Metropolis (Fritz Lang, 1927)

[miless]

Stalker was an ok transfer, but the colors were muted.

It wasn't their transfer (it's the same transfer used for nearly every DVD release that was created for RUSCICO)

[moviscop]

Stalker was an ok transfer, but the colors were muted.

It wasn't their transfer (it's the same transfer used for nearly every DVD release that was created for RUSCICO)

Kino made the only big Release of stalker for our region (America). The colors were muted in many areas and it didn't look up to its full potential.

If Criterion was able to release a version I can guarantee that they would have taken more ownership than Kino did, and have done a better job.

[aox]

The treatment of Stalker is criminally bad by KINO. Criterion should be all over that. Same with Come and See.

[HerrSchreck]

You guys arent getting the point here: Stalkers transfer from Kino is just a relicense of the RusCiCo transfer from the earlier Image disc. RusCiCo holds the elements and the distribution rights. When they get good material (like the new Paradzhanov's) the disc is good. They're not doing these transfers.

[moviscop]

You guys arent getting the point here: Stalkers transfer from Kino is just a relicense of the RusCiCo transfer from the earlier Image disc. RusCiCo holds the elements and the distribution rights. When they get good material (like the new Paradzhanov's) the disc is good. They're not doing these transfers.

well someone is gonna get blamed! i blame it on the russians! who's with me?

[justeleblanc]

Kino releases what's available and tends to be more liberal when it comes to releasing rare films without prestine prints available. Currently Criterion is sitting on the rights to Antonioni's La Notte instead of releasing it since they say the elements are bad. If Kino were to have owned the rights, they would have probably released it. It's a trade-off, but given that so many films remain unavailable, I'd much rather a film be released.

And please feel free to show me the muted colors in Stalker. I saw the 35mm print of that film recently at the DVD looks identical.

[Darth Lavender]

Stalker is a muted looking film, most of the Russian films from that period are (nature of the film-stock)
As with all Ruscico ports, the NTSCs should ALWAYS be avoided becaused they're ultra-cheap, bootleg-quality conversions of the PAL dvds. (And the PAL dvds are usually pretty bad to begin with)

[Tommaso]

Back to the 'Metropolis' transfer. Isn't it the same with this film as with the older films Kino licensed, i.e. they got the master from FWMS/Transit? If so, I can't understand how this should look bad, unless Kino managed to somehow feck it up in the process. The Transit version is one of the best looking silents I ever came across, and whatever still looks weak in it is doubtless due to the materials.

[Darth Lavender]

Back to the 'Metropolis' transfer. Isn't it the same with this film as with the older films Kino licensed, i.e. they got the master from FWMS/Transit? If so, I can't understand how this should look bad, unless Kino managed to somehow feck it up in the process. The Transit version is one of the best looking silents I ever came across, and whatever still looks weak in it is doubtless due to the materials.

According to DVDBeaver's review, Kino did manage produce a dvd noticably worse than the MoC (of the same transfer)

Apparently, Kino used ghosting to compensate for frame-rate (while MoC used interlaced. Haven't seen the Kino in motion, so I can't really comment on that) But worst of all, Kino replaced all the intertitles with english, even creating new shots of gravestones, etc.

[Tommaso]

But worst of all, Kino replaced all the intertitles with english, even creating new shots of gravestones, etc.

Yes, the old problem. I wonder whether they will continue this practice with a "Metropolis" Blu Ray... Didn't know about the problems with the transfer itself, though.

[justeleblanc]

Kino did not create new shots of gravestones.

[markhax]

But worst of all, Kino replaced all the intertitles with english, even creating new shots of gravestones, etc.

Gravestones? In Metropolis? Do you mean the inscription on the sculpture of Hel?

The intertitles are another matter, and one reason I have wound up augmenting my Kino German silents with MOC versions. But even a high-class operation like Flicker Alley used English intertitles for Phantom.

[HerrSchreck]

NTSCs should ALWAYS be avoided becaused they're ultra-cheap, bootleg-quality conversions of the PAL dvds. (And the PAL dvds are usually pretty bad to begin with)

That's quite an outa control statement

[Darth Lavender]

NTSCs should ALWAYS be avoided becaused they're ultra-cheap, bootleg-quality conversions of the PAL dvds. (And the PAL dvds are usually pretty bad to begin with)

That's quite an outa control statement

Yes, the statement you quoted is ridiculous, and whoever said that is clearly insane, ignorant or both.

But, what I said was "As with all Ruscico ports, the NTSCs should ALWAYS be avoided because..."

That's simply Ruscico's business practice, and it's well known. The master in PAL format, and then make a low-quality NTSC conversion for the American markets. I actually make it a policy to avoid Ruscico as much as possible, having been very badly burnt on their NTSC release of "War and Peace" (paying $80 at a time in my life when that was a LOT of money. And, the irony is, if only they had been honest and upfront about the NTSCs being converted from PAL, I would have just bought the PAL version)

[Darth Lavender]

Gravestones? In Metropolis? Do you mean the inscription on the sculpture of Hel?

Yes, that's it. To tell the truth, Metropolis is one of my least watched silent DVDs (House of Usher, Golem and Nibelungen getting probably the most viewing) I just vaguely remembered seeing that screencap on DVDBeaver's comparison, and couldn't remember what it was exactly.

Use of English vs Original Language intertitles is one of those things, like DNR vs Grain (on High Definition) or Dubs vs Subs (in the old VHS days) that audiences and studios are only slowly starting to change their views on.

In the case of 'Phantom' it was one flaw in an otherwise incredible effort (especially coming from such a tiny producer) so I bought it anyway.

