Metropolis (Fritz Lang, 1927)

[Tommaso]

I really don't see why having a point of view, substantiating it, and then insisting the criticism is relevant and incisive causes such indignation.

I guess it's not so much your criticism itself but the way you have put it forward.

The discussion about the correct speed of "Metropolis" is a pretty old one around here and probably a lot of other places and regularly surfaces time and again. I have agreed with you initially at least so far that I believe 24fps is wrong, especially as I've seen the reconstructed film theatrically at a slower speed with live accompaniment of the Huppertz score with full orchestra at a time BEFORE Transit released their original disc. It was performed even by the same orchestra and I believe the same conductor who did the recording for the current disc; I'm pretty sure it wasn't 16fps though, but something around 20fps, as the performance lasted around 140 or 145 min and had natural-looking movement. Hence my argument that FWMS themselves were not sure about the correct speed, at least at the beginning. Where does that lead us to? You believe that anything above 18fps is wrong, I believe that 20fps would be just right, and both of us agree that 24fps (or, godforbid, 28fps) is wrong. All this is basically a matter of 'how does it feel for you', and you're entitled to have your opinion as much as anyone else.
We won't come to a conclusion, anyway, as Koerber's points are certainly valid from his perspective, although I believe that he places too much emphasis on that premiere screening about which we don't know much (at least we don't know the reasons why it was projected that fast and whether Lang agreed with this or not). My belief that 20 fps is correct can only be backed up by comparing the flow of the images to the other films by Lang which are invariably transferred at 20fps to disc, and all look perfect to me. So if Lang made all his other silents at 20fps, even those two he made AFTER "Metropolis", I'd be surprised if he chose another speed for just this one film and then reverted back to a slower speed for his next two films.

Criticizing the MoC transfer, if there are any problems with it, is allright of course. It has been done with other films by MoC as well around here, with no bad feelings raised. But what is needed is substantiation: either post screen caps of those moments you think are problematic, or point to them by giving the exact time on the disc so everyone can check and compare for themselves. But please no ad hominem arguments and sweeping generalizations.

[MichaelB]

I actually wrote a detailed reply to your last post before having second thoughts and deleting it. But this latest one is so symptomatic of what I was criticising that I've changed my mind.

Looks like I had better add being quick witted to my list of sins then...

General rule of thumb: if you're posting in a forum you're not especially familiar with, self-deprecation ALWAYS goes down better than arrogance. I am entitled to say that Domino Harvey and HerrSchreck are quick-witted, because that's an independent opinion that's strongly evidence-backed. On the other hand, if you make the same claim about yourself, you just come across as a vainglorious peacock.

You guys do realise that it is precisely because there are so few people that have the guts to say something that is remotely contentious that so much of the cinema and television that makes it to the screen these days is completely and utterly without merit, don't you?

This is such a classic newbie response that I'd almost believe it was a calculated parody if it wasn't for the poker face. And it's just as arrogant as everything else you've been writing, because of the clear implication that we're all sheep and that only the great Robert Wheeler is prepared to nail his theses to the church door.

Actually, the reason I hang around here is precisely because of the number of independent viewpoints, accompanied by some truly jaw-dropping specialist knowledge. And I suspect I'm far from alone.

I really don't see why having a point of view, substantiating it, and then insisting the criticism is relevant and incisive causes such indignation.

I think it's fair to say that the precise point when I felt majority opinion turning against you was when you claimed "I'll add it is not personal preference that leads me to these conclusions, it is no small degree of skill, intelligence and experience."

Maybe this is true - but it needs supporting evidence and, ideally, to be expressed by someone else. In fact, you haven't substantiated many of your more elevated claims about yourself despite much prompting. Which is why I'm assuming your "work in video and film" involves tearing tickets and processing fines for late returns.

The fellow who initially responded to my first comment was very polite. Since then the responses have generally amounted to "if you think the film runs that slow you are obviously an idiot" and proceeded to "just who are you to question the deities that prepared the hallowed masters?".

