Jump to content




Photo

Eyes Wide Shut and the disappearing film grain


  • Please log in to reply
10 replies to this topic

#1 Hrishikesh Jha

Hrishikesh Jha
  • Basic Members
  • PipPip
  • 44 posts
  • Student
  • Ottawa, Canada

Posted 20 October 2014 - 02:58 PM

Hey guys,

Thanks a lot for those responses to my thread. I'll apply them someday to my project.

I found this article a fascinating read and thought of pasting it here to hear your thoughts on it:

 

 

DISAPPEARING FILM GRAIN 

Many viewers have complained that the “Eyes Wide Shut” DVD cleaned up the rich film-grain that was present in the 1999 theatrical release of “Eyes Wide Shut”. Below is a slightly edited discussion involving this issue. 

MIKE HAPES: Has Warner or the Kubrick Estate released any information as to why the Eyes Wide Shut DVD doesn’t capture the high levels of film-grain that were present when the film was released in theatres back in 1999? 

ZANDOR: When I saw the film in 99 with a group of friends, the grain was all everyone talked about. It was a murky, rich look. Warner, in all their wisdom, must have cleaned it up thinking that’s not what Stanley wanted. 

JAROD OBANNON: Maybe ask David Mullen if he’s still around. This is a question for a cinematographer. Me, I never saw the film in theatres, but the grain does seem to jump a bit in the DVD. 

ROVER: How do you know it’s not what Stanley wanted? Whenever you film in lowlight conditions (with natural lighting) you must use a fast stock (which causes significant film grain). Once you shoot any other denser stock, you will not get a scene-to-scene grain match. Therefore the whole movie had to be shot, or even stepped down to a faster stock for a consistent scene-to-scene grain match. 

JAROD OBANNON: That doesn’t make sense. You’re saying he graded the film to have “uniform grain” and then ordered the DVD’s to remove all grain? 

DAVID MULLEN: Since there have been other “night time” movies filmed at low light levels that are not as grainy as Eyes Wide Shut, one can assume that a certain level of graininess was a deliberate visual choice by Kubrick beyond his basic technical needs for low-light filming. 

JUGGERNAUT: Don't know what film stock was used, but I read somewhere that the bulk of Eyes Wide Shut was shot with minimal lighting, and the film was pushed two stops, hence the grain and the halo around the highlights. 

ALEX DE LARGE: The excessive grain is probably due to bad shoddy printing off a bad IP. The IP for this film was made post Kubrick’s death and he would have DIED seeing it and would never have approved it. Those who try to justify the grain as part of some dramatic grand design are misguided. The DVD is how it should have been released. The grain is not caused by shooting in low light, Kubrick began doing that in Barry Lyndon (candlelight) and that film is impeccable and completely without grain. 

DAVID MULLEN: How do you know this to be true? Simply for no other reason than your belief that Kubrick would not have shot a grainy-looking movie? 

Why would Warner Bros. want to release thousands of bad prints? Do you realize how much work it would take to get an I.P. to be this grainy if the film wasn't shot that way to begin with? First, you'd have to severely over-develop it and then strike multiple dupe negatives so that any other lab making release prints would also get the same results. You'd also have to have a colour timer at a major lab deliberately do this when making the I.P. (why, because he hated Kubrick?) or not even watch the first check print from the dupe negative. And remember, they always watch the print from the dupe to check for overall colour adjustments, comparing it to the answer print off of the original negative. 

It's been reported that Kubrick chose to push the negative for most of the film, underexposing by two stops and overdeveloping the negative to compensate, thereby creating a grainier print which was the look he wanted. It seems clear that the film was supposed to look dirty, somewhat dated, a throwback to the 60s. 

AKULA: I think the grain was deliberate on Kubrick's part. It helps knock down any sense of false beauty and glamour and helps bring out a story's raw edge. We've been conditioned to believe that "good" photography means anything evenly lit, properly exposed and with clean, grain free images. This goes back to the studio days when such things were required, and without question. Any variance of that meant the cinematographer was fired and replaced the next day. The experimental films of the late '50s and early '60s, cinema-verite documentaries, to name a few sources, changed that…but you know that they say…the sixties are dead. 