[Ledos]

But, what I said was "As with all Ruscico ports, the NTSCs should ALWAYS be avoided because..."

That's simply Ruscico's business practice, and it's well known. The master in PAL format, and then make a low-quality NTSC conversion for the American markets.

Correct. The NTSC versions offered on their site, or released by Image for example, are poor due to the conversion issues. Ruscico aren't that bad, just stick to their PAL editions.

[HerrSchreck]

They have some issues vs a native transfer-- but "bootleg" is a bit of a punt, I'd say. And they've begun recovering with the Paradzhanovs for sure.

[Michael Kerpan]

David Bordwell's blog (reporting from Bologna) discusses the new found print of Metropolis.

[Antoine Doinel]

Great link, and it only underscores what a miraculous discovery this is. The car journey through the city sequence sounds intriguing.

[Robert Wheeler]

I am mystified that after all this research they insist on reproducing the film at 24 frames per second. My careful reasoning leads me to believe that the intended playback speed was 16 fps. The is supported by how much better the score sounds at this speed. Also in the opening shots, the clock sound motif plays at one second intervals when played back at 16 fps. At the very least the shooting speed was 16fps.

I wrote a bit of an essay on this.

[Tommaso]

Very interesting points you make in your essay. 16 fps sounds a little slow though, given it's a 1927 film. Have you tried this procedure with 20 fps? I saw the film projected once with around that speed and it looked perfect.

Your appeal to Koerber can only be supported of course, but I think economics will be in the way. They already recorded a full 'study' version with the complete score (which runs longer than the released dvd, i.e. it has longer passages of black screens to indicate the exact length of material missing) in which they only need to insert the missing footage, which would make the new release pretty easy to do. Slowing the whole thing down to 16 fps or even only 20fps would of course be technically possible, but even with modern time-stretching audio tools would certainly result in a somewhat 'artificial' sounding soundtrack. And I can't believe that they'd do a complete re-recording of the music at the correct speed , especially if they are not sure about the speed (which they evidently are not). Such things cost money...And even if Koerber should believe that he chose a wrong speed then, without given the chance to correct it now, I'd doubt that he would admit it now. And maybe he still believes in 24fps.

[Robert Wheeler]

I looked at this very carefully. I think the fastest it could conceivably be is 18 frames per second, but looking at the numerous scenes with water 16 fps seems more likely to be the general shooting speed. Actually originally when I wrote that essay I said 18fps was the speed, but I realised the stretch had lead to a 16fps rate rather than 18.

Mathematically if the film ran 210 minutes, the closest you can get to that with a conventional frame rate on 13,823 feet of film is 18fps at about 205 minutes. When you stretch the 2001 restoration print to 185 minutes (deducting from 210 minutes the 25 or so minutes of missing footage) with the titles adapted to the slower speed, for some reason the actual playback speed works out at 16fps. I am not quite sure why this is, but this is how it works out. Not very scientific I know.

I could see that Lang may have compromised to allow a target projection speed of 18fps. I think this is where one or two people trip up when judging the correct playback speed. 20fps looks better than 24 and 25fps because it does not run stupidly fast. However, it is only splitting the difference and therefore relatively better than the shooting speed, or the target projection speed. The slower you get, the less illusion of smooth movement is apparent and as a result the film can look like it is playing slower than it actually is. I can see how people might favour 20fps over 16fps, but only if they had spent a lifetime watching the film at 24 and 25 frames per second.

So anyway, I will concede, the film may have been shot at 16fps for projection at 20 fps, but I have serious reservations that it was every intended to be shown at 20fps or above. For me, it was shot with the score being played at 16 or at most 18fps and this is the way the film should be reproduced. They certainly need to re-record the score.

[markhax]

So anyway, I will concede, the film may have been shot at 16fps for projection at 20 fps, but I have serious reservations that it was every intended to be shown at 20fps or above. For me, it was shot with the score being played at 16 or at most 18fps and this is the way the film should be reproduced. They certainly need to re-record the score.

The problem with re-recording the the score at a slower speed is that it would contradict the composer's own expressed intentions. To quote from an earlier post of mine in the 'Nosferatu' thread:

"I ran across two sources that say that Huppertz wrote a recommended projection speed of 25-27 fps onto his score, which suggests to me that Lang approved of this speed. In his book, 'Metropolis in/aus Trümmern', Enno Patalas gives the original running time as 140 minutes. With a length of 4189 meters, that comes out to 98 feet or a little over 26 frames per second, which is consistent with Huppertz's recommended projection speed."

In the critical edition DVD, which plays the entirety of Huppertz's score, even where the footage is missing, doesn't sound too fast to me.

[HerrSchreck]

Me neither. Going by "minutes" to deduce the qty of the film that's missing, and the way the action "feels" and the way the music "sounds" are not appropriate methods. To determine what is missing, the only concrete method of arriving at a tangible result-- whatever that may be-- is to use the number of reels/meters registered with censors upon release of the original version, versus the number of reels/meters of film that remain. This is the only way to determine how much is missing in flat, factual terms. And I believe this is waht Koerber/Patalas et al did when they declared that nearly XYZ of the film is to be considered lost.

[Robert Wheeler]

If Huppertz actually suggested a playback speed of 25-27 frames for his own score, then he was flat out wrong. I can only imagine he had either taken leave of his senses or was committed to tow the company line on that issue.

This is where it is unfortunate that there is no actual film producer there to oversee this project. They need someone to say that after all the academic research and all the assumptions made about the film, and not to mention the wonderful restoration work, common sense must prevail about it's presentation and delivery.

I think the score tops out at 18fps. Anything above that it becomes inane and twittering.