No-one has said anything of the sort - we're just treating your posts with healthy scepticism, especially given the blithe assumption that you're the only one here who "has the advantage" of actually seeing Metropolis at 18fps. (I'm quoting your own phraseology back at you to underscore how supercilious it is - most people would have the courtesy to ask before jumping to conclusions)

If Lang himself somehow turned up here and started posting under an assumed name, I would fully expect some of you guys to rip him to shreds.

I would certainly rip him to shreds if he started posting about his own work under an assumed name, if only for the blatant deception.

Really, it is strange how politics pervade all areas of life. You don't like somebody's point of view, you question the persons character and try to discredit them as a loon.

I prefer to think of it as an equivalent of Newton's Third Law of Motion. If you make a sweepingly complimentary statement about yourself, someone will counter by asserting the opposite. The more elevated the praise, the harsher the dissing. It's all about reaching a point of equilibrium.

[HerrSchreck]

..and there's one thing that screams out for clarification:

Having the epiphany "clearly the motion of the actors when the projection is cranked to 24fps does not correspond to the speed of the camera that captured their action on the set" is not a basis in this case to cry "foul" about the dvd releases and their fps.

Koerber, Patalas, FWMS et al mutatis mutandis (as well as Lang & Hupperz and anyone and everyone else involved with the exhibition of the film at 24fps at the time of it's release) all are and were all well aware of the fact that the motion of the actors was slightly fast when the film was projected at this rate. Noticing this phenomena now, here-- today-- doesn't constitute "a door of truth cracking open" so to speak, whereby something crucial that nobody else has picked up on suddenly becomes evident, requiring a yelling from the rooftops about what is "clearly too fast to be 'right'".

Everybody knows 24fps is a speed up of motion generally (discounting the issue of variable cranking for expressive effects) vs the average crank speed of the film. This isn't the issue that turns the tumblers and unlocks the issue, however. Take a film like The Passion of Joan of Arc by Dreyer. Slowed down to the 16-18 fps the film is interminably slow, the caricaturing of the absurdity of the interrogators loses its edge (for me at least)... and when those who'd been nourishing themselves on bootleg copies of the Oslo print at last saw the film when finally corrected and transferred at the correct projection speed corresponding to it's contemporary projection, the film, as cinema, went from "watching paint dry" to "blazing masterpiece of film art that may be the greatest film ever made-- ever".

Holding up "naturalism in movement" as your Problem-Solved! piece of evidence-- ie you are determining the solution to the historical question simply by finding "natural feel" via speedup & slowdown of fps-- reveals, rather than an expertise, a naivete. It's in fact one of the least reliable indicators for determining the speed that a film was projected, particularly one released in 1927. Koerber Patalas & FWMS and everyone on this board have long been aware that at 24fps the action plays out faster than the action which was being recorded on film.

And the act of associating yourself with Lang himself via

If Lang himself somehow turned up here and started posting under an assumed name, I would fully expect some of you guys to rip him to shreds.

puts you firmly in the Richard Cranium running. Sure, why stop at Lang in associative greatness?-- why not "and verily I say unto you, No prophet is accepted in his own country'"?

[Robert Wheeler]

Tommaso: The issue I speak of is present throughout the disc, and I have revised my opinion of what is happening with it a little bit. I think the problem is caused by how they handle grain filtering and exacerbated by the route the image has made to becoming encoded. There is a lot more grain on the MoC disc than on the original 2003 Eureka disc. I think what may have happened was that the 2003 master was a direct reduction from the 2k master. Going from 2k to PAL DVD resolutions, most remaining grain would have been filtered out as most pieces would equate to less than half or even a quarter of a DVD resolution pixel.

However, I think the MoC DVD masters were reduced from 1920x1080 domestic HD resolution masters. One of the problems of the HD formats has been that because a piece of grain is suddenly the size of a whole pixel, if it is not carefully filtered a lot more can get through than on older formats, and it can seem relatively a lot more grainy. If they used a domestic resolution HD master as the source, then in turn more grain would make it through the reduction and onto the DVD master.

