Robert Harris on The Bits - 8/3/04 column - OFFICIAL THREAD (1 Viewer)

[Eric Stewart]

Josh Steinberg ... George Lucas in his interview in Sound & Vision magazine seems to convey the general impression that he thinks digital projection is the wave of the future, because it has none of the (ahem) "negatives" of using film prints. If it does not "look remotely filmlike," Mr. Lucas is apparently unaware of it. So, how odd that your criticism concerns one of his films, THX 1138!



Darren Gross ... I wonder if the explanation is not so much spin doctoring as the "true believer" syndrome. It seems Messrs. Lucas and Lowry are true believers in the "power of digital" ... so it almost doesn't matter what film elements you start with, or what their condition is. Whatever the problems and shortcomings are, they (supposedly) can be remedied in the digital domain.



I realize I'm sounding (again, ahem) quite negative about digital projection and digital "restoration." I don't mean to be. I agree with Mr. Harris that, done properly, "digital" is a good thing.



Cheers,

 

[Robert Harris]

Without digital technology, we would not have been able to save Story of a Patriot.



RAH

 

[Ken Horowitz]

quote:I understand and can believe the original negative was shot especially inclusive of the souring CRI opticals, but weren't b & w separation masters made back in 1977?

I can't answer specifically about the film in question, but keep in mind that there are (at least) two major steps to viable separations: (1) create the three separations; and (2) test them by creating a usable element from them.



There are many films on which money was spent to create separations; but money was not spent to test them. Decades later, it was often discovered that the separations were not usable.



(Digital technologies today might be able to help in some cases.)

 

[Josh Steinberg]

Darren - yes, I did sit more than one screen height away. In fact, it looked so bad that I moved even further back, and it still didn't look any better. I think I mentioned it before, but if I didn't, last fall I saw "Alien" being projected digitally on the exact same screen at that theater, and as far as I can remember, I was sitting around the same place. Alien looked fine. THX 1138 did not.



I wish there was a way to get an official explaination for this, as I'm now becoming increasingly curious why a film being released by one of the digital format's biggest supporters looked (in my view) completely unacceptable. There weren't many other people in the theater with me, but I did overhear a couple people on their way out talking about how it looked like they were watching "TV with binoculars".

I do believe that digital technology has its place, whether it be as part of restoring films for actual film projection, or for offering the best possible home theater entertainment.

Has anything else that LDI worked on been released to movie screens after the work was done, either digitally or on film? As far as I am aware, their work has been exclusively in the realm of home theater. If this is their first attempt to present one of their projects on a big screen, in my opinion it's a miserable failure.

 

[Patrick McCart]

LDI's 2K Ultra-Resolution digital restorations of Singin' in the Rain and The Adventures of Robin Hood have been screened in digital theaters. For Paramount, they did the Sunset Blvd. and Roman Holiday 2K restorations. THX-1138 is normal 4K.

(I looked at The Adventures of Robin Hood a little closer and noticed it doesn't have the same type of LDI look.)

 

[Josh Steinberg]

quote:THX-1138 is normal 4K.

I don't want to argue with that, as for all I know, you could be right, but what I saw onscreen did not look like something that could be called 4K.



I suppose it comes down to one of two things, at least that I can think of:

-The theater was provided with low quality digital files to screen, or

-The theater's technical qualities were lacking.



Since I've seen other digitally projected films in that very same theater, I tend to doubt that the fault lies in that equipment, though of course I can't rule that out.



If any fellow HTF members were sitting with me during that screening, I'm sure I wouldn't have been the only one cringing.

 

[Michel_Hafner]

#If any fellow HTF members were sitting with me during that screening, I'm sure I wouldn't have been the only one cringing.



Makes you feel like the hero in "Invasion of the Body Snatchers" when you see something completely unacceptable to you but nobody else seems to see it, even say you are delusional.

When I saw the 35mm print of "Seabiscuit" I was disgusted at how it looked (noise reduction artifact city), but it bothered no one else. All hunkey dorey, even for professionals. I don't know to this day if there are better prints around. I'll have to check the HD/DVD some day but I won't buy it for that purpose. You have to pay me to watch this again, I won't pay to see it if there's a chance it looks like the 35mm print I saw.

 

[Michel_Hafner]

#LDI's 2K Ultra-Resolution digital restorations of Singin' in the Rain and The Adventures of Robin Hood have been screened in digital theaters. For Paramount, they did the Sunset Blvd. and Roman Holiday 2K restorations. THX-1138 is normal 4K.



I know the DVD of Holiday. If it's representative of how the 2K looks this restoration can not be considered final. There are severe artifacts in at least one scene. Some shots need to be reprocessed. According to Lowry all shots are quality controlled so they must have noticed. Either they had no time to fix it or the software was unable to deal with this scene properly at that time.

