Jump to content


Photo

DI Resolution


  • Please log in to reply
77 replies to this topic

#1 Rob Webster

Rob Webster
  • Basic Members
  • PipPip
  • 53 posts
  • 1st Assistant Camera

Posted 28 December 2009 - 12:58 PM

Hi Guys,

I'm writing my final year dissertation on the benefits of DI to contemporary cinematography, and during my research i found a bit of a disagreement over resolution.

Obviously with a traditional lab process there is huge loss of resolution between OCN, interpos, interneg, and eventually release prints (i've read up to 0.65 of original resolution).

However, a 2K DI also looses resolution over 2 generations of duplicating- some estimates put release print resolution of 2k prints at closer to 1.4K, which is much less than a S35 negative duped chemically over 4 generations to print at approx 2.4K.

So if S35mm film is approximately equal to 4K (again, this is disputed, Oliver Stapleton reckons its closer to 8k, Deakins rates 3K), then a 2K DI is roughly appropriate.

Just wanted to see what people think on this issue.

Firstly, what kind of resolution does everything think S35mm is equivalent to?

What is an ideal resolution for DI (assuming no high res VFX are required).


Obviously, this is purely academic and not really being used for testing or any practical work, so all opinions are welcome.

Thanks everyone, and merry christmas

Rob
  • 0

#2 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 28 December 2009 - 01:51 PM

One problem is that actual image resolution varies throughout a movie depending on lens used, film stock, exposure, filters, scene contrast, and most importantly, focusing. And of course resolution of presentation technologies also vary wildly.

So all you are going to get are opinions based on observation.

As you say, 35mm release print resolution, once it has gone through and IP and IN, and then projection, is quite low. But this tells us that 2K D.I.'s are probably a mistake IF any release prints are going to be struck through an IP/IN, which most of them are. Because if you start with 2K, you end up at a much lower level. On the other hand, if the release is only going to use 2K digital projectors, then a 2K D.I. makes some sense (ignoring the long-term archival issues of mastering at a lower resolution than the original film contained). And if the release is only going to be from show prints made off of a digital negative original, then a 2K D.I. isn't so objectionable, though a 4K D.I. would still be preferable.

But this assumes some sort of perfect 35mm image that is as sharp as possible. I sort of agree with Roger Deakins in that Super-35 photography hovers around the 3K range on average but even this suggest that scanning resolution should be higher than captured resolution.

Personally, I think the advantage of a 4K standard for 35mm scanning, finishing, and digital projection in theaters at least covers most best-case scenarios in that even if the true resolution of the image falls below 4K, we know that we aren't losing anything in the chain (it won't convince those that feel that 35mm is more like 6K however). 4K mastering also suggests that the archived digital master contains all the detail captured in the original.

Once we get to a 4K standard, we can argue about whether scanning should be done at even higher resolutions, like 6K, and then downsampled to 4K. Also, for a 4K RGB standard, you'd probably want to be using digital cameras that are in the 5K to 6K RAW level of resolution.

This isn't accurate or even provable, but my mental checklist is that 35mm negative is 4K at best (3K on average) whether or not you want to argue for even higher scanning resolutions, and that 35mm answer print resolution is 2K at best, and a 35mm release print from an IP/IN is less than that, let's say 1K.

And I'd roughly double all of that for 5-perf 65mm, that ideally it would go through an 8K D.I. and that 4K projection would be similar to a 70mm answer print.

