Grey Scale Imaging Performance

In our industry, we constantly decry the absence of "standards", especially in HD imaging and recording formats. Yet when we talk about what should be the basic simple terminology of our business, there are some terms which have wildly different meanings based on context or purpose. Proper depth of field. for instance, is used to describe a condition in which everything in frame is in focus as in Collateral and always as always attributed to 2/3" HD cameras; other times it is used to describe a condition in which only the subject of the story is in focus; also attributed to large format, single sensor cameras. All due to context and purpose and design. Everyone has an explanation for why they use certain terms in certain ways, but that does not obviate the confusion in language and application. In any even, most of us here understand the distinctions and make the translations to our frame of reference.

Not so well understand, I believe, and my particular pet peeve is the use of "rich" Blacks, especially in the modern era, to describe the use of either lighting or electronic processing that yield, in some cases on purpose or in others by limitations, depressed gray scale performance in the Gray and lower Gamma areas. When the story requires the feel of deep, shadowy greys and black, it is perfect. When that requirement is not present, it can be a hindrance, and even a negative attribute. In the modern era, we can measure gray scale values digitally by the bit, and the number 64 is the number commonly referred to as black, and the numbers just above it in the lower Gray and Gamma areas. But there is a very commonly held notion that "rich" blacks are the goal of every proper image. Not true.

The "milky" blacks alluded to in previous "Collateral" posts simply reflect the actual bit values present even at night in Los Angeles. The creative decision in that case was to actually allow the movie-goer the option of seeing the "place" and "character" of the city as the eye is able to, without losing those bits through natural or manipulated compression as some might have chosen to do in the name of "rich" blacks.

Compressed blacks is not an uncommon phenomenon in modern digital imaging, as we found out when we did our gray scale testing on Collateral. Any modern digital camera has the ability, through a combination of proprietary sensor and processing capabilities, to do just that either by simply using the basic setup of the camera or through the use of Black Level, Black Press, Black Gamma, and Gamma controls. The comments of Bob Kertesz and others here reveal that some cameras have less processing limitations than others, in the lower gray scale area. That is exactly what we found in the camera tests for Collateral. Even when exact gray scale and matrix matches were achieved with a DSC chart, there were profound differences between the (directly observed and filmed out) picture performances of the Viper and the F900, to the extent that in was simply not possible to reproduce some of the low end gray scale and even chromatic values on the F900. Its blacks were, indeed "rich", but suffered from an inability to reproduce the entire lower dynamic range. The criteria for initial acceptability of a digital camera was whether the digital output, when filmed out and printed closely mimicâ€™d the EK5218 printed. It was in this area that the Viper was stronger. There were other criteria, of course, such as cleaner, more artifact free resolution (resulting from the cameras' native aperture/Contour/Detail characteristics), and the ability to do 2.37:1 aspect ratio, both of which the Viper excelled in even when one notes that the 2.37 mode does slightly limit the highlight handling range of the Viper. The F900, even without wide-screen capabilities, had strengths, of course, that made it very useful for the portion of the film in which it was used.

Even in the pedestrian world of everyday sports television, it has become trendy to accept or create "too dark", gray scale compressed areas in images. When Thomson (formerly Philips) LDK HD and SD cameras are used, most Video Controllers routinely suppress the blacks to mimic the look of Sony cameras out of 1. habit or 2. some perceived notion that "rich" blacks are normal and preferable, even though lacking in gray scale reproduction capabilities.

We live in a world where even our trendy societal sweet tooth focuses on dark chocolate over milk chocolate, and many other once popular flavours. If I have one drum to beat it is that unless you can produce a full gray scale with a given imaging methodology, even when you choose to not use all of it, you are short changing your capabilities as a professional. So if you hear me whining and carping about poor gray scale performance that is what I mean. And when all cameras can do that I may not sound like such a one-sided drum beater, but until then.....

GEORGE C. PALMER
HDPIX, INC.
HD and Digital Imaging
www.hdpix.com

>> my particular pet peeve is the use of "rich" Blacks

It's certainly my tendency in TK, well actually every colorist I work with has a tendency, to immediately alter the lower end of the gamma.

Bring the lower greys down and partially crush the blacks.

It does indeed give a "punchier" image.

Of course I've also got great looks going the other way, lifting the gammas right up and then correcting the extreme blacks back down gives a great cosmetics look.

A lot of it is what's fashionable this week.

We need origination systems that can give us images that can be twisted into whatever is this weeks fashion.

Cheers

Geoff Boyle FBKS
Director of Photography
EU Based
www.cinematography.net

Geoff Boyle writes :

>>A lot of it is what's fashionable this week.

My favourite flavour of the week was the era where the agency guys would say "We like what we saw on the Avid. Can you make it look just like that?"

Steven Poster ASC

GEORGE C. PALMER wrote:

>>...Proper depth of field. for instance, is used to describe a condition in >which everything in frame is in focus as in Collateral and always as >always attributed to 2/3" HD cameras;...

Never heard of the term. "Deep focus," yes. "Proper depth of field" is even more vague than "clear pictures."

>>...other times it is used to describe a condition in which only the subject >of the story is in focus;...

