logo Sign In

Laserman

User Group
Members
Join date
11-May-2004
Last activity
6-Sep-2007
Posts
903

Post History

Post
#250861
Topic
Info: The preservation of technical flaws re the X0 project (aka Boris is DanielB)
Time
That is pretty much the case, laserdisc never really took off in the PAL territories, so releasing a killer and expensive player didn't make sense.
There was no reason for an NTSC owner to really want PAL capability back then either, their TVs wouldn't cope with it and there were very very very few titles that got a true PAL release anyway.
Making the X0 player do PAL would have cost a fortune in engineering, and had no real market. It just didn't make financial sense.

BTW I was reading that Lucas interview just now and saw this:

GL: These films are incredibly difficult to make. Normally, a director is concerned mainly with character and with telling a story. In the Star Wars films that is important, but equally important are all the details. They’re like little time bombs all over the set, thousands of them, and if you don't catch one it could do you in. When the shot moves around and there's some little thing that isn't right, it could take the audience completely out of the movie. In a normal film there isn't that thin edge. Reality, the reality of the world we know, is a tangible presence in most films. The viewer is there, it's real. But in a film like this, where we're creating a world that doesn't exist, it's very easy to puncture a viewer’s sense of reality by a missing or wrong detail.


Pretty much sums up why a glitch free preservation is important to me.

Thanks for the comments Ghostalpha, it isn't really a case of justifying myself, just letting people know what makes me tick.
The other unmentioned result I wanted from the X0 project was to inspire other people to have a go, and share some techniques. People that play with this sort of thing may end up getting interested as a career option or at the least learn some stuff and then contribute back into the community again.
Post
#250840
Topic
Info: The preservation of technical flaws re the X0 project (aka Boris is DanielB)
Time
Originally posted by: boris
I still contend that I was posting on topic in the original thread; as I was expressly about the X0 preservation.
Originally posted by: Laserman
If I didn't see the glitches at normal playback speed, I wouldn't fix them - I'm not bothering to fix any glitches I don't see in normal playback.
What I mean, is that it wasn't ever intended for customers to go through and fix the glitches frame by frame. In actual fact, it was never intended that customers fix them at all - but I chose to put it this way since this is the way in which you're going about it. Of course you can see the problems at normal speed, as can I. I put it to you, though, that film colourization in essence was the process of making the movie appeal more to today's audience; and that in essence cleaning the mattes and other technical flaws is also the process of making the movie appeal more to today's audience. You may not feel it's the same thing - the way Lucas may not feel that the Special Editions are the same either.

To put it simply, today's audience is spoiled by newer movies with more advanced effects, using the latest technology. And the average cost of producing a Hollywood movie today is much much more than it was in 1977. But in 1977 viewers had never seen the effects we now have today, and so seeing problems like matte-lines were second-nature. Today though it would be considered below par if a film was made which had obvious black matte lines. Therefore it is now second-nature to think of older films which have these defects in them as being "below par". I wonder if anyone will ever really be happy, if all they ever think about is "this is how it was meant to be"; and I wonder if the movie was really was "how it was meant to be" the first time, would it have been as grand a masterpiece as it is?


It isn't about making it acceptable to today's audience like colourisation. You seem to have a bee in your bonnet that colourisation is the same as removing a garbage matte.
It also isn't about making a movie "as it was meant to be" if they had unlimited technology. It is about taking the movie that was actually made and fixing some errors - and even that only *after* a straight untouched preservation is completed.

The problems I see in the film are the same ones that were noticed by people when the film was released. As I said before my personal version is about removing problems that jar *me* out of the viewing experience. They jarred me out of the film in 77 and they still do today. It is not about changing the film to suit a different taste, a different audience or updating to a particular fashion (like crushing the blacks on the SEs to make it look like a current film, or colourising a film to suit a new audience not used to watching B&W)

Also, one point seems to be repeatedly missed, the difference between a preservation, a restoration, a personal restoration and a fan edit.

Backing up the laserdisc is a preservation effort, which we have done and completed. As far as the DC version of the discs go, it has now been madde somewhat obselete by the release of the GOUT, although it didn't preserve the opening crawl from the laserdisc version. Backing up the SC of ANH is also done. This is still very relevant as it is the only consumer copy of the OUT with no DVNR screwing up whole scenes in the film.
So to repeat, those archives are *done* so no-one needs to stress about it. The first aim of the X0 project was successfully completed.

As presevation goes though, if you stop at that point, you have only preserved the laserdisc, not the experience of seeing the film, or in effect the film itself.

What I mean by that is the laserdisc has artifacts that were not in the film, or at least not in a good first run print of the film, it has the DVNR, has lots of dust and scratches, droputs and film damage that is not going to be on the negs or a first run print.
It also has considerably different colour. The garbage mattes are much less visible on the film prints also, and there are a lot of subtle differences between the LD versions and the theatrical prints.

By then going beyond a straight LD archive and undertaking a 'restoration' effort, you can take the various source footage available, and match it back to a theatrical print.
In effect bringing the laserdisc versions much closer to the film itself - undoing some of the problems created in the transfer to laserdisc in the first place.