But the only thing I'm really positive about is that we need to get away from 2K as a standard for mastering cinema movies at, especially those shot in 35mm (for those shot in 1080P, I'm a little less concerned.)
  • 0

#3 Paul Bruening

Paul Bruening

    (deceased)

  • Sustaining Members
  • 2858 posts
  • Producer
  • Oxford, Mississippi

Posted 29 December 2009 - 02:39 PM

Hello Rob,

It is common and practical to rate film at an equivalent digital resolution number. It is necessary so that people can make decisions that lead to profitable product. However, it is a bit more correct to think of it this way:

Film is a pan-resolution image. The grain clusters/dye clouds vary in size. Some are tiny. Some are comparatively large. Therefore, the issue is not, "What is 35mm film's equivalent digital resolution?" It is more like, "What digital resolution will satisfactorily represent the film's images?" For the most part, producers are using 2K for that representation. The decision has a bit more to do with the practical aspects and costs of computing than quality obsessions.

In the scan research that I have performed I found that it takes something more like 64K to capture some amount of those smallest grain/clouds. Even then, the tiniest 5% of grain/clouds were only represented with a pixel proxy version of it. They would require about 256K to be represented by a differential cluster of four pixels each. Of course, who can sustain a resolution of 256K in the current digital and computer environment?

So, in effect, all digital representations of film images are a collection of "pixel proxies" (pixels that collectively create a reproduction of the original film image but rarely accurately demonstrate an individual grain/cloud). Pixel proxies of a particular resolution are chosen on a "good enough" basis. I agree with David that 2K isn't enough from a quality basis. I'm in the 6K downed to 4K camp. But DP's don't often get to make this kind of decision. Producers do. They look at the cost difference between 2K workflows and 4K workflows. Then they look at the quality differences. They often choose 2K.
  • 0

#4 Phil Rhodes

Phil Rhodes
  • Sustaining Members
  • 11936 posts
  • Other

Posted 29 December 2009 - 02:52 PM

I think the main difficulty with discussions like this is that 35 itself is so variable. The results I see if I watch something at my local fleapit will be very considerably different than those I see at the Arclight in Hollywood. A non-technical friend of mine saw Avatar locally today and complained vocally of "this weird thing where you can see two pictures at once", so the technical standard really isn't that high no matter what they're doing. Watching 35 there probably yeilds, to my subjective eye, about half the resolution of what I see for the same productions in LA. The differences are presumably projector and lamphouse maintenance, lenses, and perhaps poorer prints.

P
  • 0

#5 Keith Walters

Keith Walters
  • Sustaining Members
  • 2219 posts
  • Other
  • Sydney Australia

Posted 29 December 2009 - 06:20 PM

Hi Guys,

I'm writing my final year dissertation on the benefits of DI to contemporary cinematography, and during my research i found a bit of a disagreement over resolution.

Obviously with a traditional lab process there is huge loss of resolution between OCN, interpos, interneg, and eventually release prints (i've read up to 0.65 of original resolution).

However, a 2K DI also looses resolution over 2 generations of duplicating- some estimates put release print resolution of 2k prints at closer to 1.4K, which is much less than a S35 negative duped chemically over 4 generations to print at approx 2.4K.

So if S35mm film is approximately equal to 4K (again, this is disputed, Oliver Stapleton reckons its closer to 8k, Deakins rates 3K), then a 2K DI is roughly appropriate.

Just wanted to see what people think on this issue.

Firstly, what kind of resolution does everything think S35mm is equivalent to?

What is an ideal resolution for DI (assuming no high res VFX are required).


Obviously, this is purely academic and not really being used for testing or any practical work, so all opinions are welcome.

Thanks everyone, and merry christmas

Rob

The one thing that has some of banging our heads against the nearest solid object in frustration is the boneheaded notion notion that "4K" resolution is the same thing as "4,000 lines"!
"4,000 lines horizontal resolution" means that the imaging medium can capture a pattern of 2,000 vertical black lines on a white background (or white on black, whatever you prefer).
But that is not the same thing as having 4,000 horizontal pixels.

This subject has been flogged to death on this forum previously. You can read one of my more in-depth responses to a very similar question here (Post #28).
Actually the whole thread is worth a read, it's pretty entertaining :D

Edited by Keith Walters, 29 December 2009 - 06:22 PM.