 

[Robert Harris]

I don't believe that LDI was involved with the "ultra" projects for WB. As I recall they were performed internally, based upon software conceived by AOL.



RAH

 

[Ed St. Clair]

quote:THX-1138 is normal 4K.

Oh, goodness gracious!

King Kong, will be done in 4K (correct?).:-(

 

[Patrick McCart]

quote:King Kong, will be done in 4K (correct?).:-(

It's been hinted that King Kong will be an LDI project, but it was never said what resolution it will be restored at. Although, I wouldn't be surprised if they went for 4K since LDI is working in that resolution now. This could be part of the reason why Warner wanted to wait on releasing Kong on DVD.

 

[ChristopherDAC]

Again on the subject of bad DVD presentations...

One of the numerous firebrands on this forum [not a pejorative term] is very insistent about packaging movies in a 1080p24 format [I disagree with him, not because I think he's wrong, but because I think he's missing a larger point about technology and format definitions]. In any case, he is very insistent against interlacing on two points: 1st he claims that the apparent resolution of interlaced video is reduced relative to the number of lines available as against progressive scan; and second that, he says, vertical low-pass-filtering is commonly applied to transfers for the purpose of reducing distracting scan-line "twitter" and other artefacts.

I'm not sure I believe that the so-called Kell Factor has much applicability, considering one source I've seen thought that the apparent degradation was greater for a photograph than for interlaced video, and anyway that a phenomenon noted by researchers at RCA Labs in the 50s may not be directly transferable to today's programme material [particularly film-source] and equipment. Be that as it may, if this phenomenon does occur and does tend to preferentially reduce the visual perception of vertical picture detail on interlaced displays, it may explain part of the reason why vertical picture noise [top & bottom haloes. &c.] is allowed to get out the door, when [unlike horizontal detail] it is not mimicked or masked by defects of the monitor. High frequency ringing, limited horizontal resolution, and the like may explain why technicians don't see horizontal problems, but vertical ones puzzle me.

If he is correct about that second part, all I can say is that it is a truly bad idea. Almost the only time I have ever been bothered by interlace artefacts [as opposed to NTSC cross-colour, which at least with my LDs changes direction according to the line-scan sequence] it has been due to either (a) an excessively soft transfer, or (b) clumsily applied vertical "anti-aliasing", either of which [noticably during e.g. credit crawls] cause shapes and edges to distort in a maddening manner as they drift vertically across the screen. Since this is the very impression vertical softening is proposed to ameliorate, its counterproductiveness should be manifest. So STOP IT ALREADY, anybody who is listening, OK?

 

[Michel_Hafner]

#One of the numerous firebrands on this forum [not a pejorative term] is very insistent about packaging movies in a 1080p24 format [I disagree with him, not because I think he's wrong, but because I think he's missing a larger point about technology and format definitions]. In any case, he is very insistent against interlacing on two points: 1st he claims that the apparent resolution of interlaced video is reduced relative to the number of lines available as against progressive scan; and second that, he says, vertical low-pass-filtering is commonly applied to transfers for the purpose of reducing distracting scan-line "twitter" and other artefacts.



He's right about the second. Don't know about the first.

And he's right about asking for progressive sources without

vertical filtering. Digital projectors all need a progressive

picture and for CRTs one should leave it to the player to output interlaced if desired. It should not be built into the source in compromising ways (filtering, cadence errors).

 

[ChristopherDAC]

I was just watching a 1986 demonstration LaserDisc, all video-source material which I guess must have been recorded on 1-inch helical tape originally, and I would match it for overall picture quality with any 4:3 DVD. The mastering job was just excellent, with a sharp picture and vibrant colours, and the excellent motion rendering typical of 30fps interlaced sources. Obviously, it had never been subjected to any filtering except the single low-pass filter necessary to fit it into the LD passband. Now, if film-source material is subjected to vertical filtering before encoding as interlaced video, I have to submit that it shows a lack of confidence in the mastering job, rather than a concern for any potential "interlace artefacts"; because I have seen a great deal of film-source material far softer than this video-source material, which exhibited what I may term a "total lack of artefacting". It seems that such softening would rather hide problems in the film or transfer, than "protect" the viewer. Any such filtering, then, I must consider a serious error of judgement on the part of the producers, unless very weighty reasons can be produced against me.

 

[Eric Stewart]

Christopher,





I too have read that the vertical resolution of interlaced 1080i/30 video is filtered by some 30% or more of potential vertical resolution in going from progressive scan to interlaced. This is a point that has been made many times by Joe Kane in articles written for Widescreen Review magazine.





Cheers,