camroncamera

Halos, a type of echo of portions of the image, within the image. Mostly prevalent at edges with high contrast.

Suppose there is a way to leverage SR technology to combat halos? If SR normally works by finding similarities in images and combining detail to strengthen true image information, could it not also be run "in reverse" to detect lower-detail echoes and suppress them instead of enhancing them, while still improving in true detail? 

DrDre said:

Here are a couple of screenshots at 1080p, without degraining, dehaloing etc.

http://screenshotcomparison.com/comparison/130449

http://screenshotcomparison.com/comparison/130450

http://screenshotcomparison.com/comparison/130452

http://screenshotcomparison.com/comparison/130453

 OMFG

_,,,^..^,,,_ said:

I think you should use the PAL DVD for ROTJ - at least, according to anyone in this forum, this is better than NTSC!

 Is this a title suitable for the PaNup technique (plus SR processing)? 

TV's Frink said:

 I'm not seeing much Phased Plasma Rifle in the 40-watt range.

 Fixed.

My own facepalm happens here. Clearly I had a basic SDTV spec fail. 480 would be a spec for SDTV vertical resolution, not horizontal resolution. I do still feel that a upscale factor of 4 might yield great results, glad you are open to experimenting with this.

DrDre said:

Thanks for the compliments! Glad to hear you enjoy the work and screenshot comparisons. I also often use mobile devices for the screenshot comparisons, because the zooming on mobile devices makes it easier to look at the details, than to put your face against your computer screen.

As with most techniques, there are many ways to implement the methodology. However, as a rough guide, a super resolution algorithm defines a reference frame, and subsequently estimates how objects, often defined as image patches, can be mapped onto the reference frame. The image patches are then averaged using some form of weighted averaging, where the weights are related to the similarity to the reference frame. The weighting reduces the chance of wrong detail placement, the socalled registration error. After the averaging the resulting frame is deblurred leading to the final result.

DrDre said:

I will try to create a video sample as soon as possible.

1920-wide upscales are possible, only it won't add much in terms of detail, so an HD-TV's internal upscaler will probably approach the same quality.

 Thank you for you detailed replies! One thought that I had is that multiplying the Standard Definition picture area by 4 would closely approximate "Full HDTV" 1920x1080 dimensions (though non-anamorphic black letterbox bars will not be contributing at all to vertical resolution). Would this not make the enlargement process computationally simpler than upscaling ~480 pixels across to 1280 across? A factor of exactly 4 (4 x 480 = 1920) seems as though it would be simpler (and perhaps visually "cleaner") for the computer to map pixels, rather than a factor of ~2-2/3.  The current method, though producing amazing results with a smaller output file, seems like there would be some heavier computation and possibly a SR upscale slightly less than its full potential, due to the ~2.66 x <480 computation.

EDIT: Clearly I had a major fail with my horizontal resolution spec. 480 is a vertical resolution spec. 

Would it be possible to see some 1920-wide upscales? Are the processing requirements too computationally intensive?

I love these AA improvements. Awesome work!

I continue to be amazed by these ever-more-impressive iterations. I am primarily monitoring this thread via my mobile device with retina-grade OLED screen. I especially enjoy the Screenshot Comparisons, because I can dynamically zoom in or out of the paired images and easily spot the differences when toggling between the screenshots. I do wonder about how my mobile browser handles these zooms, without "upscaling", but to my eye the image is simply being enlarged when I zoom in, preserving the screenshots as they were uploaded.

As a visually-oriented guy with no experience developing imaging software, I'd like to ask a couple of questions about what is going on "under the hood" when using a SuperResolution operation. I haven't (yet) read the SR scientific paper referenced in this thread, since I expect that it is likely to be full of equations that would be above my head.

Conceptually, though, I believe I understand how it works with motion picture images: for any given still frame in a sequence, the surrounding frames are used to recruit additional picture detail... Sort of like High Dynamic Range (HDR), but for image detail, rather than exposure detail. (I remember there was discussion that only frames following the given still frame could be used to recruit details for some reason).

So, if this is a somewhat automated process, is the software able to accurately track the motion of an "object" (such as an actor, prop, spaceship, or set) as it moves throughout a scene? I can imagine that a relatively stationary shot would be  rather simple for the automated process to work. Or, perhaps a simple camera move would also yield good results rather easily.  But is the software also sophisticated enough to recruit detail from a filmed object as it moves in 3D space in front of the camera lens, without incorrectly placing detail that does not belong? For example, if RD-D2's dome were to rotate significantly through the course of a shot that is to be processed with a SR script, how does the script not wrongly place detail of R2's eye-lens for those frames that the eye is facing away from the camera? Also, light and shadow and the sheen of metal and glossy surfaces can make pixel values change radically during the course of a shot, even for objects that only move a small amount in the frame.