As soon as you have more grain on the master, the resulting veil makes the image looks relatively darker, and the apparent level of contrast drops. The temptation is to give the contrast and top end a slight boost. With the contrast boosted you suddenly have more of an issue, as the relatively dark and random grain is now dispersed over the brighter image it sits on top of. The result is when the mpeg encoder kicks in, it has a much more difficult image to encode. Mpeg does not like very random dark spots over bright images, and I think this is part of the reason the grain on the MoC disc shows a degree of shadow which is not present on the 2003 disc. The effect on the overall image is quite a noticeable drop in image quality as so much of the bandwidth becomes devoted to trying to reproduce the tricky artifacts.

MichaelB: I'm sorry, your post is all about being personal and nothing about the topic, and really it makes no odds to me whatsoever. I think you should have wrote your detailed post instead of what you did.

[MichaelB]

Brownlow's quote strikes a chord for every film fan and cinephile who ever sat through this ludicrous "politically correct" slow speed projection regime of everything from Lang to Wellman in the late sixties and seventies. The whole experience of silent film viewing became a form of torture and must have turned a generation of people off ever watching silents again.

Tell me about it - I recently suffered through a viewing of Cecil Hepworth's Comin' Thro The Rye (the 1923 version), which had been transferred at 16fps (I can be absolutely certain about this because technical notes accompanied the Digibeta), and it was like a George Romero zombie film without the gore or social satire.

I suspect 20fps would have made a world of difference, and I would think long and hard before screening (let alone transferring) a 1920s film any slower than that.

(It didn't help that I viewed it in complete silence, which merely exacerbated the problem.)

Take a film like The Passion of Joan of Arc by Dreyer. Slowed down to the 16-18 fps the film is interminably slow, the caricaturing of the absurdity of the interrogators loses its edge (for me at least)... and when those who'd been nourishing themselves on bootleg copies of the Oslo print at last saw the film when finally corrected and transferred at the correct projection speed corresponding to it's contemporary projection, the film, as cinema, went from "watching paint dry" to "blazing masterpiece of film art that may be the greatest film ever made-- ever".

I strongly suspect this was the difference between my first and second viewing (the second was the Criterion DVD), because that was absolutely my reaction.

[Felix]

No way do I want caught up in the crossfire here, but wrt the Eureka vs MoC releases, and in the interests of fairness, it was suggested over on the Der Latze Mann thread that the Eureka of that title is better than the MoC and though I don't set too much store by screen caps, the ones I saw were enough to stop me double dipping up from the Eureka.

[Jason Gorman]

I wish to address the FPS issue and to a lesser extent the music.

Evidence: The clock just before shift change advances one second after movement ceases in exactly 16 frames so the film was shot in part at least at 16 frames per second. I have watched the film at that rate and can tell you it is accurate. My expertise is extensive in the area of video (not film) so I have spent much time considering frame rates and movement and have experimented with the film digitally.

However, it is quite clear when viewing the film at this rate, that the actors were intentionally slowing their movements at times (possibly in anticipation of the the common practice of projecting films at a faster than filmed rate).

This is most noticeable in pauses and not during faster movement. The pauses are too long during times the actors are not moving (except for Abel who seemed to be able to slow almost all of his movements). Thankfully, since these are stills, this could be easily edited.

Fortunately the technology exists to change the movement dynamically and also interpolate the frames to the common film rate of 24 fps (twixtor among others). This results in the best of all possible worlds; smooth motion, not too fast and not unbearably slow. I have been fortunate to have experienced this in my experiments.

Regarding the music, the score does not sound appropriate when slowed down digitally (I have tried this). I have heard it at the same tempo (bars duplicated) and it sounds great this way: remixed (I also have expertise with audio). Of course a rerecording would be better, but not necessary.

What is most important, is judging from the profound difficulty in producing some of the film's visual effects (research the mirror trick or multiple exposures), my opinion is that there is nearly nothing in this film that was done by mistake.

The question is: are you a purist about the film? I for one am not. I would like to see it as the vision it was intended to be; however that is accomplished. This leads me to another can-o-worms: the intertitles have got to go.... maybe some day they will find a wax recording of the actors words, they sure said enough of them in the film.

[MichaelB]

This leads me to another can-o-worms: the intertitles have got to go.... maybe some day they will find a wax recording of the actors words, they sure said enough of them in the film.

???

[Saturnome]

I hope more for the 3-D print, just like Fritz Lang saw the film when he was actually filming it.