JUGGERNAUT: I actually think this was one of his most ballsy approaches image wise. He even went so far as to do the black and white stretch printed flashback, which is something I never expected him to do. I also really dig the fact that this was shot with mostly location light and boosted practicals. Using Christmas lights for fill illumination is unbelievable. 

JAROD OBANNON: Let's face it Kubrick, wanted Eyes Wide Shut to look DIRTY. Like ARTY Super 8 footage. 

MP: Well the cinema print seemed much grainer than the film looks on DVD so the image must have been cleaned up. Mind you, I’ve only watched the DVD on a small laptop screen. The theatre could be merely magnifying the look of the grain. 

GUNTHER GLOOP: No, the cinema print was like watching a moving canvas - the whole image was made up of chunks, enhanced by the exceptional lighting in the film. I imagine watching Barry Lyndon in a proper cinema would be likewise. The dvd cleaned it up, going with the theory that clear means better. 

BILL REID: The theatrical release was MUCH grainier than the versions released on VHS and DVD. A couple/three reasons for this: 

1. ALLEGEDLY, Kubrick shot and intended the movie to be seen in "academy aperture", which is essentially the same aspect ratio (4:3) of the old CRT television screens (or the new "standard definition" digital television). For theaters, the movie was center-cropped and enlarged to fit the wider aspect ratio of movie theater screens, scalping Tom Cruise and losing some action occurring in the top and bottom of the frame, and also had the effect of enlarging any grain in the original frame. 

2. Kubrick definitely deliberately shot the movie with more grain than would be needed given modern film stocks and the lighting available for the scenes, in his end-of-life attempt to be "arty". He also deliberately manipulated the amount of grain throughout the movie: the movie is very grainy at the beginning and end, and much less so in the middle of the movie, the orgy scene, and the effect is that the entire film seems to "fade in/out" of that scene through the "veil" of graininess. 

3. Because the VHS and DVD versions were not centre-cropped and enlarged they are naturally much less grainy, plus certain technical aspects of the transfer process greatly reduced the appearance of the grain, both inevitably and by choice of the technicians performing the transfer. 

MP: I see, so it's a question of the film being enlarged for theatres (thereby magnifying the appearance of grain) and not of grain being digitally removed for the DVD. 

BILL REID: Well, no, as I said, it's not JUST the fact that the VHS and DVD frames weren't enlarged, they also appeared to have "cleaned up" the images somewhat to remove the grain that still would have been visible even un-enlarged. Some of this is just part of the transfer process, some of it is "preferential" on the part of the technicians due to their own sensibilities and making the transfer in the most efficient way possible, and for the DVD, some of it is because GRAIN = DATA, and they have to pack the entire 2 1/2 hour movie into 4.7GB of data on a standard DVD, so SOMETHING has to go, and the grain is the "low-hanging fruit", so they get rid of the grain so they can fit the movie on the DVD. 

AKULA: I imagine as part of the digitisation of film, visual imperfections such as scratches and 'bumps' are "cleaned up" by automated software/standardised procedures. Like some of the newly restored B&W transfers, I believe whoever (or whatever) did the Eyes Wide Shut transfer erred on the side of "ultra-clarity = ultra-perfection", which in this case in particular means the very deliberate canvas-like effect was lost. Personally I hate the look of the DVD and much prefer the richer, grainier look of the film in theatres. 

KEVIN: Just googled this on HomeTheater Forum: "Kubrick shot EYES WIDE SHUT using Kodak's then-in-the-process-of-being-discontinued 5298 500 ASA filmstock (it was replaced by Kodak's Vision line of stocks, which Kubrick tested but he didn't like how the Vision 500 stock's colors reacted to push-processing). This fast film was fairly grainy to begin with, and Kubrick had it pushed a full two 
stops in the lab on top of that. The result was a very dreamy, VERY grainy look to the 35mm release prints. Every scene seemed to be covered in a light haze and grain was swimming everywhere, especially in the blacks. In light of how Kubrick had the film exposed and processed, it sounds like the visual look of these HD releases replicate how the film looked in theaters.” 