If the SR process were a manual one, essentially a visual effects technique, I suspect a digital artist could take these filmmaking aspects into account and really work amazing wonders with custom-tailored upscales for each shot. However, this would obviously mean a great deal of manual work.

This bickering is pointless!

Although I too agree that there can be the appearance of over-sharpening and too much grain reduction, I am very impressed with this method of gaining additional detail from the GOUT. I don't own the GOUT DVD's, but I have owned the Definitive Collection LaserDisc set since its release in 1993. It's pretty incredible to see the additional detail that can be recruited. I hope that some strategy can be attempted that can keep the gain in true detail while also quelling the aliasing. Perhaps after rendering, two methods can be combined in a video mix using blending modes or something similar.

SilverWook said:

http://starwars.wikia.com/wiki/Aurebesh

http://starwars.wikia.com/wiki/List_of_use_of_Aurebesh

In case you ever wanted to really know what Artoo was saying to Anakin. ;)

And I do not want to see the bunny skyscraper in this thread...

 Here's what I learned about and where I downloaded the Aurebesh text used for my graphic:

http://www.echostation.com/features/aurebesh.htm

camroncamera said:

 These didn't seem to generate much of a laugh in the other thread that I posted. But honestly, this is simply how Disney should accomplish the next release:

What switches would each of you flip?

Tobar said:

The one that gets the text on Vader's chestplate correct. =P

Seeing as that's the SW77 menu, there shouldn't be any text at all on his chestplate.

Then there should be two other variations for each ESB and ROTJ.

None of which would use Aurebesh, as while it's loosely based on a SINGLE screen from ROTJ, is mostly an invention of the EU.

The original trilogy actually had it's own alien font that was used over all three films on props and the sets that has since been forgotten.

 Yep good eye. The Aurebesh text was just a little April Fool's Easter Egg... a deliberate creative decision. haha

Wazzles said:

mverta said:

X-Wings Dive towards the Death Star

_Mike

 Do you physically clean the prints before you scan them or is all your work done digitally?

 I can't answer for MVerta, but several years back I worked in a telecine bay and almost all the film we received (mostly 35mm and Super 16mm OCN from local commercial productions) went on the film cleaning machine before going up on the scanner. This film cleaner, however, had no bath. It used fluffy dacron rollers moistened with isopropyl alcohol. Even after cleaning, the freshly shot and processed negative would still have some dirt and debris. The scanner itself had slightly tacky rubberized rollers to further aid dirt removal as the footage rolled. After each reel or two was transferred, the dirty tacky rollers would be replaced with a freshly washed set. 

I can only imagine how much embedded dirt there is on 35-year-old projection print. I wouldn't even know what method of chemical cleaning - if any - would be safe for a Tech IB print.

77FN said:

Some of the ON was destroyed for good too, according to this article:

http://arstechnica.com/gaming/2013/09/once-lost-star-wars-related-short-to-screen-for-the-first-time-in-33-years/3/

"Tanaka: I remember when we were working on the Star Wars restoration, that was a different process. I think we optically recreated interpositives. But in order to do this, it went through some kind of warm chemical bath cleansing. The weird thing about Star Wars was that it was made up of different film stocks, so it went through this bath and they didn’t know what would come out on the other end...

 

Tanaka: There’s a space battle shot and a close-up on Hans Solo, and the original negative is coming out of this cleaning solution and it’s just acetate.

Parker: It’s all clear. Oh no, did the bath dissolve it?

Tanaka: Yeah, it dissolved it, depending on the film stock."

 I get the impression that they are referring to the shot where Han Solo is in the turret gunner seat and the frame is rocked side-to-side as an optically-printed animated effect as the attacking TIE Fighters blast away at the Millennium Falcon. Pretty sure that's how it plays in the GOUT, anyway. That shot could have been printed to CRI from camera negative and then the CRI subsequently faded away to clear years before it even reached that cleaning bath. If that's the case, the actual OCN for that shot - without the animated shake - might have been in the same faded-but-usable condition as much as the rest of the OCN.

emanswfan said:

SilverWook said:

They are people too. If they prefer the SE's (all three versions and counting) and the prequels, then more power to them.

It's when people insist the OOT stay in the vault I can't fathom. I can't see Disney passing up a revenue stream that's been untapped for years though. If they have the rights free and clear that is...

 Completely agree.  It's less that the SE's are bad, and more so that the OOT just isn't available.

Gosh, I really hope with all my heart they announce something at Celebration.  It's far more important to me than some new TFA trailer right now.

 These didn't seem to generate much of a laugh in the other thread that I posted. But honestly, this is simply how Disney should accomplish the next release:

What switches would each of you flip?

"Who's the more foolish, the fool or the fool who follows him?"

slask said:

Anyway, I'm still convinced a "combo" box is the best solution: everyone will stick to its favourite version.

