved or new filtering.

Author
Time

Spaced Ranger said:

This particular "fix" is for chromatic aberration fringing -- see post on the previous page describing the particulars: http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/624565/#TopicPost624565. I wouldn't expect such a problem to be on a professional release.

Unfortunately, it is on the two BDs I described earlier. I believe I've heard it referred to as Technicolor Dye Registration errors. A color sometimes peaks out far beyond where it should be. Universal just used old masters for both and didn't bother fixing. As I am very fond of both movies, Rope more than the 1943 Phantom of the Opera, I'd love to see if there's a way to make a corrected version.

Anyway, thanks for the informative post. I didn't know EE could be reduced as well, that's great.

Author
Time

After checking blu-ray.com for the reviews on both those movies ...

"Thin edge halos appear here and there, color fluctuations and minor bleeding seep into the presentation ..." - Phantom of the Opera (1943) Blu-ray

"... and red color fringe affects several shots. (Rope was shot using three-strip Technicolor, a process which hinged on the perfect alignment of three color separation negatives. If the red layer is a hair off, thin red lines occasionally appear that behave a bit like edge halos.)" - Rope (1948) Blu-ray

(even visible in their 720p "thumbnail": http://images.static-bluray.com/reviews/6908_4_large.jpg)

... then, yes, the chromatic aberration correction would allow you "correct" such alignment (moving the affect separations) and/or shrinkage (resizing the affected separations) "errors" after some by-eye experiments. Keep in mind the separations could be "misaligned" for any reason (shift, rotated, size, warp), not just the simple resizing for the THX 1138 film-capture here.

Author
Time

I pulled that screenshot into photoshop and the red layer is definitely warped, because even when you simply resize and move it so that it lines up at the edges of the screen, it doesn't line up in the middle. Plus depending on what caused the misalignment, the amount could differ from shot to shot and if that was so, it would be a huge pain to try and correct. In the unlikely case that it would be the same throughout the entire film, it wouldn't be too hard to fix in After Effects.

http://img9.imageshack.us/img9/8075/69084largecopy.jpg

Author
Time
 (Edited)

@ Harmy

Even though done on this 720p sample, your example fix is a significant  improvement and probably good enough for the full 1080p. I think you've made someone happy. :)

 

msycamore

Spaced Ranger said:

Interesting that they look allot like edge-enhancement halo's.

A great storyteller always use foreshadowing. :)

Back on page 42, I demo'ed the DeHalo_alpha Avisynth filter on a noticeably EE'd laserdisc snapshot -- with mixed results.

Now, on the film capture, the de-focused edges in 2 of the 3 R/G/B separations have added different color fringing to the chromatic aberration problem. Applying DeHalo to the Red and Blue separations reduced the level of peaking on those bad edges:

                              BLUE separation                                                    BLUE separation DeHalo'ed

In this Blue separation, it flattens out the light and dark peaks. Of course, being in "alpha" development, it produces some artifacts of it's own. But those are less obtrusive and get lost when joined back to the other separations.

Using the same DeHalo settings on both Red and Blue (Green didn't have the edge damage) cleaned up what the chromatic aberration corrections couldn't (notice the shoulder and the collar across the 3 sections -- color fringing reduces as you move along):

            color corrected (cc)           cc & chromatic aberration correction (cac)             cc & cac & DeHalo

Note: color correction was applied later, in a paint program.

I did make one change in applying the chromatic aberration correction. Red seemed to be a very minor player and resizing it, even at a single pixel, didn't produce a noticeable improvement. So I adjusted only Blue and that seemed enough.

Here's the Avisynth script that produced this. DeHalo_alpha is an add-on script (not mine):

##====================
## THX 1138 16mm
## AviSynth 2.5.x script
##==========

## source
##----------
clip1 = ImageSource( "THX1138 16mm -0000.png" )
v1 = clip1.ConvertToRGB32()

## get RED / GREEN / BLUE channels
##----------
v1_R = v1.ShowRed()
v1_G = v1.ShowGreen()
v1_B = v1.ShowBlue()

## DeHalo & resize (chromatic aberration fix) each channel
##----------
## DeHalo_alpha()
## clip clp, float "rx", float "ry", float "darkstr", float "brightstr",
## float "lowsens", float "highsens", float "ss"
## - rx, ry [float, 1.0 ... 2.0 ... ~3.0] Default 2.0
# As usual, the radii for halo removal.
# this function is rather sensitive to the radius settings. Set it as low
# as possible! If radius is set too high, it will start missing small spots.
## - darkkstr, brightstr [float, 0.0 ... 1.0] Default 1.0, <0.0 and >1.0 allowed
# The strength factors for processing dark and bright halos. Default 1.0 both
# for symmetrical processing. On Comic/Anime, darkstr=0.4~0.8 sometimes might
# be better ... sometimes. In General, the function seems to preserve dark
# lines rather good.
## - lowsens, highsens [int, 0 ... 50 ... 100] Default 50
# Sensitivity settings, not that easy to describe them exactly ... in a sense,
# they define a window between how weak an achieved effect has to be to get
# fully accepted, and how strong an achieved effect has to be to get fully
# discarded.
## - ss [float, 1.0 ... 1.5 ...] Default 1.5
# Supersampling factor, to avoid creation of aliasing.
##----------