Never heard that, either. The term is, "selective focus." I guess George has to work with some image illiterates. I know nothing about COLLATERAL except what I've read on this list, but don't recall anyone mentioning "proper depth of field," maybe I missed it. They simply described the great depth of field on many shots. Some merely observed that to achieve it the F900 must have been used, a reasonable assumption.

>>...The "milky" blacks alluded to in previous "Collateral" posts simply >reflect the actual bit values present even at night in Los Angeles....

I don't know what the posters saw that they described as "milky blacks." They certainly could be expecting crushed blacks, since that is such a popular approach. But that term should be reserved for blacks that have been lifted to the point that they are transparent but lacking in any detail, i.e., they should be black. A milky black is a black trying to become a lower gray scale tone, but doesn't qualify. Any tone lacking in texture or detail (if that part of the subject contains it) must be black. The fact that it lacks detail means it is black. Underexpose film and print it up, you'll get milky blacks.

>>...We live in a world where even our trendy societal sweet tooth focuses >on dark chocolate over milk chocolate, and many other once popular >flavours....

Then why do I have to search in vain for the dark chocolate amongst the milk chocolates? I haven't found this trend.

But actually I'm on your side, George. The ability of the equipment to render a full gray scale should be one of the primary capacities of the system and the use of it can be very beautiful.

Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614

Thanks for the feedback Geoff. Punchier images seem to be part of the escalation of the senses that everyone seems to pursue in art and cinema today. Notice that your colourists all seem to automatically do those things.

Did you ask them to do that, or do you think they just presumed that was what you wanted or that "everyone" would want that look?

GEORGE C. PALMER
HDPIX, INC.

One of the things I found visually interesting in "Collateral" was the texture represented in the LA night sky, which like the night sky of any city is, of course not "black" (barring major power outage).

It made me think of the haunting opening of William Gibson's "Neuromancer" : "The sky above the port was the color of television, tuned to a dead channel"

I mean, we tend to think of film characteristic curves as "natural" "organic" but they editorialise, albeit in interesting ways, do they not ?

Sam Wells

>>One of the things I found visually interesting in "Collateral" was the >texture represented in the LA night sky, which like the night sky of any >city is, of course not "black" (barring major power outage).

The LA skyline is anything BUT black. In truth, the LA night skyline hovers around the dangerous areas of all forms of photography, but then photographers have a fascination with the dangerous. It is seductive to consider your options as a Cinematographer setting out with the task of giving some kind of new life or new look to the oft photographed LA cityscape, and Collateral succeeds on a grand scale at rendering that cityscape in a new way.

One can only wonder what legacy Weegee would have after wandering the streets of LA with a Viper set to FilmStream mode and a disk array to record on.

Dave Stump ASC
VFX Supervisor/DP
LA, Calif.

>...interesting in "Collateral" was the texture represented in the LA night >sky... of course not "black"

> "...The "milky" blacks alluded to in previous "Collateral" posts simply >reflect the actual bit values present even at night in Los Angeles..."

Don't know about anyone else, but those are not the tepid blacks which I saw in "Collateral" - and I think a print on Premiere could've solved it. I can appreciate the glow of the sodium & mercury vapour contaminating the city skies, and being able to barely expose for them, but the bottom-end of the picture was often not at black, but pushed to a higher point in the realm of "black/gamma" but without any detail to show for it.

Put it this way, hopefully one day with a single chip camera that's 800-1000 ISO out-of-the-box, and with a 50mm t1.0 prime, and the shutter at 1/30th and the gain lifted, all recorded 4:4:4...Well then we're not going to have this problem. Either that or we need more street lights.

>One can only wonder what legacy Weegee would have after wandering >the streets of LA with a Viper set to FilmStream mode and a disk array >to record on.

Weegee with a Viper documentary would be interesting although it seemed he loved to "poof" his subjects with a flash and capture that perfect moment (he liked to freeze the action with a tight shutter speed & flash when needed - and I suspect he liked tight grain too). I've always loved Weegee's work, more for his camera placement where everyone in the frame has the most perfect expression frozen in time - he really had a knack for that. Sort of anti-Cartier-Bresson but in a good way.

You know who would've loved it : Camille Silvy. Did a phenomenal photograph of a Parisian street at twilight that could've been an inspiration for "Collateral". This was mid-1800's, film speed was surely in the single digits, the fastest lenses around 5.6+, and it required 4 negs printed together - one for the lamp-post, another for the sky, and 2 more for different parts of the street. One section had the blurred ghosts of people walking away on the sidewalk (had they crossed the street parallel to the film plane they would've disappeared altogether). Guys like him and Gustav LeGray (also from mid 1800's who did similar dusk double-exposures) would be all over the Viper... well, if they can forgive the regretful loss of detail - they were raised on large format after all and the clarity in their prints is amazing even now. I mean, its gotta be 10K ! Just saw some at the Ghetty last July.

"It is the Artist who is truthful, and photography which lies, for in reality time does not stop." -- Auguste Rodin

Mark Doering-Powell
LA based DP

David Stump wrote :

>One can only wonder what legacy Weegee would have after wandering >the streets of LA with a Viper set to FilmStream mode and a disk array >to record on.