  • 0

#6 Rob Webster

Rob Webster
  • Basic Members
  • PipPip
  • 53 posts
  • 1st Assistant Camera

Posted 31 December 2009 - 12:15 PM

Thanks for the responses so far everyone,

As cinematographer's then, what would you say the main reason for using a DI on a film would be?

Resolution? (assuming we agree a digitally projected 4k print IS better than a s35 neg processed through IP/IN)
Creative Controls? Replicating old stocks (i.e. the Rob Richardson's work on The Aviator, being able to replicate technicolor two and three strip); Colour Timing that would be IMPOSSIBLE in the lab (i.e. O' Brother Where art thou?), Saving time on set by flagging, finessing contrast in post, power windows, luma-keys etc

Do you think there is a difference between "resolution" and "quality"? Deakins touches on this in his DI article, but do you feel that if a picture was finished chemically, and timed to match exactly with a DI (resolution difference is negligible), the chemical version would be better "quality"? I'm just interested to know whether the innate "quality" of having everything going through celluloid is (a) noticable and (B) desirable (in comparison to the digital intermediate route).

Do you feel the DI will lead to the democratising of the cinematographers role?

Thanks again for all your responses.

Happy New Year,

Rob
  • 0

#7 Paul Bruening

Paul Bruening

    (deceased)

  • Sustaining Members
  • 2858 posts
  • Producer
  • Oxford, Mississippi

Posted 31 December 2009 - 12:43 PM

Thanks for the responses so far everyone,

As cinematographer's then, what would you say the main reason for using a DI on a film would be?

Resolution? (assuming we agree a digitally projected 4k print IS better than a s35 neg processed through IP/IN)
Creative Controls? Replicating old stocks (i.e. the Rob Richardson's work on The Aviator, being able to replicate technicolor two and three strip); Colour Timing that would be IMPOSSIBLE in the lab (i.e. O' Brother Where art thou?), Saving time on set by flagging, finessing contrast in post, power windows, luma-keys etc

Do you think there is a difference between "resolution" and "quality"? Deakins touches on this in his DI article, but do you feel that if a picture was finished chemically, and timed to match exactly with a DI (resolution difference is negligible), the chemical version would be better "quality"? I'm just interested to know whether the innate "quality" of having everything going through celluloid is (a) noticable and (B) desirable (in comparison to the digital intermediate route).

Do you feel the DI will lead to the democratising of the cinematographers role?

Thanks again for all your responses.

Happy New Year,

Rob


Control has got to be the number one reason. DI gives a project powers that are difficult to impossible in an all optical route. The big categories include: Color timing (color grading), image adjustment (gamma, brightness, color replacement, texture, anything a bit of software can do to change the "look" of the image- far too many to list, here), composites, motion tracking, CGI, and easier access to the new projection mediums in varying and evolving file types and formats.

The resolution/quality topic erupts here every once in a while as it should since it is an important concern. The folks with a technical, mechanical, mathematical or scientific bent lean in favor of talk about resolution and related. I listen to their talk since I know far too little about that kind of stuff. The folks with an artistic bent tend to talk about aesthetic qualities. The technical folks kind of just put up with us. I'd suggest you do some searches through the site. We could pile up 1,000 posts regurgitating what we've already covered related to the resolution/quality issue.
  • 0

#8 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 31 December 2009 - 12:44 PM

Main reasons for me:

1. Image manipulation -- you can adjust parts of the frame separately from other parts, you can adjust contrast, black level, and saturation to get shots to match each other better, you can fix color problems. You can also do more extreme manipulations more easily (desaturation, etc.)

2. You only color-correct the final movie once for both theatrical, home video, trailers, commercials, etc. rather than go through the photochemical color-timing process and then the home video color-correction process. Of course, this isn't exactly true because these different formats use different color spaces, but usually the conversion is pretty straight-forward and doesn't involve retiming the movie from scratch

3. Avoiding optical printer dupes and blow-ups. Not just for the Super-35-to-anamorphic conversion, but also for "opticals" -- transitions like fades, dissolves, etc. Though many of those could be done with A-B roll printing, a lot of the studios require a single-strand negative be delivered.