People might think cleaning up the film damage is somehow 'not original' but every film print struck will have differences.
When restoring old paintings in Museums sometimes new paint has to be added to get it back to as close to original as possible - a pure restoration project is a similar process.
Adjusting the colours to a known print, and removing film damage is just getting it back as close as possible to original.
This process is regularly undertaken when archiving films for the AFI etc.

Then you get into the area of a personal restoration effort.
If the original is archived, then you are not taking anything away by doing a personal restoration.
I consider a personal restoration effort that includes fixing matte lines, garbage mattes and other glitches to still be a restoration, albeit one of a slighlty different type.
Fixing items that are purely post production errors or glitches differs from a reworking like the SE in one major area.

For a 'personal restoration' The 1977 theatrically released film becomes the script for your project. You are not changing any dialogue, any character interaction, any creative decisions, removing or adding any scenes. You are not adding characters, changing music or sound, or altering the mood through colour correction.
Colourisation however is akin to an SE style change as B&W shooting requires you to make particular choices about how the movie will look, and to colour them arbitrarily later totally changes the movie.)
In short for a preservation the film's story, pacing, characterisation, length etc. etc. remains unchanged,( mess with any of that stuff and you have a fan edit).
For your preservation you are taking the finished product and purely fixing errors. This is different from trying to 'guess their intentions' about a scene. It is easy to know what is a glitch and what isn't. Trying to guess how they *might* have shot a scene, or what they *might* have added to a scene if the tech was available is a totally different thing.

The limitations of the technology made the film be shot in a particular way, certain camera angles used and so on.
A personal preservation effort doesn't try to change any of those things as they make up the way the movie was shot, and no one can say how the movie would have been different if those limitations were not there.
It keeps the film locked into the way it was shot, but removes any technical errors.

One way cleans up glitches out of a movie but leaves the movie itself unchanged, the other changes the story and it becomes an entirely different movie.

Then of course you get into the grey area of fixing other types of errors, like continuity problems (like the disappearing/re-appearing cloak in Kenobi's house), like the R2-D2 being blue when seen behind Luke, but black in the next scene, the jump cuts for some of the lightsabre on/off scenes and so on.
I still think this belongs more in the restoration camp than in a fan-edit category as once again you are not changing any part of the story, characters etc. You are changing it from the original in that you are removing or colouring elements within a particular scene, and you have to really do your research to ensure that you are not changing something that was meant not be there (for example researching to find out that R2 was only black because of the bluescreen, that he wasn't black because he was 'in space'. That the green blobs are unwanted garbage mattes, not 'shields' on the ships).
People's opinions will always differ on this sort of stuff, but the idea is to be as careful as you can, and if you can't be *sure* then to leave it how it is.
Then you end up with the same movie, but distraction free.

A project definately moves out of the realm of a preservation effort once you start 'editing' the film. Cutting or adding different scenes, or changing the grade to change the mood (like making the tantative very cold and blue), putting in new characters and so on really becomes a re-edit, or a different cut of the movie, and belongs in the realm of a fan edit.


All of this is semantics anyway, people will always have different ideas of how to categorise something, (and could argue it forever) but to say fixing a garbage matte is the same as colourising a movie, or the same as doing a total re-edit like the SEs is beyond stretching an analogy.

I will state again to make sure there is no confusion.

1. The X0 project was about archiving the footage first (done) and then creating our own preservations second (in progress).
2. The laserdiscs are not exactly what you saw at the cinema.
3. Different prints abounded even in the 70s so there is no single, definitive ANH as such, but they all were essentially the same and nothing like the SEs.
4. Our preservations are being matched to a set of theatrical prints to get as close to the 1977 cinematic release as possible.
5. The GOUT and the DE LDs are a long way from the theatrical prints I have seen both in colour and of course the massive loss of detail and artifacts caused by the DVNR process. The SCs are much closer to the prints.

Arguing what category a particular effort falls is somewhat pointless as anyone can call them what they like, and belongs in threads like this rather than clogging up project threads with endless arguments.

I don't think I (or the X0 guys) have ever tried to muddy the water as to what we are actually doing though.
Post
#250824
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Mielr
I think Boris (or anyone starting from scratch) would be better off using the DVDs, rather than using the LDs and buying a $?,000 LD player.


If the SC or any version without DVNR was released on DVD that would be true, but if you don't use the other laserdisc versions then you are stuck with 4 eyed stormtroopers and the smearing and trails.
Post
#250639
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Yes colouring a few panels that were meant to be blue is different from colourising a film that was never meant to be in colour.

Visible garbage mattes were not *necessary* they could have been avoided even in 1977, they certainly were not meant to be visible. They are highly visible on the GOUT, go calibrate your TV.

As I said, my changes are for me, and I am restricting them to things that were inarguably not meant to be seen on screen. This is different to making new *creative decisions* which is what you are implying. Most of the 'extrapolations' you are making are changing creative decisions.
Others may decide to make all sorts of things including total re-imaginings, that is fine by me. As long as the original is preserved then people can do whatever they like for their own viewing pleasure, it doesn't hurt anyone else.