You can be certain it would involve bent plexiglass...

Jeff "prefers the earlier Weegee to the experimental Weegee" Kreines

Mark Doering-Powell wrote :

>Just saw some at the Ghetty last July.

Is that the new museum in Compton?

Jeff "only been to the other Getty" Kreines

>One can only wonder what legacy Weegee would have after wandering >the streets of LA with a Viper set to FilmStream mode and a disk array >to record on.

IMHO Weegee's true heirs are the guys riding around all night in unmarked ENG vans, hoping to capture a horrific enough crash, killing, or fire that they can sell the clip to the affiliates morning news shows. I tried to Google up a few names - you see their bugs upper right all the time - but they only seem to show up on the web when they sue clipping services. I can't imagine them making much use of a Viper, let alone an S2 box. They seem more like the types to shoplift DV tape at Costco.

Tim Sassoon
SFD
Vfx & creative post

>One can only wonder what legacy Weegee would have after wandering >the streets of LA with a Viper set to FilmStream mode and a disk array >to record on

Thankfully WeeGee used real 4x5 negatives to capture photos that no digital could.

Walter Graff
BlueSky Media, Inc.
www.bluesky-web.com

> They seem more like the types to shoplift DV tape at Costco.

And re-use it...

Jeff Kreines

IMHO Weegee's true heirs are the guys riding around all night in unmarked ENG vans, hoping to capture a horrific enough crash, killing, or fire that they can sell the clip to the affiliates morning news shows... they only seem to show up on the web when they sue clipping services. They seem more like the types to shoplift DV tape at Costco. (SNIPPED)

---They're called "stringers" and as much as folks seem to enjoy maligning them it seems that the public appetite for their wares just gets bigger and bigger. It's just a business like any other, and in an industry where predatory behaviour is the norm, such maligning sounds a lot like the "pot calling the kettle black".

It's okay if you "look down your nose" at the lowly stringer...he could careless. He makes enough money that he probably doesnâ€™t need to rip off Costco for DV tape. Besides....most of the "predatory" clipping services tend to "look down their noses" at DV....unless it is a clip they "gotta have"...then even ordinary VHS would do.

So yes, you're correct, Weegee would fit the bill nicely, and so would anything else that fits the price/practicality requirements of stringer news services. By the way, part of the practicality would include the ability to take extremely severe physical punishment. I lost a BetaCam during the LA riots....seems they haven't made tape transports that can withstand bullets yet.

Four bullet holes in my unmarked "ENG" van too, but the van was still usable afterward.

Jeffery Haas
freelance editor, camera operator, former news stringer
Dallas, Texas

You know you can all say what you want about the blacks this - the highlights that - the grain this - the noise that...

I saw it last night, digitally at the Below The Line Screening. I thought it looked really great as far as all of that is concerned. Compositionally, editorially it was as great as Michael gets. And I thought it was a unique look at LA. And I was on the edge of my seat a number of times.

However, it still looks like video to me. And I'm worried that we may never get over that problem. It's what I've been talking about for 15 years now. There is a perceptual difference how a viewer is effected by film and how that viewer is effected by video.

Steven Poster ASC

The writer in me suggests you meant affected Steve rather effected.

The filmmaker in me suggests that while video may always look like video, there is no reason that digital capture (data, 2K, 4K, 8K, whatever) will. The experts among us may always be able to tell where the edges of the matte painting begin and end or see the seams in a special effect. However, those who can tell are few and far between when those effects are done extremely well. To predict what the future holds on the basis of one example - an example that more than a few have commented on re: the choices Michael Mann made - is impossible.

Will digital technology ever look like exactly like chemical technology? No. Though I believe for the foreseeable future digital and chemical will be married. Film or Digital for capture, digital for post, and film for release.

There are those who say that CD recording is ruinous to music - too cold without life. Only vinyl can deliver the warmth necessary for the reproduction of great music.
I have a shelf of albums dating back 30 years and the pops, clicks, scratches I can hear in those recordings outweighs all the subtle warmth that's supposedly buried in there. When I put on the CD version of the same album and hear it as I heard it when I first played that vinyl recording without the pops, clicks, scratches, etc - I'm glad I own the CD version.

Digital projection has its advantages - no scratches running through the print like the film I saw in the theatre yesterday - and it's disadvantages - no true blacks and no rest for the eye.

So at the moment, I'm glad I watch film prints most of the time no matter how the image was captured, manipulated or processed.

Robert Goodman
Author/Filmmaker
Philadelphia, PA

Robert Goodman writes:

>The writer in me suggests you meant affected Steve rather effected

You are correct...But please note I made no value judgement about the way the movie was captured. I am just concerned that we may eventually change the way movies are perceived. Or maybe it doesn't matter. I sure would love to find some way of testing that outcome.

Steven Poster ASC

I'm certain things will change. I wish I knew how to test the outcome, too.

Yet, do you really think that at the budget level of the A pictures you shoot that it makes a whit of difference whether it's film-capture and digital post or digital-capture and digital post?

The cost differences on an A picture are a rounding error in the budget.