4. Enabling a digital cinema release. I prefer digital projection over 35mm release print projection in most theaters, due to the fact that most prints still come from IP/IN dupes and many theaters have lousy 35mm projectors and have no trained projectionists.

5. Ability to mix formats.

6. Ability to mix titles and visual effects at the same generational quality and match them to surrounding non-efx footage.

Main disadvantages for me:

1. There is a "deadening" or flattening effect caused by the process of recording digital files to color intermediate stock -- I can't put my finger on it, but it could be due to the density standards that everyone uses for optimal LAD figures, I don't know. Photochemical dupes through the IP/IN process suffer a similar fate so it could be the nature of dupe stock.

2. Deviations from normal color response of the stock. The flip side of being able to manipulate the color, contrast, saturation, of a stock is that you can so easily drift from a "normal" look that you find yourself sometimes struggling to get back to that unmanipulated look that film stock would normally give you in a contact print. The main way this problem shows up is in "unreal" looking fleshtones, which is a hard color to get right if you are doing it by eyeballing it more or less. You find yourself talking to the colorist about what color the skin should be to look normal. "Maybe it needs more magenta but less chroma?" The end result may be a painted "band-aid color" look to the skin, almost like a tinted b&w photo.

3. While the graininess of Super-35 conversions to anamorphic have improved due to the use of D.I.'s rather than optical printers, thus improving the format, the D.I. process has also reduced the quality of 35mm anamorphic photography, either due to lower resolutions being used (i.e. 2K) or just that deadening effect of recording to intermediate stock. The net result is that Super-35 and anamorphic are closer in quality than they should be, making it harder to make the case to shoot anamorphic. Combine that with the fact that D.I.'s have made shooting 3-perf as easy as shooting in 4-perf, to the point where many producers automatically budget for 3-perf, and thus if you want to shoot anamorphic, you have to convince them to spend the money on 4-perf as well.

4. 2K as a standard. Bad idea unless maybe if you are only shooting in Super-16 and HD cameras.

5. The high cost of doing D.I.'s, especially if you want to scan at 4K, etc.

---

The "quality" issue is hard to define. If you are comparing an optical printer blow-up of Super-35 to a D.I. conversion, it's a bit apples and oranges because the optical will be grainer but it may look better in other ways.

I don't see much of a "democratising" effect -- bigger-budgets still get to use better equipment, have longer color-correction times, can perhaps make all release prints from multiple digital neg originals, etc. I've even noticed lately in my movies that have gone to DVD that lower-budget movies have worse DVD masters in terms of the bit rate, etc. So the "ghetto-ization" of independent filmmaking never seems to stop.

Now I suppose for the bottom-end of things, new technology has allowed them to work at a higher level. That's the nature of digital -- it lifts the bottom to the middle but it doesn't lift the middle to the top -- the top will always be rare because at that level, you have a unique combination of exceptional human skill combined with working with the best tools made. I still don't think there has been a modern digital sound mix to match what Walter Murch did in analog for the 6-track 70mm version of "Apocalypse Now", but your average Adam Sandler movie sounds better that it would if made in the 1970's. The bottom rises to the middle -- now your average low-budget movie can mix many tracks of sound, etc.
  • 0

#9 K Borowski

K Borowski
  • Basic Members
  • PipPipPipPip
  • 3905 posts
  • Camera Operator
  • I.A.T.S.E. Local # 600 Eastern Region

Posted 31 December 2009 - 02:10 PM

1. There is a "deadening" or flattening effect caused by the process of recording digital files to color intermediate stock -- I can't put my finger on it, but it could be due to the density standards that everyone uses for optimal LAD figures, I don't know. Photochemical dupes through the IP/IN process suffer a similar fate so it could be the nature of dupe stock.


