Test results...

[robinbuday]

Did we ever find out what happened with those tests Kodak conducted? Were they ever made available?

Sorry if this has been answered already.

[Rocco Schult]

If nobody else is - I am interested too, ambush or not.

[Greg Voevodsky]

Kodak vs RED tests revealed... soon...
you might have to join digital cinema society though if you want to go.
Notice - RED and Kodak are invited to the results:

Info here:

http://www.digitalcinemasociety.org/

I thought this was the "ambush" test... sorry, if this is wrong... but no one was giving much details as to what happened and where.

[David Mullen ASC]

The Kodak tested involved other digital cameras than the RED and will probably take months to do the post on -- Kodak generally doesn't rush any public demos and tests. And it was for internal purposes only, then it doesn't matter how fast they work on them, we won't see them.

The DCS demo is probably more informal, of the "let's see how the camera looks" variety.

[Michael Schrengohst]

David has probably shot more Kodak film then most
posters here. I shot a fair amount myself but we are
not in LA or NYC and usually don't get the high end spot/
TV work here. Kodak probably did the same thing when
the Digital SLR's hit the market. That is why they probably
went head long into setting up digital printing stations.
I also shot a fair amount of kodachrome 25 and nothing
beats that image. I like digital OK but it does not have the
same feeling.....but if I where to show up on a photo
assignment now with Kodachome 25 my clients would be saying "WTF"??
Same now in my digital world. The Kodak tests will only confirm to
Kodak that film will eventually be replaced and they need to figure
out how to keep a foot in the market. Same thing with the film labs
and film to tape transfer. It will all still be around for a while but
at what degree??? I am sure Kodak will spin some marketing
around these tests and proclaim that film is the "superior capture
medium" - so - my Nikon F2 and Kodachrome was superior at one time....
It's just a matter of time - how much is anyones guess.

[Tom Lowe]

David, I did not follow this whole saga closely. What do you think of Kodak testing the RED camera without even speaking to them first? Is this normal for Kodak?

Because the camera is so new, basically still in a beta phase, I can see why Jannard was perturbed.

[Michael Schrengohst]

What difference are Kodak tests going to make??
Will it make someone say Hmmmm....
should I buy a RED and capture images that
rival or go beyond 35MM without the additional
cost? Or should I get that used Arri on e-bay??

[Tom Lowe]

In a sense it is funny. This would be like Pepsi hosting a "fair and balanced" soda competition.

[Michael Schrengohst]

I have done a bunch of work for Frito-Lay, they have
consumer taste tests and crap like that I have taped.
It is just - well who are we kidding, Frito-Lay crushes the
competition anyway they can. They have such power they
kick the competition all the way to the middle of an aisle
way up high. Kodak can rant & rave about any thing they
find out with their "test". Once the "films" shot on RED start
hitting the screens Kodak can say this "film" shot on RED
was transferred to Kodak film.

[Rick Darge]

That's not going to happen because RED is going to kill them once again with their low cost 4k projectors. Goodbye Sony too.

[Jay A. Kelley]

Last time I looked, I noticed Sony made more than a couple items. But I could me wrong!

:)

[Steven Parker]

The DCS demo is probably more informal, of the "let's see how the camera looks" variety.

Davis is exactly right. Jim Mathers' tests are not the "ambush" tests apparently sponsored by Kodak. Jim put #30 thru its paces and is doing a 4K finish and film-out to compare RED 4K footage screened with a 4K projector vs RED 4K footage screened on a 35mm print. The test is for his own information, and for the benefit of DCS members.

http://digitalcinemasociety.org/

[Jonathan L. Bowen]

I think basically Kodak is the anti-christ. I can't think of many more companies I hate more than Kodak besides Micro$oft and maybe B&H Photo Video in New York, for reasons that cost me money in the last week, maybe I forgive them eventually. And I still like Bill Gates and really respect him, I just hate Microsoft products and think they have no clue what they're doing with, well, anything. Everything they make is a total disaster (the Zune? Crap; Xbox 360? Doesn't work, fails immediately, etc., had it happen personally; Windows anything? Sucks hardcore, etc.).

With Kodak, I saw this comparison test in film school of their film stock with various digital cameras, and it was presented to us like a "look and comparison of digital and film." It wasn't a comparison at all. It was completely propaganda from MORONS. Only the stupidest DP in the world would say something as inane as, "Oh film will ALWAYS be better than digital, you cannot beat the film look." Yeah you can. Film is just an ancient recording medium on its last legs and digital can mimic it in EVERY respect WHATSOEVER and surpass it in cost and efficiency. There isn't anything about Kodak film that is "special," unless by special you mean like riding the short bus to school. Those bunch of tools need to go sell tomatoes or open an office supply store, because film is dead as a doornail. Anyone who says otherwise is in total denial.

[Michael Brennan]

.....With Kodak, I saw this comparison test in film school of their film stock with various digital cameras, and it was presented to us like a "look and comparison of digital and film." It wasn't a comparison at all. It was completely propaganda from MORONS....