Am I worried about digital projection in its current state? Absolutely. the weak link clearly. And the one that distributors have the most incentive to change.

Making and shipping 3,000 prints is expensive. Replicating and shipping 3,000 DVDâ€™s is really cheap.

The weather vane in this is Roger Ebert. When digital technology becomes acceptable to him, 35mm will join 65mm on the shelf. And that will be a sad day for us all.

Robert Goodman
Author/Filmmaker
Philadelphia, PA

Steven Poster ASC writes :

>However, it still looks like video to me. And I'm worried that we may >never get over that problem. It's what I've been talking about for 15 years >now.

I agree, I liked the movie, but it felt ..."strange"

John Babl
Miami

Yes,

There were many parts of the Viper footage that did look like video â€¦

But to me, that "video give-away" was the open shutter look, which to me screams 60fps video with those "film-blur" plug-ins attached to them in After Effects or whatever software package you want to give it a "film-look"; never works IMHO and looks awful.

I'm really quite curious what the perception would have been if they had stuck with a 180 degree shutter and upped the gain. Might have been noisier, but I would rather have that noise than the "smeary" look that a full-open or 270 degree shutter gives.

BTW, I myself have made the mistake of letting the shutter lie full-open with my first go-round on the F-900. Got confused by the menu saying that the shutter was on, but not realizing that there's a master switch that over-rides the menu settings. I'm never making that mistake again.

On another side note, many of the guys I work with who aren't into this whole HD-versus-film stuff, thought the movie rocked, and there was no mention among any of them about either how "video-y" the film looked-they just though it looked "cool"-whatever bearing that might have on audience perception.

Jason Rodriguez
Post Production Artist
Virginia Beach, VA

Robert Goodman writes :

>Making and shipping 3,000 prints is expensive. Replicating and >shipping 3,000 DVDâ€™s is really cheap.

Sort of, but for who? Distributors have traditionally made a killing marking up prints. A large-format distributor recently told me how CFI had ruined his business by publishing their pricing table for 70mm print - one reason he was switching to FK. The costs of any future digital media will be pretty transparent. And 65/70mm is far from dead - both CFI and FotoKem have invested heavily in large-format film infrastructure in the last couple of years.

IMHO it's time for tent pole features that can afford it to consider 65mm 5-perf origination and 4K compositing for 35mm/digital final, to really separate themselves from the pack with the cleanest possible projected image. Spiderman 2 made a quite successful stab at this, but 4K compositing of 35mm 4-perf only provides a marginal improvement over 2K. The StEM illustrates the benefits very well - the 4K 65mm 5-perf original jumps right out against the 35mm (scanned at 6K!). Of course, there aren't enough cameras, but... (Dalsa? Kinetta?)

Tim Sassoon
SFD Vfx & creative post
2525 Main Street
Santa Monica, CA 90405

P&A hidden profit margins won't go away but when the hard costs are significantly lower the risk is lower. Does that mean distributors will stop pocketing every dollar they can get their hands on? Of course not.

>And 65/70mm is far from dead

I didn't say it was dead - just up on the shelf with a lot of other technology that isn't used as often as it once was.

>IMHO it's time for tent pole features that can afford it to really separate >themselves from the pack with the cleanest possible projected image.

I would agree - a fewer choices of donuts on the chow wagon and it's paid for.

Robert Goodman
Author/Filmmaker
Philadelphia, PA

>IMHO it's time for tent pole features that can afford it to consider 65mm >5-perf origination and 4K compositing for 35mm/digital final, to really >separate themselves from the pack with the cleanest possible >projected image. Spiderman 2 made a quite successful stab at this....

Not really. All of the VFX work was done strictly in 2K. Yes, the scans were technically 4K scans, but they were immediately downrezed to 2K for all effects work, only to be uprezzed to 4K again for the DI.

Personally, I agree with John Dykstra's comments regarding the whole "4K pipeline" (really just 2K for at least the 40-60% of the movie that comprised the effects shots). Someone asked why it was done and he replied "because they could." Maybe he was watching Bill Clinton earlier that day, but the comment was very on point.

Mike Most
VFX Supervisor
IATSE Local 600
Los Angeles

Mark Doering-Powell quotes:

>It is the Artist who is truthful, and photography which lies, for in reality >time does not stop." -- Auguste Rodin

Show me one kinetic sculpture by Rodin and I'll gladly agree.

Dan Drasin
Producer/DP
Marin County, CA

>They seem more like the types to shoplift DV tape at Costco.

But in total darkness their single-chip nightshot Sonyâ€™s will shoot circles
around a Viper.

Dan "always use the right tool for the job" Drasin
Producer/DP
Marin County, CA

Steven Poster ASC writes :

>But please note I made no value judgement about the way the movie >was captured. I am just concerned that we may eventually change the >way movies are perceived.

I believe that it does not ultimately matter to the vast majority of the audience and that they will adapt to (and prefer actually) what they have become accustomed to seeing. An example is football, and sports in general. Many decades ago, when much football was seen on film, the audience was quite happy with that. Today, if you showed them film rather than video, most would revolt and complain about how bad it looked...

Tom Tcimpidis

>But in total darkness their single-chip nightshot Sonyâ€™s will shoot circles >around a Viper.