Very democratic of you coming up with (almost) as many discrete disadvantages of DI as disadvantages to contact-printing, David, but I think you are exaggerating some of the weaknesses of DI to make up this list. I'd say that there are fewer big problems, maybe the two you mentioned about color response and killing off a lot of real anamorphic work. BUT, the problems with digital color are still huge on some of the films I've seen.

I think contact still looks better than just about all 2K (except, of course, for S35 blowups), but I freely admit a photo-chemical finish is more difficult to fine-tune and has all the problems with scratches, dirt, and the risks of physical handling that film-makers want to be rid of completely these days.


Anyway, I don't think that what you are referring to, this "flattening" effect is due to problems with recording to IP stock. Intermediate stock is finer-grained, slower, and despite it's low contrast can MAINTAIN contrast in a printing chain, especially if you are building it up in two subsequent stages of printing. As long as a film-recorder that has the proper software and hardware to produce an image geared to the stock's color sensitivity, I certainly don't think outputting onto camera negative, even 5201, would be better looking.

The real problems I have seen are/were in the release-printing and IN stages, which are done in a hurry with far lower quality control. Having a 1-2/3 stop drift on the "grey" dot towards red shows me that the chain completely falls apart at the end. Granted you want to see a print that is on the warm side, color-wise as opposed to being cool, but it seems like anything more than half a stop bias towards a certain color on a grey patch is just sloppy work, no matter how many thousand release prints you have to churn out. At the same time I just saw a print of "Up In the Air" that looked really cool, and immediately noticed it, so maybe I am part of the problem with this trend. Proper density of the "grey" (now pink) aim dot on most reels seems to be a much smaller problem. They tend to be printed more along the dense side, but no more than maybe 2/3 of a stop too dark.

Ideally, if you knew what lab you were using for a release print run, as they use the HIT dots and Shirleys on the ends of the reel primarily, and spot check the rest, you could just compensate for this warmness, density overcompensation by making the IP darker and warmer to compensate, but not on the charts, of course. This assumes that someone at the lab doesn't take steps to undo your work anyway. I wonder if Deluxe and Technicolor would deviate from their normal printing even if the studios told them to.


I know it is more expensive, but I don't really understand why they just strike multiple INs on film recorders, instead of making one or two IPs. It will cost more, but not nearly as much as striking every release print in a film recorder. The highest IN number I ever saw, at a second-run, was #9. Anything over a fourth or fifth internegative, at least in the U.S. first-run market, was uncommon, at least as of four years ago. So you'd have to print five INs instead of two IPs to remove a generation from the process completely. Have any films ever done this?


Even if you finished on film (to making a timed master positive), wouldn't that still be cheaper to make into a 2K or even 4K digital file than doing a DI from the negative, and having to do all the timing digitally? I am still, frankly, surprised to see movies with practically no SFX opt for DIs over traditional finishes. It is almost surprising to see a movie NOT go through a DI anymore :-(

Edited by Karl Borowski, 31 December 2009 - 02:11 PM.

  • 0

#10 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 31 December 2009 - 07:14 PM

The studios are penny-pinchers in some areas -- one of which is the cost of making multiple original digital negatives, the other is making release prints. It costs them around $10,000 to strike another IN off of an IP, but it costs them around $40,000 to record out another IN from a digital master. It's also why they may do the whole D.I. with Vision 2383 print stock as the target and then make all the release prints on Fuji or Agfa stock.

In regards to why low-budget movies shot in 35mm even go through a D.I., part of it is that many directors come from music videos and commercials and demand digital color-correction, feeling that it's the only way they can get the final look they want. The other is that many up and coming editors make decisions without a photochemical finish in mind.