And Laserman, I'm aware you're doing it both ways, and I have been for quite some time - I just don't understand why.
Then you must have trouble reading, I spelt it out in detail in the last post, and in many previous posts.


Your project is supposed to be a preservation project, and as I understand it you feel that Lucas can make changes to his films - but that he should also preserve the OOT, correct? Well that's the way I feel


Good then we agree on something.

If you have any real questions I am happy to answer them, but I'm not going to answer any more of the type that I have already answered ad nauseum. Please re-read the posts in the thread, it is all covered in great detail.
Post
#250603
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
They are in lossless Huffyuv and the only processing done was an IVTC pass to get them back to 24fps progressive.
Re the pre THX versions (like the SC) we are doing a lot more processing on them to try and get the scenes ready to drop them into the X0 capture to replace the DVNR affected scenes.
We should have some stuff on the site within the next week or so (it's a balancing act between writing site stuff and getting the project done) showing that exact process on some of the SC footage.
We currently have a 2 stage food chain for the SC, basic cleanup and noise reduction and then the 'blackmagic' process to combine various sources to get a really clean detailed version. We don't want to go too far in stage 1 and end up losing the detail as we need it for stage2.
Post
#250469
Topic
Info: Auto-correction from SE colours to GOUT colours (lots of information)
Time
Originally posted by: superrune
What if the 2004 version was digitally stabalized reducing camera wobble, and the OUT wasn't?


Sorry about my explanations, english is not my first language.

With wobble I mean that the information within the frame is shifted in relation to each other, like distortions under water, and that this relative shifting differs from one frame to the next. If the movie was doing something like that it would be very obvious. That's why I questioned the method of tracking several areas of the frame, instead of a single location (the stabilization you need if the frame is just shifting). I would assume the warping that is in the frame to be lens-distortions from the old telecine (at least that's what I think I see), and perhaps some corner squeezing in order to get more of the image inside the safe frames.

Some noise-reduction routines and DVD compression can produce an effect similar to wobble, making the dark areas lag one frame behind the rest of the image. But I haven't examined the movies enough to confirm that particular effect.

Rune


It is a bit of both.
Some scenes require only a 'set warp' to align them, but most require a tracked warp.
I think this is because we have multiple problems.
1. Lens distortion - if this was the only problem then you could devise just one warp for the whole transfer.
2. Film movement in the gate. Both sprocket wobble and the film not being perfectly 'flat' causing planar distortion in Z rather than just wobble in X&Y
3. Stabilisation differences.
It is easier to track the warp to get the 'set warp' done anyway, rather than trying to do it manually.
A simple warp would probaly work OK though, the resulting artefacts probably wouldn't disturb 90% of viewers, especially when moving.
It would certainly be worth a try, and would be a lot less effort.

The 2004 release seems to be somewhat stabilised, this is why I stabilise all footage before combining, so that at least movement within the frame is locked down and you aren't chasing a moving target.

Boris the flaws you discovered where already clearly pointed out by Rune and myself beforehand, as Rune said when mentioning the limitations of the process in his first post he said that just "Transforming the GOUT is too simple, you need a warp to properly match the two images"


Post
#250467
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: borisUntrue. Monty Python and the Holy Grail has no delted scenes - every scene they shout for the movie they used in the movie (of course there'd be out-takes, though).

Sorry Boris, you are wrong, there is at least one deleted scene (the Anthrax bit) and it was deleted because it just wasn't very good (personal opinion after watching the deleted scene). Most movies have a ton of scenes that never make it. It is called editing.

Scripts, from what I hear, are about 120 pages long. That's less then 1/3rd the number of pages in a novel - and from what I've seen of scripts they contain much less words on each page than a novel does - and many of those words are wasted on describing lighting and costumes and such. I don't agree that they start out deliberately with this flexibility to remove stuff - that's just stupid.

Sorry Boris, you are wrong on a few counts here too.
All films (there may be an *extremely* small number of exceptions) have a shooting ratio. i.e. the number of scenes shot vs the number that make it into the movie. Many scenes are shot *specifically* to have the flexibility to swap scenes in or out later - it depends on the director.
An extremely efficient director may have a ratio as low as 4:1 (i.e. they shoot four times as much footage as they use). John Cassavetes is at the other end of the spectrum. For his film "Husbands" he had a shooting ratio of *over 100:1*. He shot well over 200 hours of footage for his two hour movie. For "Faces" it was around 70:1.
So what? of course they weren't intended to be obvious, but then again neither was leaving Jabba in ANH on the cutting room floor. Lucas had obviously intended to put it in if he'd shot the scene...
Not necessarily the case at all, there were plenty of scenes shot for Star Wars that never made it into the film.