I wonder what information leads you to believe that?

GEORGE C. PALMER

>An example is football, and sports in general. Many decades ago, when >much football was seen on film, the audience was quite happy with >that.

NFL films shoots most of their stuff S-16mm , and they have wide distribution. They shoot most games I think. Their stuff looks GREAT...

Nick Hoffman NYCDP

>However, it still looks like video to me. And I'm worried that we may >never get over that problem. [...]

I remember seeing a demo a few years ago of what is now called SDS-70. 5-perf 70mm projected at 48fps at 35fL. And with all of it's great increase in brightness, sharpness and clarity, what did it feel like --- hmmm --- super detailed video! It was an odd sensation, not what I was expecting at all from 70mm projection.

Jim Houston
Pacific Title Imaging

>...MHO it's time for tent pole features that can afford it to consider 65mm >5-perf origination and 4K compositing for 35mm/digital final, to really >separate themselves from the pack with the cleanest possible >projected image. ...

Alas, there has not really been any excitement in the 5/65 end of Panavision for a while...it seems to me that the cameras go out for VFX background plates once in a while, but that seems to be about it...there are a bunch of Arri 765 pkg's available for rent for about the same cost as a modern 4 perf 35 pkg from them - you only pay a bit more for film stock

The StEM really did show the image clarity advantage that 5-65 origination has...who do we write to in order to start the movement

Mark Weingartner

>An example is football, and sports in general. Many decades ago, when >much football was seen on film, the audience was quite happy with >that.

Ask the average sports fan in this country that question and they would disagree... Studies and surveys have clearly shown such to be the case

Tom Tcimpidis

>5-perf 70mm projected at 48fps at 35fL. And with all of it's great >increase in brightness, sharpness and clarity, what did it feel like --- >hmmm --- super detailed video! ...

As with Showscan (5-perf 70 at 60 fps) and the Luxor Ridefilms (8 perf 35 48 fps) there is a really noticeable difference in the perception of motion between these higher sample rate systems and 24 fps, even when projected with a two blade shutter (same number of flickers per sec as 48fps single blade shutter)

A lot of what we think of as filmic is the particular way that 24fps capture handles motion.

Mark Weingartner
LA based

>[[ "It is the Artist who is truthful, and photography which lies, for in reality >time does not stop." -- Auguste Rodin ]]

> Show me one kinetic sculpture by Rodin and I'll gladly agree.

I'm not so sure that Rodin was referring only to his own work which tends to be static. There was a lot of resistance to early photographers being accepted as Artists - but the idea of "movement" as interpreted by a painter or sculptor is far from what a photographer's up against: somehow capturing "reality" in just the right way. Sure you can have motion blur or a series of photos showing a galloping horse - but in general, you've caught time in your hands.

Its a really great quote when you think of it - and the accusation of photography lying which is normally considered a more scientific and accurate depiction of things. What then do painters and sculptors do ?

Are they not guilty of stopping time merely because they've interpreted it ? Are paintings and sculptures more "real" ?

Oh boy, really far off topic. Listmums are coming... um, well then, what did Rodin think of motion pictures? Lets put a Viper in his hands too.

Mark Doering-Powell
LA based DP

>A lot of what we think of as filmic is the particular way that 24fps capture >handles motion

We equate lower frame rates with the sensation of dream and memory, and higher frame rates with reality. Thus we prefer storytelling at lower frame rates, and Survivor and the evening news at higher. A reason that the flicker of firelight makes us sentimental, ridefilms generally run at 30fps, and Hitchcock step framed Grace Kelly leaning in to kiss James Stewart in Rear Window. In the future, mixed frame rate shows will be common, but will also drive post supervisors nuts.

Tim Sassoon
SFD Vfx & creative post
2525 Main Street
Santa Monica, CA 9040

>In the future, mixed frame rate shows will be common, but will also >drive post supervisors nuts.

Not on a Quantel

Jason Rodriguez
Post Production Artist
Virginia Beach, VA

Tom Wrote :

>An example is football, and sports in general. Many decades ago, when >much football was seen on film, the audience was quite happy with >that.

What studies and what surveys? NFL Films Game of the Week which airs on the NFL Network was the highest rated show in the Networks' inaugural season when it began last October on Direct TV. 44 minutes, 90 percent film, 10 percent video...er, sorry...digital.

HBO "Inside the NFL" which showcases every game, every week has been on the air for several years, and has won several sports Emmy's using S16 capture. Do your research.

Steve "38+ SR's, XTR's, A-Min's, F900's and counting" Richer
Camera Tech
NFL Films

NFL Films :

http://www.nfl.com/tvradio/story/5745241

http://www.kodak.com/country/US/en/motion/newsletters/inCamera/

july99/nfl.shtml

Steve Sabol :

http://www.miscmedia.com/nfl.html

http://www.kodak.com/US/en/motion/sabol/index.shtml

At NAB2004, Kodak showed a demonstration produced by NFL FILMS that compares Super 16 film and digital images displayed on HD monitors. The demonstration included side-by-side comparisons that ran the gamut of production situations captured with film and high-definition, NTSC and PAL format video cameras. The test scenarios included interior and exterior dramas, nature and action sequences.