I did a movie that was shot without a D.I. planned ("no way", said the producer) and then the editor and director cut it in such a way that there was no way to not finish it with a D.I. (digital zooms, shot repeats, slowdowns, speed ramps, etc.) Apparently this editor had done the same thing to the company's previous movie, forcing them to do a D.I. by the way it was cut, but they didn't learn. Some newer editors don't even know about A-B roll printing and being limited to fades and dissolves of certain lengths, etc. And they use transitional effects with a heavy hand.

I've also noticed a trend with directors to "fix" almost every shot -- they don't like the color of some shirt that an extra is wearing, they don't like some sign in the background, they don't even like the other actor in the frame anymore, creating this huge laundry list of digital efx touch-ups that then have to be prioritized and argued over, because it costs money. There used to be the attitude of working with what you shot, and if necessary, do some reshoots. Now there is an attitude that you can keep remaking the movie in post, keep making changes to the shots. Maybe I'm just old-fashioned and this is just the way all movies will be made, but it drives me a bit batty.

Most of this has nothing to do with the cinematography. So a D.I. is as much a tool for the director and editor as it is the DP's.
  • 0

#11 John Sprung

John Sprung
  • Sustaining Members
  • 4635 posts
  • Other

Posted 31 December 2009 - 07:40 PM

I'm pretty sure the major studios all have DI in their pattern budgets by now. I know for sure of one that's had it for a couple years. So, if you feel that a straight contact printing finish is best for your production, that becomes something that has to be talked about in meetings. If you go that route, take care not to leave the money on the table.



-- J.S.
  • 0

#12 K Borowski

K Borowski
  • Basic Members
  • PipPipPipPip
  • 3905 posts
  • Camera Operator
  • I.A.T.S.E. Local # 600 Eastern Region

Posted 31 December 2009 - 08:29 PM

The studios are penny-pinchers in some areas -- one of which is the cost of making multiple original digital negatives, the other is making release prints. It costs them around $10,000 to strike another IN off of an IP, but it costs them around $40,000 to record out another IN from a digital master. It's also why they may do the whole D.I. with Vision 2383 print stock as the target and then make all the release prints on Fuji or Agfa stock.


I can understand the high cost of an IN filmout, but not the $10,000 for making an IN from an IP. Isn't this just un-timed contact printing?

Granted it has a gate instead of being continuous motion, like IN to RP, but I don't understand why it would cost that much, even with a clean room environment and careful handling procedures.

Anyway, $30,000 extra is a lot better than I thought. I know this doesn't work into "pattern budgets" that John has brought up more than once, but surely you can save $120,000 elsewhere (I'd even use Agfa stock if I could save a generation, personally). I know that's not the way accountants think; maybe they should though.

Speaking of DIs, I saw a movie, actually a trailer for a movie, one of the "attached" films on Sherlock Holmes here in the U.S. that looked to be a DI but the trailer was "full aperture." I would say I could tell it is a DI because they had some sort of numerical information at the top of the frame, movie started slightly out of frame, and then the projectionist apparantly had trouble framing it himself, because he had to frame it up and down several times with text flashing on screen before he could tell for certain it was centered.

So everyone is scrimping to save money where image quality counts at the studios, but then they apparently are shooting soft-matted 4-perf. for the DI :blink:



As for Vision versus Fuji release printing, isn't it still just a matter of getting the grey HIT at the tail of each reel to register as grey? I know intermediate stocks aren't necessarily designed for ease-of-use with a competitor's product, but the color shifts I saw on release prints were SHOCKING sometimes, not three or four points of color shift (by point I mean thirtieth of a stop).

The only possible reason I can think of that release prints would be off as far as they are is if the charts were printed inconsistently from the actual on-screen material. But this is more or less what I saw with every print.


I would *hope* that this all happens at the IN to RP stage, but I suppose color discrepancies could creep in earlier in the chain.
  • 0

#13 K Borowski

K Borowski
  • Basic Members
  • PipPipPipPip
  • 3905 posts
  • Camera Operator
  • I.A.T.S.E. Local # 600 Eastern Region

Posted 31 December 2009 - 08:32 PM

If you go that route, take care not to leave the money on the table.