When I worked on Babe Pig In The City, a *lot* of scenes were shot and never used, we even built a beautifully elaborate set for the opening sequence, had the motion control camera all set up, did a few test passes and then the director changed his mind.
He decided that as it was a sequel the opening sequence was no longer required, so it ended up not being shot, even though we had spent 3 months setting the shot up and were ready to roll camera. Had he been on another part of the set that day it would have been shot anyway, and then in edit may have ended up in the film.
Pre-vis has allowed shot ratios to be reduced in some instances, but it is a staple of film making to get as much 'coverage' as possible. Films cost big bucks, so shooting some extra scenes when you already have the set, the actors and the crew in place makes a lot of sense. Pickups and reshoots cost a fortune in comparison to grabbing an extra couple of scenes when you have the chance.
Even in post a lot of CG work was created that was never used, but was wanted so they had options when going into the final edit.
No one really know exactly how a film will turn out, so it makes sense to have more scenes available than you finally use in case the script just doesn't work when it becomes a movie.
Just on the script thing scripts are about 90 minutes long for a typical movie. 1 page = about one minute of finished picture, so a 90 minute film has roughly a 90 minute script. A 120 minute film would have roughly a 120page script. Scripts change massively throughout the course of the film, you see scripts with all sorts of colour coded pages to let you know where the (constant) revisions are. No one except really broke people (or arty guys) just shoot the V1.0 script and no scenes deviating from it.



I think preserving the matte issue is kinda ridiculous. We're not talking about something that is a result of special effects limitations of the day. You don't see similar issues popping up throughout the film (not that I'm aware of). Somebody screwed up, and the crappy mattes made it into the finished film. As sort of a side thing, I can understand the preserving of the mattes in order to keep that "opening night" feel, but as far as the finished film, no.

Boris: "Somebody screwed up"? No, that's just an unwanted by-product of the special effects of that day. It's like filming a TV... it may flicker... or filing a helicopter - depending on the camera used the helicopter blades may look fluid - or they may look uncharacteristically stilted. The mattes, on the other hand, are the result of somebody in the SFX department being asleep at the wheel.
Boris: They were a necessary by-product of the special effects. Just like the black R2 unit used in blue-screen shots, next you're going to say he should be digitally colourized?


They weren't a 'necessary' by product, just an unwanted one.
We have always said, the straight X0 captures were done and archived in the name of the purists that want all the glitches intact. It is interesting from a technology point of view, and therefore the glitches are worth preserving for that alone.
True the matte lines were somewhat visible at the cinema, (and still were when the SE's were released in cinemas!) but far less so than on the laserdisc/video and DVD releases. However, they were never intended to be an integral part of the film, they are akin to a processing error, or a scratch on the film, i.e. something that may have been unavoidable, or missed, but definately unwanted.

Boris:
Besides which, if you've seen the OUT DVD you'd know they're barely visible at all, except over other spaceships (and how do you expect the X0 team to fix those ones? but levelling out the blacks... that won't do there!)

I found the Matte lines very visible on the GOUT, not sure what you were watching if you can't see them? Perhaps you need to calibrate your TV.
Fixing the garbage mattes over other spaceships is simply an application of a little roto and some colour correction. Easy.
Tedious, but easy.

I repeat - straight captures of the DC and SC are done, they are archived and finished, so purists can rest happy - we will not be 'cutting our negatives' so to speak.
The straight laserdisc captures will stay as they are in all their glitchy glory, matte lines, jump cuts, miscoloured sabres, garbage mattes totally untouched. This was always one of our goals.
i.e. The transfer is a straight calibrated capture from the X0 so it can be watched without side changes and looks better than the laserdiscs did for anyone except the X0 owners!

But for the cleaned up version we are focussing on fixing just the glitches that should never have been there and make the movie harder to watch now.
That is just fixing the problems that were *definately* not intended to be in the film when it was released. Not adding any new creatures, or changing story arcs, or throwing in rontos or messing with the story in any way.
However removing matte lines, film burns, garbage mattes etc. definately brings the movie closer to what the guys in post were trying to do, but in some way failed (but not taking it to the level of what they *might* have done, which is where I see the SEs).

The difference is we are just redoing exactly what they actually *did* do in the film with better tools (e.g. redoing an optical comp digitally).

Boris: Cleaning out matte lines/boxes is just like colourizing a film.


Not at all. Colourising means *totally* changing the intention of the DOP and the director. B&W films are shot utilising the medium for certain effects, and a certain look - and colourising changes the intention of the film maker drastically.
The films made in black and white were intended to be made in black and white. Had a director been *able* to make a colour film, he may have chosen to do so, but would have shot and composed it completely differently. He would not have just made the film the exact same way but in colour (unless the DOP was truly useless).

This is the core difference between a 'corrected' or fixed version of the OT vs a re-imagining of the OT which is what the SEs are.
One fixes technical errors and byproducts of the optical processes of the day, the other changes the story, composition, characterisation, pacing etc. Two very different agendas, and two very different results.
But at any rate, the archive capture remains untouched.

For my own personal version I am also fixing jump cuts and other effects that look cheesy to me today, and once again, obviously are not meant to be a part of the film.
On the weekend I was deciding wether to colour R2's panels or not (for my own version) as I find it distracting that they are blue when seen behind Luke's head in some of the 'in cockpit' shots, but are black in most of the external shots.
For me I want the film to be watchable with no moments that pull me out of the viewing experience, wether that is a poorly executed jump cut, or a horribly obvious new CGI addition it has a similar effect on me - it pulls me out of the movie.