I've watched the dvd a few times on a 9ft hi-def screen via a hdmi-connected dvd player. There is no grain in it. The image is so clean you'd be forgiven for assuming it's a textbook perfect transfer if you hadn't already seen the movie in a cinema. 

For that reason I can see why people would see the grain as a distraction if it were to be "brought back" in a hi-def transfer (again, I haven't seen the Blu-Ray transfer yet so I'm not saying it has or hasn't), but for me the grain added a depth and richness to the whole experience of the film that is lacking on dvd. I suppose the same can be said for all truly-cinematic films. Maybe it isn't possible to transfer that experience to the home. I know Kubrick thought so - albeit before certain advancements and the rise in big-screen home cinema setups. 

RYAN: Here's the thing though, that people keep forgetting...a DVD is NOWHERE near a "textbook perfect transfer" of ANY movie. For that matter, neither is a VHS, super-VHS, or even Blu-Ray. 

What you are watching is actually missing anywhere from 40-90% of the colors, and 40-80% of the resolution. For DVDs (and Blu-Ray), you are missing far more color "depth" than VHS, and dramatically more than "laser-disc". You are watching a fairly obvious FAKE picture CREATED by a software algorithm that BY DESIGN sacrifices picture quality to reduce the total data size and instantaneous data rate to levels that are able to be handled by the capacity of the storage medium and decoding hardware. 

Don't get me wrong, the MPEG "codec" is pure genius and brilliantly uses every trick in the computer science textbook (and cunning knowledge of human perception) to present a picture that "fools" most casual observers in most situations, but not only are the "numbers" nowhere near the original film content, a side-by-side comparison or just a careful examination of certain types of "problem" scenes reveals a host of digital "artefacts". 

Film is analog chemical, and so is the human nervous system, digital isn't, so if you really want to re-create human experience (especially dreams) you're better off working with film. Regardless, I’m pretty sure the reason the original film was so grainy was due to the negative stock (Kodak 5298) being pushed forward two stops during processing because Kubrick shot the film using the lowest possible light levels, just the in-scene source lighting in most cases, a process that would increase graininess significantly. 

COSMIC GNOME: While 'conventional' digitization of film onto DVD clearly blocks such fine grain resolution, surely grainyness can be 're-created' as a 'special effect' in most digital post-production houses? Or did the DVD eejits, in their delirial eagerness to sanitize everything, think all that grainyness was simply an 'error'? 

[BTW, I love watching 'old' films with loads of blips, scratches, etc, just as listening to a vinyl LP/album with its occasional 'snap, crackle, and pop', far from ruining the listening/viewing, actually enhances it, partly because it draws attention to the ghostly uncannyness of the technological medium itself, rather than pretending it's not there by seamlessly shutting down all evidence of its presence]. 

AKULA: I'd think would be a tad difficult to 're-create' the grain of the original if it were processed and identified as 'noise', i.e. if the texture of the original was not captured in situ or as part of the signal. It would be akin to removing the germ from wheat and reinserting a reduced portion of it when making bread. 

BILL REID: I think Kubrick was playing a game with the graininess, because I got the general impression that it was more grainy at the beginning and end of the movie and less grainy in the middle, in other words the orgy scene. Kubrick was, I suspect, playing a deliberate game with the grain. 

But be aware that grain, like every other moving and transient picture element, requires extra data to be recorded onto the DVD (and since there is so much transience and movement of grain across the entire picture, it requires a LOT of data), and since a DVD has only a finite amount of data that can be recorded onto it, what you would gain with the grain would be lost somewhere else in the picture: fewer colours, more "macro-blocking", etc. 


  • 0




#2 cole t parzenn

cole t parzenn
  • Basic Members
  • PipPipPipPip
  • 287 posts
  • Other

Posted 20 October 2014 - 07:16 PM

https://www.theasc.c...9/sword/pg1.htm


  • 0

#3 Mark Dunn

Mark Dunn
  • Basic Members
  • PipPipPipPip
  • 2182 posts
  • Other
  • London

Posted 21 October 2014 - 03:04 AM

No mention of grain there. Can't say I noticed it in 35mm.

Kubrick would have protected his negative, but of course there's no enlargement in widescreen. The crop is top and bottom so there's no need to enlarge to fill the screen.