John Pytlak
Eastman Kodak Company
http://www.kodak.com/go/motion

>However, it still looks like video to me. And I'm worried that we may >never get over that problem. It's what I've been talking about for 15 >years now.

I'm a bit late to the party here, but had to put on my Nomex suit... oh, and see the film...

I'm glad you said that. I thought the film was fun and the choices made were what made the film what it is, can't dispute that, but it really showed where these formats fall apart.

The motion issues were really awful. It was like watching "Cops" at times.

The noise was unappealing at best. To me it didn't create a light night sky effect at all, it didn't feel atmospheric just washed out and noisy. It didn't look like a new way of seeing to me.

The chromatic aberration in out of focus highlights and backgrounds was a bit distracting and the star effect was kind of interesting but not a side effect I'd want - is that a prism issue? It didn't look like an iris shape. The DoF did look very good though. Can I get a Pro HD adapter for my film camera?

As far as the Jazz club scene, it was clean but probably one of the least interesting in look, and it was an easy environment. That doesn't prove the quality of film over HD, shooting film in the cab and night exteriors could have. HD would have looked very good in the Jazz club. I know that they did tests and did what they liked and that's fine.

Again, agreeing with Steven, it looked like video. You can say that video looks more "real" but not to my eye - I don't see color, contrast or motion that way (and not because I watch a lot of movies).

Steven, how would you rate the Genesis footage in regard to the "video look"?

This isn't meant to be a "vs." post, just my observations. People get so caught up in the technology that they forget that people have been shooting gorgeous film with no lights for how many years? Shouldn't sacrifice quality for novelty.

Best Regards,

Anders Uhl
Cinematographer
ICG, New York

>Again, agreeing with Steven, it looked like video. You can say that video >looks more "real" but not to my eye - I don't see color, contrast or motion >that way (and not because I watch a lot of movies).

Couldn't most of the motion problems be alleviated by the use of a 180 or less degree shutter (1/48s or 1/50s)? For me that was the most distracting part between the film versus video, going from one scene with smear everywhere on any motion in the frame, to the crisp judder of the 180 degree shutter on the film camera.

J Rod

>Steven, how would you rate the Genesis footage in regard to the "video >look"?

At the risk of controversy (which never stopped me before) What I saw of the Genesis was not a definitive test. Just a demonstration. But I have to say that I did not see the same "Video" look that I saw in the movie. It was hard for me to tell the difference between the film material and the Genesis footage. But maybe we shouldn't call it footage....

Steven Poster ASC

Steven Poster said....

>However, it still looks like video to me.

And Anders Uhl said...

>But it really showed where these formats fall apart.

Both of you guys - haven't you seen better Viper and HDCAM? Either projected direct, or from a transfer to 35mm?

Why condemn the medium/format - which was admittedly pushed to its limits - when you're really criticizing the choice of the filmmaker?

Dale Launer
writer.filmmaker
Santa Monica

>Both of you guys - haven't you seen better Viper and HDCAM? Either >projected direct, or from a transfer to 35mm?

Of course. But not under those circumstances. I have seen great film under those circumstances however.

Steven Poster ASC

Dale wrote :

>Why condemn the medium/format - which was admittedly pushed to its >limits - when you're really criticizing the choice of the filmmaker?"

Steven wrote :

>Of course. But not under those circumstances. I have seen great film >under those circumstances however.

That was my point as well.

Neither of us was "condemning" the formats.

As I said (and was quoted on): "...it really showed where these formats fall apart". That's what I'm interested in, where it falls apart. What can it do technically that film can't? Some suggest that you can shoot with no light. Clearly that isn't true.

I'm not criticizing the choice of the filmmaker. It's not a choice that I would have made, but it wasn't my film. I enjoyed it for what it was, if different choices had been made it wouldn't be the same film. My comments were technical and aesthetic observations and not a criticism of anyone.

Best Regards,

Anders Uhl
Cinematographer
ICG, New York

> Neither of us was "condemning" the formats.

Sorry - it kinda read that way.

>As I said (and was quoted on) : "...it really showed where these formats >fall apart". That's what I'm interested in, where it falls apart.

Absolutely. Me too. (even though some liked the way that looked, I could think of other ways to push the camera that might look more interesting).

>What can it do technically that film can't? Some suggest that you can >shoot with no light. Clearly that isn't true.

I'm confused then - because you're saying you it can't shoot in dim available light (which is what I'm guess you mean by "no" light) - is it that, or is it the settings and choices you're not happy with?

Cranking up the gain and the detail, and putting green light on your subjects will give you that look, but if you didn't crank up the gain that high, and used either natural light or added full spectrum light, low or no detail - would it be acceptable to you? Or more acceptable?

>I'm not criticizing the choice of the filmmaker. It's not a choice that I >would have made, but it wasn't my film.

Then I'm really confused, what are you criticizing if not their choice?

Don't get angry with me - I'm not challenging you - I'm just missing something here.

Dale Launer
writer.filmmaker
Santa Monica

>I have seen great film under those circumstances however.
>Steven Poster ASC

Yes!!!