So just to be clear, (even if it is on the front page of the thread for all to see) we are keeping an archive of the captures untouched as a 'purist' archive, and are doing a 'technical errors repaired' version, as we have stated many times.

So *please* for any future discussion on morals, idealogies, scriptwriting, film making etc. start a new thread for them Boris.

We are getting into the home stretch for ANH and I would like to keep the discussion focussed on the project happenings.


Post
#250131
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Dufusyte
Thanks Zion.

Btw, may I suggest that XØ Project place at the top of their work list the scenes that have been altered in the SE, for example the Greedo shooting scene, the end of ROTJ, etc. This way, in the unlikely event that the XØ Project dissolves before completing all three movies, at least it will have produced clips of the original movies that are most needed for possible insertion into other projects (see my "Dream Project" above - Rune's type project would need polished clips of the "un-SE" version of the scenes Lucas has tampered with).


Personally I want to see ANH completed in its entirety, but for the other two movies I will be focussing on doing just the changed scenes first.
Post
#250130
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: Dufusyte
Gentlemen of the XØ Project, I would be interested in your thoughts on this approach. My dream project would be to record the SE version being televised this November in HD, then color correct it using the GOUT, and for frames where the SE is divergent (extra cg critters etc), use restored frames from the LD or GOUT (which ever cleaned up better).

I applaud the concept of preserving the best LD possible, but I think we are all primarily interested in preserving the original Movie, as opposed to preserving a particular media (LD/DVD/VHS/reel/etc). You are such a talented crew on the XØ Project, I am mildly concerned that strictly limiting your focus to the XØ LD transfer may no longer be the best strategy, as far as restoring the original movie is concerned, not to mention it is an extremely time consuming and tedious approach. Are you guys primarily concerned with restoring/preserving the Movie (regardless of source media) or are you more interested in preserving the LD (even if you recognize it is not the best way to produce the highest quality restoration of the movie)?

Thanks again for your great efforts. I hope they are spent in the most efficacious manner.


I posted an answer of sorts over in that thread also, there will be an article shortly on a similar approach that we have been messing with for a while.

As to your last question, my main aim is in preserving the movie.
The problem in the past was the laserdisc and DVDs were far too far apart in terms of quality to do a hybrid style project like ocpmovie attempted. I applaud his attempt, but I found the result unwatchable, as I did with my own attempts to do the same type of thing. (i.e. mix the LD caps with the official 2004 DVD).

I realised early on that the only real hope was to get the original elements in whatever form possible and find ways to squeeze the highest possible quality from them.
So although from the outside we look obsessed with sticking with the laserdiscs, the fact is for some scenes the laserdiscs are the only source you can find them (other than a 35mm print). The blackmagic system and some other tools are a direct result of us sticking with working with laserdisc technology and working out ways around its all too obvious limitations.
Hopefully the resulting images will be of a quality to make a hybrid project possible. At the very least we need the SCs and other disc sets captured, cleaned up and stabilised to be able to use them even as an 'automated colour reference' for other sources such as the official DVDs for example.
No matter which way I look at it, it always seems to come back to needing the laserdiscs processed as part of the equation, so we might as well finish that task before moving on to other restoration projects.

Like in production, we are building up toolsets that will be valuable for the next phase whatever that might be.

My hope is that when we are finished with the LDs we will have a suite of tools (and a selection of scenes) that will allow us to tackle a hybrid restoration with a detail level close to or exceeding the 2004 DVDs.
The truth is that a year ago, we didn't have enough to do much more than an archive and touch up of the DC. Embarking on anything else would have been premature and have resulted in something interesting but probably not watchable along the lines of OCPs sets.
It would have been a case of "You're all talk, Hamill. You never even finished Jedi school."

The road has been long, but the lessons learned have been very useful, and the input and ideas from everyone here have been indispensable.
So yeah, I intend more than just a laserdisc restoration, but I still think it is the necessary first step.

I'm looking forward to the HD broadcasts in November to add to the mix...


Post
#250128
Topic
Info: Auto-correction from SE colours to GOUT colours (lots of information)
Time
Originally posted by: THX
Impressive... most impressive. Presumably with this system you could incorporate PAL and NTSC captures?


Yes, but of course the different framing causes issues. I don't want to hijaak this thread, so I will try to keep comments useful to this type of process.

With the PAL discs the framing is a little tighter (i.e. there are bits 'missing' from the extreme edges comapred to the NTSC discs) which leaves you with two choices.
1. Crop the frames to match the PAL versions
2. Use a soft rolloff matte (basically a 20px or so gradient) on all four sides to allow the resultant image to blend into the NTSC image to restore the edges.
I've tried both and sometimes option 2 works perfectly, other times it doesn't, if I was smarter option 2 would work more often I suspect.