# DeHalo RED
v1_R_DH = v1_R.ConvertToYV12()
\.DeHalo_alpha( rx=5.0, ry=5.0, lowsens=99, highsens=99 )
\.ConvertToRGB32()
#\.Lanczos4Resize( 718, 480 ).AddBorders( 1, 0, 1, 0 ) #not needed

# do nothing to GREEN
v1_G_DH = v1_G

# DeHalo & resize BLUE
v1_B_DH = v1_B.ConvertToYV12()
\.DeHalo_alpha( rx=5.0, ry=5.0, lowsens=99, highsens=99 )
\.ConvertToRGB32()
\.Lanczos4Resize( 726, 484 ).Crop( 6/2, 4/2, 720, 480 )

# recombined "fixed" separations
v1_DH = MergeRGB( v1_R_DH, v1_G_DH, v1_B_DH )


## display the various results
##==========

#showclip = StackHorizontal( v1_R, v1_R_DH )
#showclip = StackHorizontal( v1_G, v1_G_DH )
#showclip = StackHorizontal( v1_B, v1_B_DH )
showclip = StackHorizontal( v1, v1_DH )
#showclip = v1_DH

showclip



##====================
## De-halo

# Here are the parameters:

# - rx, ry [float, 1.0 ... 2.0 ... ~3.0] Default 2.0
# As usual, the radii for halo removal.
# Note: this function is rather sensitive to the radius settings. Set it as low
# as possible! If radius is set too high, it will start missing small spots.

# - darkkstr, brightstr [float, 0.0 ... 1.0] Default 1.0, <0.0 and >1.0 allowed
# The strength factors for processing dark and bright halos. Default 1.0 both
# for symmetrical processing. On Comic/Anime, darkstr=0.4~0.8 sometimes might
# be better ... sometimes. In General, the function seems to preserve dark
# lines rather good.

# - lowsens, highsens [int, 0 ... 50 ... 100] Default 50
# Sensitivity settings, not that easy to describe them exactly ... in a sense,
# they define a window between how weak an achieved effect has to be to get
# fully accepted, and how strong an achieved effect has to be to get fully
# discarded.

# - ss [float, 1.0 ... 1.5 ...] Default 1.5
# Supersampling factor, to avoid creation of aliasing.

function DeHalo_alpha(clip clp, float "rx", float "ry", float "darkstr", float "brightstr", float "lowsens", float "highsens", float "ss")
{
rx        = default( rx,        2.0 )
ry        = default( ry,        2.0 )
darkstr   = default( darkstr,   1.0 )
brightstr = default( brightstr, 1.0 )
lowsens   = default( lowsens,    50 )
highsens  = default( highsens,   50 )
ss        = default( ss,        1.5 )

LOS = string(lowsens)
HIS = string(highsens/100.0)
DRK = string(darkstr)
BRT = string(brightstr)
ox  = clp.width()
oy  = clp.height()
uv  = 1
uv2 = (uv==3) ? 3 : 2

halos  = clp.bicubicresize(m4(ox/rx),m4(oy/ry)).bicubicresize(ox,oy,1,0)
are    = yv12lutxy(clp.expand(U=uv,V=uv),clp.inpand(U=uv,V=uv),"x y -","x y -","x y-",U=uv,V=uv)
ugly   = yv12lutxy(halos.expand(U=uv,V=uv),halos.inpand(U=uv,V=uv),"x y -","x y-","x y -",U=uv,V=uv)
so     = yv12lutxy( ugly, are, "y x - y 0.001 + / 255 * "+LOS+" - y 256 + 512 / "+HIS+" + *" )
lets   = maskedmerge(halos,clp,so,U=uv,V=uv)
remove = (ss==1.0) ? clp.repair(lets,1,0)
          \        : clp.lanczosresize(m4(ox*ss),m4(oy*ss))
          \            .logic(lets.expand(U=uv,V=uv).bicubicresize(m4(ox*ss),m4(oy*ss)),"min",U=uv2,V=uv2)
          \            .logic(lets.inpand(U=uv,V=uv).bicubicresize(m4(ox*ss),m4(oy*ss)),"max",U=uv2,V=uv2)
          \             .lanczosresize(ox,oy)
them   = yv12lutxy(clp,remove,"x y < x x y - "+DRK+" * - x x y - "+BRT+" * - ?",U=2,V=2)

return( them )
}

function m4(float x) {return(x<16?16:int(round(x/4.0)*4))}
##====================

Author
Time

That is awesome work, Spaced Ranger! will test your script and see how it looks in motion, I'm not so fond of the slight fuzziness it introduces going by that sample pic but it may be a worthy tradeoff I guess. Thanks! :)

Will post a sample of your script...

We want you to be aware that we have no plans—now or in the future—to restore the earlier versions. 

Sincerely, Lynne Hale publicity@lucasfilm.com

Author
Time
 (Edited)

msycamore said:

... I'm not so fond of the slight fuzziness it introduces going by that sample pic but it may be a worthy tradeoff I guess.