Imagine if Michael Chapman had 5218 to shoot *Taxi Driver*...

Or...Maybe even better : 79 push 1...

David Perrault, csc

Despite what has been conjectured here, there was a good deal of Viper footage in all of the club scenes, even cross shots in which film and Viper were cut together.

>As far as the Jazz club scene,...and it was an easy environment. That >doesn't prove the quality of film over HD, shooting film in the cab and >night exteriors could have.

Regarding the video v.s. film "looks", if find it interesting that in all of the discussion of the Genesis/film "comparison", there seems to be an underlying assumption that the DI from the Genesis was an untouched, straight across transfer with no correction, and that the film was transferred straight across to a DI with no correction or preconceived timing reference. I don't know how the latter is technically possible unless it was done with some random "one light" settings which just happened to exactly match the uncorrected Genesis DI.

I'm REALLY not trying to start a fight, but you must all at least entertain the possibility that such a combination of events is, at minimum, a very chance driven probability.

GEORGE C. PALMER
HDPIX, INC.

>there seems to be an underlying assumption that the DI from the >Genesis was an untouched, straight across transfer with no correction, >and that the film was transferred straight across to a DI with no >correction or preconceived timing reference.

Dear George,

Nobody ever said that there wasn't any color correction. Of course color correction was used to make both the film and digital look their best I'm sure. Why do you think that negates this demonstration?

What I did say was that the digital and the film were mostly indistinguishable. That's not what I said about Collateral.

Steven Poster ASC

>Nobody ever said that there wasn't any color correction. Of course color >correction was used to make both the film and digital look their best >I'm sure. Why do you think that negates this demonstration?

A bizarre question, given that the George's beloved Viper would also need to be color corrected.

I think it's fair to compare the best post path possible for any format -- and equally valid to compare lower-end paths that might actually be used.

I mean, there are folks going from film to HDCam to NLE to HDCam to DaVinci to film recorder... which is obviously inferior to SR or even D5 or disk, but it's done because that's what's available to them.

Jeff "big picture" Kreines

>Nobody ever said that there wasn't any color correction. Of course color >correction was used to make both the film and digital look their best >I'm sure.

Steven :

I'm sorry, I did not mean that you personally or singularly had created that assumption, and of course I know that color correction is necessary for any negative. But there are many folks who have been characterizing the Genesis/ film comparison as some kind of "pure", film/video comparison with minimal touch-ups.

The difference is that most of the other tests were self characterized as demonstrating the potential, with appropriate post production correction/processing of HD cameras, of achieving film-like performance in those cameras. In fact, it was not any lack of potential interoperability, the sensor size, focal plane characteristics, or the in-camera processing design of the Viper on "Collateral, but shutter angle choices, which tipped some off to a perceived video look.

So I agree that, given those limitations and the presence of differing post processing for each original media, it is extremely possible to achieve ALMOST seamless interoperability of HD and film, IF all "video look" parameters are matched (i.e. shutter angle).

My only objection has been to the notion that, in the Genesis test we saw the results of a comparison of two virtually equally treated raw originating media, rendering actual objective comparison data. That is just not possible with the Genesis or any other in-camera processed camera, UNLESS the HD camera setup is created to produce a pre-corrected DI which matches the desired look of the corrected film negative OR unless the HD in-camera setup is further post corrected to match the DI created from the timed/corrected film. The former was pretty much the case in "Collateral". So, of course, both scenarios can work, and neither negates the test, only some of the implied conclusions.

Now if the Genesis outputted unprocessed 4:4:4 FilmStream data, as the Viper can, exactly analogous to the data created from the transfer of film negative to a DI, so that both could have the same post production timing/corrections applied to both, THAT would be a really interesting test that could provide truly objective results for hard comparisons.

GEORGE C. PALMER
HDPIX, INC.

>THAT would be a really interesting test that could provide truly objective >results for hard comparisons.

That would be a useless and unnecessary test with no real world conclusion possible.

Steven Poster ASC

George C. Palmer wrote :

>...it was not any lack of potential interoperability, the sensor size... or the >in-camera processing design of the Viper on "Collateral, but shutter >angle choices, which tipped some off to a perceived video look.

1/24th shutter was part of what tipped me off, and what may have bugged some others the most. Its tough to resist when it lets the chip gather another stop of available light - while depleting motion sharpness. I read this thread with interest - and I liked the movie - and at the very least the challenges they embraced presented risks and took a lot of expertise to present as good as it was.

but the other "inconsistencies" people have mentioned were the weak blacks (sans detail) and at times very weak colors. The extreme amount of gain in some shots didn't help - the noise sizzled in the weak blacks of some underlit night/EXTâ€™s. And if Dale's right that they intended it to look that way, well then great. I am talking how the results are perceived, not privy to nor care as much about the intentions.

With all good intentions gone bad, by trying to clarify this extensive topic it might seem that we're bashing the film, but really we're not - or at least I'm not. Parts of it looked really great and appro the story (whether film, Viper or f900/HDcam). Just pointing out what some have seen as inconsistencies in the look - which others actually embraced as the aesthetic ! and again, the film was very well done overall - I really liked Foxx.