-- J.S.


I'd go the optical route and then use the money I saved specifically for bribes to allow the negative to be delivered A-B and to get the lab tech at Deluxe or Technicolor to make sure the grey dot was within five points of grey :P
  • 0

#14 Keith Walters

Keith Walters
  • Sustaining Members
  • 2219 posts
  • Other
  • Sydney Australia

Posted 31 December 2009 - 08:56 PM

Do you feel the DI will lead to the democratising of the cinematographers role?

Not sure how that follows, since in any significant production you are talking about two entirely separate parts of the process.

"Democratising" is one of those unfortunate "house-of-mirrors" expressions that crop up all the time on various cinematography-related forums. Like many similar expressions they are more likely to have originated in the overactive imaginations of barstool producers than industry people.

The usual implication is that by reducing the cost of just one small part of the cinematic process, (camera, post production, distribution etc) it's suddenly going to some become possible to make a $100 million movie for $100 thousand (or on certain forums, $10,000) :lol:

Essentially the DI process allows movies to be post-produced the same way TV programs have been post-produced for decades, but at full cinema resolution, and in theory on someone's desktop PC. The reality is currently very different.

The problem is that neither the cost of cameras and recording media, nor the cost of post production have any massive impact on the the overall cost of making cinema release films. Sure, there have been a few gimmicky exceptions, but in general, much of the cost is in paying all those names that roll on for several minutes at the end of the average feature....

They're going to be there regardless of what the feature is shot/posted on.

Edited by Keith Walters, 31 December 2009 - 08:59 PM.

  • 0

#15 Dominic Case

Dominic Case
  • Basic Members
  • PipPipPipPip
  • 1357 posts
  • Other
  • Sydney Australia

Posted 01 January 2010 - 12:17 AM

There used to be the attitude of working with what you shot, and if necessary, do some reshoots. Now there is an attitude that you can keep remaking the movie in post, keep making changes to the shots. Maybe I'm just old-fashioned and this is just the way all movies will be made, but it drives me a bit batty.

Well said David! But unfortunately it's the modern way of thinking. Keep options open until the very end, postpone any decision until the last possible moment (or later). I don't see any benefit that has come from this.
  • 0

#16 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 01 January 2010 - 01:37 AM

Well said David! But unfortunately it's the modern way of thinking. Keep options open until the very end, postpone any decision until the last possible moment (or later). I don't see any benefit that has come from this.


I think it's just a new way of thinking about cinema, post George Lucas, that what you shoot is merely raw material for future image creation. Which is fine for a fantasy / sci-fi movie that will have to create wild environments with CGI anyway, but for a run-of-the-mill narrative, it's a bit annoying, this constant need to change the shot, not necessarily for the better, but to change it because you can. Directors start to feel that they won't really live or die by their choices on the set because there is always post to change their minds about everything -- props, color of wardrobe, etc. I've even seen whole characters erased out of a scene.

It even starts to drive editors nuts because rather than make good edits based on the material, the director or producer asks them to make bad cuts because they've changed their minds about the scene, about the dialogue, about the staging, about the cast, etc. So the editor gets asked to cut out and around what was shot, making cuts that they would never normally make, like cutting to the back of someone's head so they can slip in new dialogue, or jump-cutting an actor as they move across the room in order to get rid of dialogue.

I've seen wide shots cut out because they've decided they don't like the set, and they don't like the time of day established, so they only use the close-ups, and then say they don't like the make-up or hairstyle and want to digitally alter it. They start fixating on unimportant and minor facial flaws that they would normally have to live with in a contact print.