DeHalo_alpha seems to be the best they've got. I looked around for updates or other filters but none to be found. Even tried one that forced me to add sub-filter after sub-filter to eliminate successive "function not found" errors but it proved to be only fair at best.

So, I tried twiddling the parameters and discovered, for this application, the radius was too high. For the optimum setting, "5.0" is it -- lower and it doesn't do enough; higher and the "ringing" artifact starts to show.

                            rx=9.0, ry=9.0                                                                    rx=5.0, ry=5.0

As it's less drastic, the new correction isn't as overall soft, while doing a slightly better job reducing the leftover color fringing (no "added ringing" this time):

                     original                                        DeHalo @ 9.0                                             DeHalo @ 5.0

I'll update the previously posted Avisynth script parameters from "9.0" to "5.0" to keep the code correct.

Author
Time
 (Edited)

And, of course, I had to revisit that old proof-of-concept on the U.S. laserdisc edition of THX 1138 -- now that I'm getting the hang of DeHalo_alpha settings. Back then, I max'ed it out to wipe the edge enhancement entirely (it can't -- the author described it as a "trick" function). Now, after a feel for what the numbers are doing and a minimum setting just enough to show improvement, all the unsightly edge-peaking is nicely diminished. It turns out that the only setting away from the defaults was to "turn off" (0.0) dark-edge processing, to operate only on the over-bright edges:

This is the only line of code (along with the already posted DeHalo_alpha script) to do the job:

video_DH = video.ConvertToYV12().DeHalo_alpha( darkstr=0.0 ).ConvertToRGB32()

Author
Time

Yeah, that looks a little better than before but the tradeoff isn't worth it IMO but others may disagree. If it's something you all want me to apply then I will of course do it. But I would rather not.

Here's a video sample of Spaced Ranger's superb work on the lost scene, utilizing the last correction of his script:

http://www.sendspace.com/file/wdz2xg

Your fix for the color fringing made wonders, definitely a keeper, love it!

We want you to be aware that we have no plans—now or in the future—to restore the earlier versions. 

Sincerely, Lynne Hale publicity@lucasfilm.com

Author
Time

Thanks, glad to help.

Yep, it looks good! :)

BTW, I was looking over the sample shot from the U.S. & U.K. laserdiscs and discovered something interesting:

1991 US LD

1992 UK LD

1997 Bravo TV

2004 DVD R1

2010 Blu A/B/C

Notice? The top spacing of the doorways within the frames?
Lucas' Director's Cut of THX 1138 looks like it's from the U.K.-laserdisc source, not the U.S.-laserdisc source as one would expect.

Author
Time

I didn't even know THX had been shown on any cable network in letterbox in the late 90's. Where does that Bravo screenshot come from?

Forum Moderator

Where were you in '77?

Author
Time
 (Edited)

Are you talking about the different cropping between releases or is there something else I don't see? The mad revisionist cut was sourced from the negative whereas the letterbox LD's was most likely from an IP or IN. The UK LD comes from the same source as the US/Jap but is just a different telecine - different cropping and a less nr of frames missing between reel-changes.

We want you to be aware that we have no plans—now or in the future—to restore the earlier versions. 

Sincerely, Lynne Hale publicity@lucasfilm.com

Author
Time
 (Edited)

@ SilverWook

You remember this from earlier; pictured here (towards the middle):
http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/585765/#TopicPost585765
and dated here:
http://originaltrilogy.com/forum/topic.cfm/THX-1138-preservations-Italian-Cut-available-see-1st-post/post/585909/#TopicPost585909
Recorded on VHS tape, with "coily", but missing the "lost scene". Too bad ... "GL was here".

msycamore

The source is the question. George Lucas complained that his Star Wars original was falling apart (let's just play along with that claim for now) and Star Wars was years afterward. In what condition might the THX 1138 original be, as well?

Why would a director make a "director's cut" (aside from marketing to fill his coffers) and arbitrarily trim the frame throughout his otherwise non-problematic movie? Could it be the negative was gone (for whatever reason) and the choice for source was the U.S. edge-enhanced-DNR'ed(?)-mess master or the U.K. cropped-but-non-processed master?

Author
Time

I may be misunderstanding what you're saying but it seems that you're under the impression that they used a laserdisc transfer to make the Directors Cut. This of course is absolute nonsense - they made a new high definition scan from the original negative for the 2004 DVD release and probably reused the same master for the BluRay.

Author
Time

If anyone ever wants any 16mm or Super8 scanned to help with this, just let me know. I have a clawless flashscan type system, so worn sprocket holes are no problem.

Donations welcome: paypal.me/poit
bitcoin:13QDjXjt7w7BFiQc4Q7wpRGPtYKYchnm8x
Help get The Original Trilogy preserved!

Author
Time

1991 US LD and 1997 Bravo TV are the same source; the 2004 DVD R1 and 2010 Blu A/B/C are the same source as the 1992 UK LD.

The implication is that the original source was lost/damaged/whatever, and so they used the same source that was used in the 1992 UK LD.