And vbangle, to be fair, my drafts are *awful* and poor Zion and others have to spend a lot of time turning them into the professional looking articles that he produces.

My mum use to say 'you can't polish a turd', but Zion often surprises me!

Rune, I'd love to see your files and find out if you can feed the tracker to whatever the morpher/warper is in fusion to get better internal mapping between the two streams.
Post
#250082
Topic
Info: Auto-correction from SE colours to GOUT colours (lots of information)
Time
Nice stuff!
We have been using a similar technique for a while in a system we developed that we have nicknamed 'blackmagic' that uses Shake and a bunch of custom scripts to allow the combining of multiple sources to get the best out of it.
Basically we are taking in the various telecines (i.e. the various laserdisc releases), do a quick transform to get the corners roughly aligned and then doing a stabilisation pass to try to remove as much gate movement etc. as possible (this simplifies the next steps enourmously).
We then pick one transfer to be the reference master and, using the tracker, pick suitable track points in each transfer. Once a solid tracking solution is found, we then feed the trackpoint data to the warp (morph) function to get the picture elements to align within frame as close as possible.
Then stealing an idea from the astronomy boys, try to treat it as a multipass capture of the same source (which it basically is) and create a composite image throwing away noise and gathering detail to get a final result that is a lot cleaner and more detailed than any of the individual laserdisc transfers are on their own.

I've been writing up the process and sent a draft of it to Zion a couple of weeks ago for him to turn into one of his slick looking webdocs for the X0 site, so it is great to see someone else looking at a similar process using different software.
We had been discussing wether using free/cheaper programs the same thing could be achieved (although Shake recently became a lot cheaper anyway).

You should be able to achieve the same thing in Digital Fusion, but I haven't played with Fusion for years, so I'm not sure how much control you get with its tracker and warp tools (does RE:Flex run on fusion?)
I remember back in the mid 90s when one of the guys that wrote fusion was demoing a beta of V1 at a little trade show. I watched him run it through its paces and the damn thing was so *fast* for the time. I went up to him and commented that "now that is why people spend so much on SG workstations" he lifted the tablecloth and showed me the standard PC it was running on (it was a 486 DX4-100 or a Pentium 60, I can't remember which) and I was literally standing there with my mouth open.
It was the first time I thought PCs might actually be able to be a real force in post. It still astounds me how fast they came on. In 95 if you had said to me that 10 years later people would be able to re-edit and add effects to feature films in their bedrooms on their home PC's and see most of the results in *real time* I would have laughed out loud. If you had said they could then put it out to a DVD atthe same quality that the studios get I would have thought you were truly crazy. It has been an amazing transformation for just a decade.
Post
#248642
Topic
OUT: PAL or NTSC?
Time
Originally posted by: Darth Editous
Moth3r:
I don't know for sure, but I suspect that CRT TVs sold in PAL countries have shadow masks/phosphors optimized for ~576 scanlines.


I have to say I doubt this. I don't think a CRT could ever be "calibrated" in that way since the tube is affected by the Earth's magnetic field. If it was in any way "optimized" everyone would get a sub-optimal picture when they got the set home (especially me - my room is near a small electricty substation and my picture is sheared by a few degrees).

DE


They certainly are affected by the earth's magnetic field. We had to recalibrate all the magnetics in a bunch of broadcast monitors we brought in from the US.
A CRT from the Northern hemisphere is setup for the earth's magnetic field in that hemisphere, take it south and you get all sorts of colour problems. Same is true in reverse. You don't wanna import CRTs across the equator unless you know how to adjust them at the other end.

Back on topic the scaling on the PAL versions doesn't appear to be done very well, and some form of vblur has also been applied by the look of it. That is why I sent my PAL version back and got the NTSC version, I'd rather do any upscaling myself and get it as right as possible.

Post
#248622
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Well, a few posts slipped in while I was writing mine.

Just for the record, I don't hate Lucas, I don't hate the SEs (although parts of them drive me nuts), but I don't enjoy them either.
I don't hate the prequels, but was very disappointed with them.

I am highly disappointed with the GOUT release, I always think if you are going to actually do something then you do it properly. If you are going to release the OUT then at least do a new telecine of it with current technology and make it an anamorphic release just the way any other major film from the 70s has been treated on DVD.
Matte lines, corrections soundtracks etc. are all open for personal preference, but doing a direct dump of a 1993 laserdisc master tape is just a poor effort.

The film deserved at least the same treatment that a standard film to DVD transfer gets.

I think it is pretty clear where I stand on all this stuff anyway, so I'm not going to go on about it any more here. If someone wants to start up a 'philosophy of the SW fan' thread then I'll chat there if I get time.
Post
#248619
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: boris
What I find disturbing is the fact that so many people here intend to freeload off the X0 project as an alternative to LD's/DVD's. And the fact I can't even post in fun anymore without being told I'm a troll.

I know you're saying that you're cleaning up the video frame-by-frame, however Lucasfilm did this in 1993, and *most* of what you're cleaning up is LD defects, rather than film defects.
No actually Lucasfilm didn't do any kind of frame by frame cleanup for the 1993 laserdisc release - that is obvious from watching them.
And no, most of the defects we are cleaning up are film problems (dirt, dust flecks, scratches, splices, cue dots, film burn). We are cleaning up the laserdisc problems as well while we are there.