[Tommaso]

I have the strong feeling that it was Lang's intention to have the film in 2.35 scope format, given the monumental character of "Metropolis". This part of restoring Lang's vision would be easier achieved than the rediscovery of the lost words.

[domino harvey]

Really, none of us will ever be able to "experience" Metropolis as it was first shown unless we are there when it is first shown. I refuse to view the film until science perfects time travel, and anyone who dares to watch the movie without first going back in time doesn't really care about cinema

[MichaelB]

I have the strong feeling that it was Lang's intention to have the film in 2.35 scope format, given the monumental character of "Metropolis". This part of restoring Lang's vision would be easier achieved than the rediscovery of the lost words.

You're right - Lang was an outspoken champion of widescreen aspect ratios, and it's a tragedy that his Scope masterpiece Funerals and Snakes is now believed to be a lost film.

[Tommaso]

You're right - Lang was an outspoken champion of widescreen aspect ratios, and it's a tragedy that his Scope masterpiece Funerals and Snakes is now believed to be a lost film.

No, he remade it as "Die Nibelungen", but stupidly the producers insisted that the snake wasn't exciting enough and had him replace it with a dragon, thus spoiling all his original intentions. The dragon would also have looked much better in scope, of course. But at least Lang got the funeral parts right.

[HerrSchreck]

No, he remade it as "Die Nibelungen", but stupidly the producers insisted that the snake wasn't exciting enough .

That's also not accurate-- Lang liked unexciting snakes, and went out of his way to insert one in, for example, the color Indian Tomb saga where Lil Dagover-- that's Lil Dagover-- did the humpty hump in fronta the statue. Notwithstanding the stuffed tiger.

As my experitise resides in the zone of stuffed animals (being from the Bronx streets of 70's/80's , we've stuffed a lotta animals inta the back of cars and cement filled drums) I can authoritatively say this without eye twitch..

[Tommaso]

Lang liked unexciting snakes, and went out of his way to insert one in, for example, the color Indian Tomb saga where Lil Dagover-- that's Lil Dagover-- did the humpty hump in fronta the statue.

Who's the snake? Lovely Lil? Unexciting?!
But I agree that's a scene that cries out for widescreen. Vertically, I mean. Though it would have to be called tallscreen, then.

[Jason Gorman]

Here are a few 3D images from Metropolis for those of you who expressed an interest.

[Saturnome]

Great!! Now all I need is the 4 strip technicolor print, or even 5 why not. The more the better!

[Jason Gorman]

Great thing about the film as transferred at 25 fps: we have access to all of the individual frames on the PAL version with having to reverse pulldown. This way, anyone can extract the frames and view it the way they wish. This is something that is impossible to accomplish on the US version though. In fact, there are problems with that transfer making frame blending permanent and impossible to remove.

[HerrSchreck]

This thread/film sure attacts some doozies.

[marknyc5]

I just found this thread and I have to say I completely agree with Robert Wheeler's take on the speed, based on my own experience with the film.

A couple of years ago, I decided to make a camera-speed version of Metropolis, since I love the film and just can't stand watching people move like Keystone Cops.

So I sat down with my editing software and began choosing frame rates based on my eye - speeds that produced natural movement. Time and again, the rate I ended up choosing was 16fps.

I was not trying to make an historically accurate version - just one that I would want to watch. So I replaced the intertitles with subtitles (heresy, I know), and slowed down the undercranked scenes until they produced realistic movement. I wanted to see what Maria's chase in the catacombs looked like if she ran at the speed at which she was filmed. That speed ended up being 12fps, and for the first time, the chase looked frightening, not silly, to me.

So I am absolutely certain the film ran through the camera at 16fps and was undercranked to 12fps for all the chase scenes.

We'll never know for certain at what speed it was originally projected, but I can tell you one thing: everyone I've shown it to agrees it is much more powerful at camera speed. Not only does Freder not run like an idiot, but all the reactions and facial expressions are more believable.

Surprisingly, replacing the intertitles (except for the animated ones), led to a running time about the same as the 24fps version.

I've asked Martin Koerber if there's any chance we could see the restored version released at a slower speed, and he says no, that I'll just have to revise my camera-speed version when the new footage is released.