  • 0

#4 Freya Black

Freya Black
  • Basic Members
  • PipPipPipPip
  • 4032 posts
  • Other
  • Right on the edge in London

Posted 21 October 2014 - 08:07 AM


ALEX DE LARGE: The excessive grain is probably due to bad shoddy printing off a bad IP. The IP for this film was made post Kubrick’s death and he would have DIED seeing it and would never have approved it.

 

I'm sorry I'm willing to believe that Stanley Kubrick may or may not have liked grain but I'm NOT prepared to believe it was responsible for his death. My understanding was that Mr Kubrick died in his sleep.

 

Also the suggestion he died as a result of the print would appear to involve time travel which I am also skeptical about for reasons I won't get into here.

 

I'm not sure why grain gets such a bad rap.

 

Freya


  • 0

#5 Mark Dunn

Mark Dunn
  • Basic Members
  • PipPipPipPip
  • 2182 posts
  • Other
  • London

Posted 21 October 2014 - 08:57 AM

You've misinterpreted that statement. He uses the past conditional. It's died as in 'died of embarrassment'. It's a figure of speech.


  • 0

#6 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 18789 posts
  • Cinematographer
  • Los Angeles

Posted 21 October 2014 - 10:31 AM

There are a number of issues being mixed up here.

 

1. The 500T negative was pushed two-stops.  So there would definitely be visible graininess no matter what!  And I doubt that Kubrick was thinking along the lines of "I don't like the way that looks but at least on home video I can get rid of that grain."  So let's just assume that some grain level was expected by Kubrick because he tested the hell out of the stocks before production.

 

2.  Was the grain in the release prints even heavier than Kubrick intended or wanted?  Hard to say -- some people who saw a contact print said that the grain wasn't so strong as in the release prints.  I also heard a rumor that Kubrick shot the negative in some odd way - common top framing between 1.85 and TV, or Full Aperture, etc. - that required that the IP be made in an optical printer as with Super-35 movies. Optical printing can have the effect of sharpening the same grain compared to contact printing.  This grain sharpening effect may have been unintended by Kubrick since he never got to the finishing stages.

 

3. Did Kubrick want so much grain reduction applied to the video transfer?  Hard to say again, since we can't talk to Kubrick.  My suggestion would be, when in doubt, just faithfully represent in the video master the amount of grain in the negative, which would be heavier than what was released on home video and less than what was seen in the release prints.


  • 1

#7 cole t parzenn

cole t parzenn
  • Basic Members
  • PipPipPipPip
  • 287 posts
  • Other

Posted 22 October 2014 - 12:39 PM

I also heard a rumor that Kubrick shot the negative in some odd way - common top framing between 1.85 and TV, or Full Aperture, etc. - that required that the IP be made in an optical printer as with Super-35 movies. Optical printing can have the effect of sharpening the same grain compared to contact printing.  This grain sharpening effect may have been unintended by Kubrick since he never got to the finishing stages.

 

The AC article indicates otherwise: "Kubrick framed Eyes Wide Shut in the standard 1.85:1 format;" I wonder how that rumor started. Could you elaborate on the effects of optical printing?


  • 0

#8 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 18789 posts
  • Cinematographer
  • Los Angeles

Posted 22 October 2014 - 01:33 PM

"Framed for standard 1.85" doesn't rule out shooting in a non-standard way, "framed" being the operative word.

When copying an image using an optical printer instead of contact printing, each frame is focused through a lens onto the receiving piece of film, which can give the grain a crisper edge than in a contact print.
  • 0

#9 Tyler Purcell

Tyler Purcell
  • Sustaining Members
  • 2363 posts
  • Other
  • Los Angeles

Posted 25 October 2014 - 02:24 AM

Funny enough, I read Kubrick shot Eyes Wide Shut in full frame in an ASC article. He did that on many films including; The Shining and Full Metal Jacket due to his belief that cinema's of the day were inept at presenting wide screen images to specific aspect ratio's. He simply framed scenes for 1:85 and if the theater used the wrong cropping (1:66 or 1:85 or full frame) it didn't matter. It's the same reason why most of his 35mm release prints were in Mono, yet the film's were all meticulously produced in stereo and converted. Kubrick believed most theater's wouldn't have stereo audio, so it would be better to mix in mono to produce accuracy across the board. 