Originally posted by: boris
Yes I'm a critic of this project, I don't think it's a bad project or anything - but I just think that a lot of forum members here are:

1. seeing it as an alternative to the OUT rather than a preservation...

and/or

2. seeing it as being much better then the OUT.

I'm posting here as a critic. The project does have some fantastic points about it, don't get me wrong... I just don't think it would be worth it to me to spend as much time as you 4 have preparing a preservation which is only going to be slightly better then the retail release, at best.

I think we constantly make it clear that this is a restoration effort, not a commercial product to buy instead of the GOUT - I don't think anyone is honestly confused about that.
I also see it as quite a reasonable step up in quality from the GOUT. I wasn't sure it would be when the GOUT was announced, I really hoped what they released would absolutely trounce what we were working on, but now that is out I know it isn't the case.

I've got no problem with criticism, I welcome it wholeheartedly. Many members here have pointed out things we missed, offered up ideas we hadn't thought of, and questioned in a constructive way some of the decisions we have made. This in turn has resulted in quite a few changes in direction for us. If not for the criticism in this forum, the X0 project would have been a straight transfer from the definitive set with some cleanup done. It now has turned into a full blown restoration effort pulling in as many sources as possible to bypass the shortcomings of the Definitive LDs.
In short, criticism is welcome, poking fun is welcome, but stuff that doesn't add anything to the discussion, or is totally made up just messes up the threads.

Originally posted by: boris
I also think sometimes, you guys are misleading people about the X0 project's quality. I mean, wasn't it you who said the OUT '77 crawl was recreated using CG? AND mverta? I know you guys didn't start the controversy, but still...


I don't think I have ever tried to mislead, I try to be as honest and thorough as possible, but I certainly don't know all the answers and can and do get things wrong just like anyone else.
At a first look the 'original crawl' looked suspiciously CG to me, the text looked far too clean and the gate weave looked far too pronounced. I didn't know for sure, but it really didn't look right, that is why I asked a question rather than making a statement. I believe I said something like 'is it just me or does the original crawl look CG' (I can't find the exact post). I was looking for other opinions and trying to find out if anyone knew the answer for sure.

If anything I keep trying to play down the quality expectations of people, I have plenty of posts here telling people to keep their expectations reasonable . I'll say it again, it sure aint gonna look like a HDTV presentation, most of our sources are laserdiscs, but I now know it will be noticably better than the GOUT as some of our sources are better quality than it at the outset, and with cleanup and correction get better still.


Originally posted by: boris
Zion says it's like a hobby for him, so fair enough. But if you, Laserman, are investing thousands of hours and blood sweat and tears to produce something that may at the very very best - be marginally better then the OUT - then I think you really have lost grip on reality. To spend so much time, and so much effort, for so little to me seems futile.

I have a lot of respect for you Laserman, and for Zion, and for mverta. You are all fantastic people, so please don't act as if I'm here throwing insults.. it is not my intention.


Well seeing no insults are being thrown, I guess I'll take having lost my grip on reality as a compliment then

Seriously though I've got no problem with your opinion or anyone elses that I am wasting my time, but I don't see that as particularly constructive to the threads, or adds to the discussion in a valuable way - but hey you are free to post it. I'd prefer it if people didn't just make stuff up, or argue that black is white, but that is the nature of public forums.

To spend so much time for so little?
Star Wars quite literally changed the path of my life. It set me on a career path that brought me a lot of joy and success. Had that movie not existed I don't know what profession I would have ended up in, but I'm pretty sure it wouldn't have been as fulfilling as my time spent working in film. It created entire communities, and much to my surprise when I was literally in my darkest hour, this community right here reached out to me and helped me back up again. The influence of this movie on my life and others extends far past the film itself.

It is one of the most important films of all time, it changed film making completely (for better or worse) and enthralled and entertained a good part of the world.

For a film of such importance, there is no good quality copy of the film as it was, or even close to the way it was released outside of a film institute, and even then are only achived on film which degrades over time (I do not know if they have made digital archives of the prints they received or not - it is possible they have, but I do not know). I cannot think of another imprtant film, or artwork where this is the case, it really does boggle the mind.

I feel this is wrong, and that such a groundbreaking and historically important film deserves to be preserved, the laserdiscs won't last forever, by bringing what is publicly available into the digital realm, combining them to be the best and closest to original version possible, then at least there will be one version that will not degrade over time.

It is my way of giving back to something that gave a lot to me.

If I knew how much work it was going to be at the start I never would have started in all likelyhood. I've thought of giving it up many a time. If I thought the result would only be marginally better than the GOUT I would probably just keep the X0 transfer itself and give it up now.

I think the end result however will be worth the effort and considerably better than the GOUT, the removal of the ghosts and trails alone would make it worth it to me to continue.





Post
#248485
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
Originally posted by: THX
Well, that might give him the edge.