Mark Doering-Powell
LA based DP

George C. Palmer wrote:

>Despite what has been conjectured here, there was a good deal of >Viper footage in all of the club scenes, even cross shots in which film >and Viper were cut together.

OK, so my point was very accurate. It was an easy environment, one that would accommodate a digital format very well.

>Regarding the video vs. film "looks", if find it interesting that in all of the >discussion of the Genesis/film "comparison",

Just for the record, since I was quoted, I didn't say anything about the Genesis/Film comparison and didn't see it. However, it's crazy to assume that you can/should follow the same finishing path with the different formats - that wouldn't give you a more accurate comparison but rather a less accurate one. The format dictates the path and each has it's optimum and even correct post process. For that matter, if film grain is part of the aesthetic for better or worse, leave it alone. Trying to make them look the same isn't exactly a comparison of the qualities of either format.

Best Regards,

Anders Uhl
Cinematographer
ICG, New York

Come to think of it, I wonder what was the reasoning in the decision to shoot the jazz club on film...(?)

John Babl
DP
Miami

Jeff Kreines writes:

>I think it's fair to compare the best post path possible for any format

Hear, hear.

I think it's also fair to read American Cinematographer. The recent article on Collateral explains their entire rationale for their choice of media.

(My apologies if someone HAS mentioned this article -- I just don't recall seeing it here.)

Dan Drasin
Producer/DP
Marin County, CA

John Babl writes:

>I wonder what was the reasoning in the decision to shoot the jazz club >on film...

See page 40, et. seq, of the August, 2004 American Cinematographer.

Dan "it's a grand old rag" Drasin
Producer/DP
Marin County, CA

Anders :

I see your point, but the club was anything but easy, as the highest light levels on the set were not on the floor of the club, but on and in the set pieces that were in the background. The most light I ever saw on the floor (the principals and the extras was, maybe, 4 fc. Not an easy environment without pushing ASA either film or video.

The hard part was rendering dark complexion with both appropriate and accurate luminance and chrominant characteristics, then being able to cut together cross shots of actors talking to each other without them seeming like they were shot in two different environments. Easy if you are shooting with only film or only HD, but to make them cut is the more difficult part.

I didn't mean to mix apples and oranges with your comments, but some other folks on CML are quite willing to talk and conjecture about the merits of the equality and interoperability of film and electronic capture HD with a camera like the Genesis but seem to have difficulty admitting to or acquiescing that is possible with the Viper, even when it has actually been done (perhaps not to their taste, and not, perhaps without some tips to the HD footage like 270 deg. shutter). So, sorry, I didn't mean to suggest that you had seen the test or were referencing it, and I agree that when PROCESSED HD video and film are the media involved, the two most probably cannot follow the same finishing path, simply because of the many variables (limitations) added by the in-camera processing, but, when one follows the UNPROCESSED FilmStream path together with film, you have the possibility for a very accurate and analogous result if you follow the same finishing process for both, because the FilmStream path yields a directly analogous (to film) electronic camera negative.

At that point, the finishing process can be applied almost exactly, equally, and any processing and grading that is applied to one (with minor variations to account for the limits of sensor dynamics) can be applied to the other, especially if the same nominal ASA (ISO or whatever other rating system you choose) is chosen for both media is the same. And, by the way, that is usually, nominally, between 375 and 500 (based on the quantum characteristics of all CCD class sensors no matter what the size) for any CCD sensor (without adding gain our highlight and amounts of knee that unduly alter the lower gray scale, highlights or flesh tones), since the film will, of course have the latitude to exceed those numbers in either direction.

So I would suggest that it is not the format (film or electronic HD) that dictates the path (grain and noise considerations aside), but the ability to choose whether to use OR not use in-camera processing that determines the efficacy of parallel finishing processes. The choice of format size, in the case of the Genesis, is a creative one, since the latitude of all CCD's, regardless of the array size, is a matter of the limitation of pixel level physics. Larger pixels may, arguably, yield more dynamic latitude, but not more pixels. More pixels may however, if the color filter chosen is perfect and creates no chrominance aliasing, and if the recorder is totally uncompressed and employs no resolution limiting, increase resolution but it does not increase latitude. Especially if the CCD is an IT/FIT sensor, as it probably is in the Genesis, because the IT/FIT sensor actually has smaller pixels (by at least 40%) than FT sensors (as in the Viper).

At the end of the day, the choice of a larger format sensor is based much more on the ergonomic comfort of shooting with PL mount lenses than the technical merits of a larger sensor anyway.

GEORGE C. PALMER
HDPIX, INC.

George C. Palmer wrote:

Hi George,

Thanks for the lengthy reply.

>I see your point, but the club was anything but easy, as the highest light >levels on the set were not on the floor of the club,

Well, it was a controlled environment and certainly should have been...incidentally I was talking about the Jazz club and not the dance club.

The dance club, no, not as easy. I was referring to the Jazz club because people had commented on how good it looked and that it was shot only on film. The dance club didn't look so smooth.

>I didn't mean to mix apples and oranges with your comments,

Quite alright, I just wanted to clarify.

Best Regards,

Anders Uhl
Cinematographer
ICG, New York