Oh, well, back to the editing room! =)

Mark

[Rufus T. Firefly]

According to this a 9.5mm print of Metropolis has been found in Chile. They think it may be another copy of the complete 1927 version and are sending it to Germany for verification. Actually, it sounds like they don't know what it is as they can't project it.

[HerrSchreck]

Famine to a fucking deluge in a single quarter.

[marknyc5]

Babelfish translation:

Original Metropolis Found

A presumably original edition of the film Metropolis, directed by the German Fritz Lang in 1926 and one of the symbols of the expressionist cinema, was discovered in the Film library of the University of Chile.

The film, in the 9.5mm format, remained for decades in the warehouses of the Film library of the University of Chile and in 2006 was found to be an edition similar to that exhibited in Germany for its commercial opening in 1927.

According to the newspaper La Tercera, Luis Horta, curator of the Film library, said the cans it was stored in said only "Metropolis" on the cover.

“We checked and we saw that it was odd - it was 9.5mm, a format that is obsolete, which has a perforation in the middle, not to the side like 35mm film, which makes it impossible to project,” he said.

The expert explained that the discovery was not made public immediately, (in 2006), because the Film library had other priorities like the recovery of films by the Chileans Raul Ruiz, Helvio Grove and Miguel Littin.

In any case, he announced that next week the film will be sent to Germany, to the Murnau Foundation, owner of the rights of Metropolis, for verification.

According to Horta, only that foundation will be able to determine if the one found in Chile is the 1927 edition, after a verification process that will last between six months and a year.

He stated that since the 1940s, only an edition of 120 minutes released by the American company Paramount circulated, but in 2001 the Murnau Foundation made a restoration on the basis of several copies and managed to add some scenes that increased the duration to 147 minutes.

Nevertheless, he recalled that version, released January 10, 1927, was a reduction done by Lang himself, whose original lasted 210 minutes.

Horta said that the film was forgotten for decades, because after the military coup of 1973 (the time of book burning and destruction of the culture of Chile) the then-director of the Film library, Pedro Chaskel, changed the labels of some films to avoid their being destroyed by the military.

[HerrSchreck]

How in the fuck-- even then-- did they project a film w perfs down the middle.

[marknyc5]

From other forums:

9.5mm features were exceedingly common on the European home market in the 20s and 30s. Because they are still relatively easy to find, there are others copies of METROPOLIS out there. If the 9.5mm print were the original uncut version, we would have known about it long before now.

These 9.5 versions of feature films were similar to the ones sold here on 8mm or 16mm for showing at home and were so short that I do not think that it would have any lost scenes. The reels that I have seen play for about 3 1/2 minutes, similar to the old 8mm reels used for home movies but some may have been longer.

All true; but sometimes original 9.5mm releases went out before a film was, say, re-released; if all that survives of a 35mm film is a re-release edit or TV print, that may mean that the 9.5mm versions carry different scenes or sequences to the archived 35mm prints.

A 1933 Pathe 9.5mm catalog I got from the UK has METROPOLIS listed but shortened to five "super reels", the equilivant to five 1,000 foot reels of 35mm. Roughly 50 minutes at 24fps, depending.

One thing I want to add here, about METROPOLIS, is that the film distributor has to be the same one like it was in Argentina, Cinematográfica Terra, which produced their Spanish version following the original German version. If so (and this is not in the article), the print has to be similar to the one located in Buenos Aires. Film disitribution and exhibition in South America in those days was concentrated in the same hands (and to some extent, it is still the case).

This film was not shown theatrically. 9.5mm was Europe's equivalent of 8mm, and was sold to the home market (not unlike Blackhawk, Castle and Ken Films). Pathe produced and distributed these "home movie versions" all throughout Europe. AFAIK, there was no 9.5mm market in South America (correct me if I'm wrong), so this film must have come over on the boat from France, Germany or England.

The real mystery to me is why the Chileans believe this is an important find. Maybe it's the 9.5mm gauge. But a 9.5mm condensed version of METROPOLIS is the same thing as an 8mm condensed version of METROPOLIS... available to everybody, and neither rare nor complete.