 

It surprises me Kubrick never returned to 65mm shooting after 2001. Projecting on 70mm would give him all the specifics he needed, but I gather his budgets weren't quite big enough. I for one would love to have seen Barry Lyndon shot in 65mm… that would have been very interesting. ;)


Edited by Tyler Purcell, 25 October 2014 - 02:26 AM.

  • 0

#10 Mark Dunn

Mark Dunn
  • Basic Members
  • PipPipPipPip
  • 2182 posts
  • Other
  • London

Posted 25 October 2014 - 04:14 AM

 He simply framed scenes for 1:85 and if the theater used the wrong cropping (1:66 or 1:85 or full frame) it didn't matter.

It did matter to him - he used to send notes to cinemas with precise details. He reportedly even sent aperture plates to some.

His letter for Barry Lyndon says it was shot 1.66 and should not be projected at less than 1.75. I have a frame enlargement with gate hairs suggesting that it was shot hard matted, not 1.37 as The Shining undoubtedly was. It's about 1.6:1 but the edges are cropped.


  • 0

#11 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 18789 posts
  • Cinematographer
  • Los Angeles

Posted 25 October 2014 - 10:27 AM

"Full frame" is a bit vague of a term.  What Kubrick did in "The Shining" and "Full Metal Jacket" was compose for 1.85 cropping but let the whole Academy 1.37 area get exposed, which was fairly common.  In fact, most 35mm films exposed the silent "Full Aperture" 1.33 area of the 4-perf negative but since the lenses were shifted over for the sound aperture, the fact that the soundtrack area was exposed didn't really matter.

 

Almost no theater would "accidentally" show a 1.85 film without some widescreen mask (1.85 or 1.66) because if you pulled it completely you'd see the soundtrack stripe on the screen and if you put in the thin Academy 1.37 mask the image would still be taller than the screen and spill off above and below the screen masking.  Kubrick was very meticulous about such things so it's not accurate to say he didn't think it mattered if the wrong mask was used.

 

For example, he shot "Clockwork Orange" and "Barry Lyndon" (and parts of "Dr. Strangelove") with some degree of in-camera masking (hard matted gates) and wanted "Barry Lyndon" to be shown worldwide with a 1.66 mask in the projector and sent a Warner Bros. rep to every theater showing "Barry Lyndon" to make sure they had a 1.66 mask.

 

But "The Shining" and "Full Metal Jacket" were shot without any camera mattes and just composed for cropping in projection to 1.85.  The confusion stems from the fact that for years, he insisted that any 4x3 video transfers of those two movies be made without any letterboxing, thus showing a 1.33 image with a lot of headroom.  With the other movies shot with some camera matting, he wanted the 4x3 video transfers to show those, so you see a small amount of masking in the image, sometimes variable depending on the camera he was shooting with.  Lots of theories as to why Kubrick avoided letterboxing.  One was that he loved the old 1.37 Academy format and a TV presentation was one way of getting a version of the movie to look like that.  Another was just that he hated electronic matting for some reason.

 

He started to shoot "Full Metal Jacket" in 65mm b&w for a week and then switched to standard 35mm color.  He also planned to shoot "Napoleon" in 65mm and owned some 65mm cameras and lenses for that project.

 

All of this just shows that Kubrick had a repeated style but also had variations in that style from project to project.


  • 0


Visual Products

Glidecam

Aerial Filmworks

Abel Cine

Willys Widgets

Ritter Battery

Zylight

rebotnix Technologies

Paralinx LLC

Broadcast Solutions Inc

The Slider

Technodolly

Rig Wheels Passport

CineTape

Pro 8mm

Tai Audio

CineLab

Broadcast Solutions Inc

Technodolly

CineLab

Zylight

Pro 8mm

Visual Products

Glidecam

CineTape

Tai Audio

Paralinx LLC

Abel Cine

The Slider

Rig Wheels Passport

Willys Widgets

rebotnix Technologies

Aerial Filmworks

Ritter Battery