No comment on the luma/chroma thing?


We looked at doing a lot more than just that.
The SC is very detailed but has quite a bit of video noise. Most of the noise however is in the chroma channel.
Using Shake you can track the two different video captures and warp the chroma from one to match the luma from the other.
This allows you to take a relatively clean chroma from the source with good chroma and layer it into the relatively clean and more detailed luma from the other capture.
But you don't have to stop there, once you have a way to align multiple captures you can start to do other things, like take advantage of the different exposure levels and the different aspects (different noise, dropouts, dirt and scratches etc.) of separate transfers and combine them in a way to try and get a final product that is more than the sum of its parts so to speak.

There are also some other interesting discoveries we made re the particular noise signature of some of the laserdiscs.
Post
#248484
Topic
.: The XØ Project - Laserdisc on Steroids :. (SEE FIRST POST FOR UPDATES) (* unfinished project *)
Time
On the other hand, I am very happy with the DVD's I purchased. What interesting about people like you vbangle, is that you can't seem to understand that people like me regard the X0 project as a great idea, but with little actual interest in it. The lifesaver's flags at the beach are a great idea, but I don't use them. And no, I don't get tempted to go and swim between them.


As someone who spent 5 years risking my own life pulling people out of the water that decided to swim outside the flags and got swept out in a rip, all I can say is people that do that drive us crazy.

I'm glad you enjoy the official GOUT release, a lot of people will be happy with it. The NTSC version does look OK on a reasonable sized TV if you don't mind the problems with it.
A lot of people don't mind PAL speedup, others can't see NTSC judder, some don't notice the DVNR and the trails and smearing it causes.
If you are in that group then more power too you, enjoy it.
If however you are one of the unlucky ones that can't *not* see the DVNR smearing, the four eyed stormtrooper, or the over the top grain then I enjoy the fact that there are fans out there who are trying to do something better, and some that even share the 'how to' aspect of it. Zion is right it is kind of a curse once you notice these things, it is hard to un-notice them and enjoy the presentation.

I'm glad you like the idea of the X0 project, and it doesn't bother me in the slightest if you have no real interest in it - but I'd also add that if anyone has no real interest in a preservation or fan edit, then they probably wouldn't be all that interested in reading or posting a whole lot in the threads about them.
The threads are designed so that people with an interest in a particular topic can get info and discuss that topic.
I've got no problem with people asking questions in threads, even seemingly 'ignorant' questions (who was it said that we are all ignorant, just about different topics), but I figure if someone doesn't have a real interest in a thread or isn't really adding anything that hasn't been covered already, that they wouldn't frequent that thread much.

I try to answer questions on the basis that the person really wants to know the answer, or as much of an answer as I can give, even if I sometimes suspect they might be stirring the pot, I try to answer as honestly and thoroughly as I can.

I don't visit many of the threads here as my interest is pretty narrow, and for a lot of the threads I wouldn't have anything meaningful to add. I probably only read 4 or 5 regularly at the most. These threads get pretty long so trying to stay at least vaguely on topic, and not arguing just for the sake of it keeps everything a lot more enjoyable.

I'm sure the X0 guys will have something to show soon so you can all get a look at a scene or two and hopefully offer up constructive comment.

Post
#247589
Topic
Star Wars prequel film noir (* unfinished project *)
Time
You may want to grab some of the 1080i versions floating around before you start.
Once you turn it back to into 1080P you will have a lot more scope for creative work if your final output is going to be 16x9 at 720x480
For example if your starting images are HD then you can zoom in on a characters face by a factor of at least two and still not get noticable pixelation at SD res, it will allow you to crop shots etc. without degradation in your final piece.

I'd go Sin City on it and go mostly B&W with some colour or tints for particular scenes, a bit of red never hurts when travelling down the dark side.
Post
#247361
Topic
Info: Anyone Got A Spare 1300 Bucks? Star Wars 16mm on ebay...
Time
Originally posted by: ronlaw
Karyudo: some of the equipment on that site is 700 line broadcast cameras. Thats a lot better than SD, no?
Yes, I was referring to the Sniper/SniperPro, at $3000. It's hardly prohibitive. Nearly anybody could raise that kind of cash if they were enthusiastic enough.
I'm sure it would be difficult and relatively expensive, but why all the negativity? I think we are being pretty realistic here. It just needs someone (or a group of someones) to organise it.

Only the legal part is really scary, and to be honest I don't think that Lucasfilm will go after an anamorphic SD scan if they don't bother with all the transfers already done.


Well the Sniper Pro is $5995 and is still SD, you would need to add another $2-4K for a HDV camera and about another $500 or so for a suitable lens to have a hope of getting something better than what we already have. I would think you would be lucky to get away under $9000 (assuming you already had a print) and then a *lot* of post production work. As there is no way you could legally sell or distribute the final product it is a lot of money to pony up.

Even then the snipers don't have infrared capability so you are going to have a very scratchy dirty looking transfer that will need a lot of cleanup even with a really good print.

So kind of like Karyudo said, possible but difficult and expensive with no guarantee of a great result.