The end result is 180 degrees opposite from the controlled filmmaking of the past. Imagine Gordon Willis ever having to put up with this.
  • 0

#17 Keith Walters

Keith Walters
  • Sustaining Members
  • 2219 posts
  • Other
  • Sydney Australia

Posted 01 January 2010 - 02:54 AM

Hi Guys,

I'm writing my final year dissertation on the benefits of DI to contemporary cinematography, and during my research i found a bit of a disagreement over resolution.

I don't want come across as patronizing and I don't know the actual details of what institution you are producing the dissertation for, but it occurred to me that you might have to be a bit careful how you present the information you get here.

There is absolutely nothing wrong what anybody has said on this thread thus far (it's still early days :rolleyes: ), but as you've noticed, it does tend to conflict somewhat with the current body of "common knowledge" on this subject.

I've run into problems in the past helping kids with homework assignments where I know from personal experience the information I have given them is 100% accurate, but they got poor marks because the ignorant teachers "knew better". And this was from an expensive private school.

You've probably realized this already, but part of your dissertation must include an acknowledgement that the most commonly perceived version of cinematic "reality" conflicts somewhat with the real situation, and appears to be in large part the wishful thinking of cinematic wannabes. Even the mighty Scientfic American aren't immune from this, and routinely ignore letters to the editor to the contrary.

Actually the whole "wannabe vs reality" syndrome itself would be an excellent subject for a scholarly dissertation. :rolleyes:
  • 0

#18 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 01 January 2010 - 12:35 PM

If he's not writing from personal experience, then I think he simply has to present the best arguments from a wide range of "experts" (with citations, footnotes, whatever) even when they are contradictory, to create a complete picture of the state of the industry and current trends and opinions. He can even go back to these experts a second time to question them about conflicting data and opinion.

The only danger is trying to draw firm conclusions; I think that's the point where the paper could get into trouble if it attempts to say THIS or THAT is the final truth on the subject.

I'd take a look at some of Matt Cowan's articles on the subject:
http://digitalconten...inemas_special/
http://www.etconsult...0Resolution.pdf
http://digitalconten...ideo_di_primer/
  • 0

#19 Rob Webster

Rob Webster
  • Basic Members
  • PipPip
  • 53 posts
  • 1st Assistant Camera

Posted 01 January 2010 - 01:34 PM

I'm not at all intending to draw any conclusions about whether DI performs better of worse than photochemical processing. My argument is more along the lines that the DI is a tool that should be used discerningly, and why it has its applications in certain scenario's and for certain people. Presenting contrasting opinions over the reality of "resolution" and "quality" will definitely help me, as i have not been able to do a digital intermediate on any of my films (film school budget doesn't quite stretch that far), hearing the experiences of those who have (even if they are more emotive responses about the "feel" of a DI print for example) is what i am interested in.

Aside from all the data, there is an argument for and against the DI that is purely based on personal experience and preference, and i think these ideas are equally valid.
  • 0

#20 Rob Webster

Rob Webster
  • Basic Members
  • PipPip
  • 53 posts
  • 1st Assistant Camera

Posted 01 January 2010 - 01:41 PM

And of course all opinions and information will be acknowledged properly
  • 0


rebotnix Technologies

Broadcast Solutions Inc

Ritter Battery

Media Blackout - Custom Cables and AKS

FJS International, LLC

Aerial Filmworks

CineLab

Metropolis Post

Glidecam

Paralinx LLC

Wooden Camera

Opal

Rig Wheels Passport

Technodolly

Willys Widgets

Gamma Ray Digital Inc

Tai Audio

The Slider

CineTape

Abel Cine

Visual Products

Ritter Battery

Glidecam

FJS International, LLC

Technodolly

Willys Widgets

CineTape

Media Blackout - Custom Cables and AKS

Metropolis Post

Tai Audio

Broadcast Solutions Inc

Paralinx LLC

Abel Cine

The Slider

Visual Products

rebotnix Technologies

Gamma Ray Digital Inc

Opal

Rig Wheels Passport

Aerial Filmworks

CineLab

Wooden Camera