 I can't think of a combo box better than this one:

Joel said:

TServo2049 said:

OK, perhaps SW was scanned at 4K, but I thought digital intermediates were still at 2K? Correct me if I'm wrong here.

I seriously doubt it was scanned at 4K for the SE in 1996/97. 

Snow White aside, I worked for Technicolor when the DVD masters were made and that was *right* when 4K was being widely adopted, around 2004. Also, if they already had a 4K scan of all three films why go back and pay for a new 1080 scan for the DVDs? 

 Agreed. In fact I believe that ILM was doing heavy R&D and built their own 2K scanners at the time. Truly groundbreaking accomplishments for its day, but as mverta has said, film scanning technology has come a long way since then. Which makes sense, considering it is nearing 20 years since the Special Editions have been put together. 

I see that msycamore has already shared the youtube video of the Tech I.B. process, but here is a short associated article to go along with it:

http://www.redsharknews.com/technology/item/2371-remembering-technicolor-dye-transfer-printing

ww12345 said:

Whoa, I never knew that was supposed to be a circular building!

 Agreed, I am just now seeing that for the first time. 

"They were cones!"

CatBus said:

Alderaan said:

I think watching Star Wars is a little different than watching something like The Godfather, don't you think?

Absolutely. One of them looks and sounds like it did in theatres (although some argue this point because they don't think the yellow tint is from bulb-matching), and one of them clearly doesn't, which is rather the point.

EDIT: If you're talking about "looking dated", you simply need to look at the haircuts.  They are movies from the seventies.  The VFX, editing, haircuts and fashions are all dead giveaways.  I'm sure it's possible to give everyone a digital haircut, freshen up the fashions and color schemes a bit, do some more jump cuts and fewer wipes, and remove the matte lines, but that would just turn it into a seventies movie trying to be a modern movie (and failing).  Which, in a lot of ways, is one of the primary problems with the Special Editions.  To focus exclusively on matte lines is really no different than focusing exclusively on haircuts.  It's the same "I don't like the look of films from that period" writ small.  Not that it isn't a valid opinion to have, but that's what it is.

It's also not my intention to be dismissive of your opinion, just of the idea that you can extrapolate/evangelize your opinion to what everyone else might want. Heck, I wouldn't mind changes to the originals myself (no cue marks, no gate weave, no burn marks, some dirt & scratch removal), but I know there's plenty who would disagree with me on some or all of these.

 Largely I agree with Alderaan, in that newly composited original elements (but no CGI embelishments), put together as a "Showcase" version for a modern DCP/Blu-Ray/4K, would be ideal for a wide release of the Classic cuts. Though CatBus makes a perfect point about preserving the film(s) as a product of their time (and where would Disney hypothetically draw the line should they decide to play in the Lucas sandbox?), I believe tidying up the original compositing would eliminate the biggest remaining distraction for modern audiences. Everything about set design, hairstyles, costumes, etc., should remain as it is, because, those are aspects that contemporary audiences *expect* to see, and will not find jarring, in the way that a blurry landspeeder (1970's VFX) or ronto creature (1990's VFX) would on a modern screen. In other words, cheap sets and costumes, by themselves, don't degrade the image quality in the way that photochemically-composited VFX may have (in some cases). I think that most of us here, though, understand that the original VFX needs to be acknowledged and preserved, whether seamlessly branched or presented on an alternate disc, as a concrete duty to film history.

lovelikewinter said:

Handman said:

Roger Ebert agrees: http://www.rogerebert.com/rogers-journal/lucas-demonstrates-potential-of-digital-video-with-attack-of-the-clones

He seems to claim that AOTC's fuzziness on film was part of Lucas's agenda to push digital projection.

 

I remember that.  I saw it opening day after school and it was fuzzy.  Of course theforce.net tried to blame it on my theater.  

Maybe they stretched pantyhose over the projection lens, for that soft 70's look. :P

Tobar said:

I wish I knew how much of a delay there was between the two of them. That seemed like a mighty long pause before he responded...

I hate watching broadcasts of live interviews conducted by satellite. The satellite delay makes a conversation seem very awkward, and the participants don't always seem to understand that the interviewee is hearing the voice of the interviewer a couple seconds later than when the audience does. This gives the impression of an unresponsive guest.

As far as the subject mater of their talk, I would say it is quite encouraging.

moviefreakedmind said:

I recently watched the new youtube video from Zombie on Star Wars News and Views about the 4K restoration. He mentioned the resumes from RMW workers so I looked it up online and found a linked in profile where the lady lists that she returned to RMW to work on "frame by frame image restoration" on Star Wars in early 2013.

m_s0 said:

Yeah, someone pasted her resume in one of the threads here. Believe me, every bit of information available out there that could've been discussed already has been. Ten times over :) We need new info.

Just conveniently posting the link here to said new Youtube video:

https://www.youtube.com/watch?v=kwcR4x1c2pc&list=UUXbMeL5_F6HWln8_sWA0PGg