Reminds me of another companies chat to students.
A few years ago Arri conducted a seminar in Denmark to the National Film School where they screened their own f900 vs 35mm test.

The students (used to watching locally shot movies digibeta to 35mm) were so surprised at the results they went and shot their own tests.

The conclusion of at least a couple of the students was that the transfer to film that Arri had done (in house) was well below industry standard and that the presenters who were senior Arri exectutives were both biased and ignorant.

I loved the quote to a young filmmaker from a senior Kodak rep at a BFI event in 2002; "you have to go to the USA to get HD transferred to film"

Young film maker had just shot filmout tests in Denmark and UK the week before.


Mike Brennan

[Stephen Williams]

It was completely propaganda from MORONS. Only the stupidest DP in the world would say something as inane as, "Oh film will ALWAYS be better than digital, you cannot beat the film look."
Yeah you can. Film is just an ancient recording medium on its last legs and digital can mimic it in EVERY respect WHATSOEVER and surpass it in cost and efficiency.

Hi Jonathan,

Very interesting comment, I look forward to seeing the work you shoot with your camera when it's delivered.

Stephen

[M Most]

I think basically Kodak is the anti-christ. I can't think of many more companies I hate more than Kodak besides Micro$oft and maybe B&H Photo Video in New York....I just hate Microsoft products and think they have no clue what they're doing with, well, anything. ...Film is just an ancient recording medium on its last legs....Anyone who says otherwise is in total denial.

There you go, showing that maturity again. Thanks for your insightful comments.

[Jay A. Kelley]

Only the stupidest DP in the world would say something as inane as, "Oh film will ALWAYS be better than digital, you cannot beat the film look." Yeah you can. Film is just an ancient recording medium on its last legs and digital can mimic it in EVERY respect WHATSOEVER and surpass it in cost and efficiency. There isn't anything about Kodak film that is "special," unless by special you mean like riding the short bus to school. Those bunch of tools need to go sell tomatoes or open an office supply store, because film is dead as a doornail. Anyone who says otherwise is in total denial.

Wow... Those are some big statements there buddy.. You must be up there with Spielberg and the other heavy hitters to speak with such confidence experience.

I myself am not near as experienced as you. I've only been doing this for 20 years and I have not shot every type of film, or used every model of high-end camera, therefore I can't speak on this with the same authority you can.

Send me a link of some of your work dude!! It must rock and I would LOVE to see it.

Thanks! Have a great weekend!

Jay

:)

[jaadgy akanni]

oops, sorry

[jaadgy akanni]

I for one don't give a rat's stinky ass. I began my sound designing career using Protools since the early 90's and have never had to go through the time- consuming, awkward, and torturous nightmare of working with analog tape, thank god.
Now that I'm ready to make my entrance into Cinema I feel lucky that I will never have to work with film because, especially now that there's RED. RED has it's own aesthetics and I'm in love with it. Now, those who every now and then say "film is better than digital-you will never beat the look of film" are making a subjective statement, and that's cool. But it also holds true that Film will never beat the look of RED-and I prefer RED. That's me being subjective, like all true artists are.
Digital will prevail in the end. Why? because we humans are evermore in need of the easy way to do things. We need to do away with the cumbersome, the time-consuming, those things that are not so great that they're worth the hell they put you through.

[Mike Smith]

Moderators: Any chance we can have a kiddies section added to this forum? ;-)

[chuck colburn]

Moderators: Any chance we can have a kiddies section added to this forum? ;-)

I don't know Vmax..... Mayby he's right. Looks like I got out of the motion picture (film) camera business JUST in time! Although that last camera I assembled (65mm shooting at a standard 60 f.p.s.) did look kinda nice on the big screen. Who knows, mayby Doug Trumbull was just an idiot for even coming up with the concept. Oh well guess I'll just go catch me a salmon for lunch.

[Mike Smith]

Ha Ha!

[Chris Pickle]

. . . unless by special you mean like riding the short bus to school.. . .

Disgusting comment not warranted by professionals.

[Chris Kenny]

The tone of some of the anti-film folks here might be a bit out of line, but it's hard to reasonably argue with a lot of what they're saying.

If one looks at how far digital has come in the last ten years, it's nearly impossible to avoid the conclusion that the digital acquisition systems available ten years from now will vastly exceed the quality that film can deliver on every technical characteristic -- and provide enough resolution and dynamic range that any desired aesthetic characteristic of film can be emulated in post. Film will continue to improve as well, of course, but, as has been the case over the last 10 years, at a much slower pace than digital.

If one brings discussion of workflow issues into the picture, the conclusion that digital will largely replace film becomes even more inescapable. You'll probably be able to set up a 4K uncompressed workflow on a laptop in 10 years. Without external storage. No joke. Just project out what has happened over the last ten years in the storage and graphics processing markets.

It's even likely that 4K displays will be widely available in 10 years -- probably not because video content will be regularly delivered at 4K, but simply because there's going to be a big push in the next few years to move computer displays from ~100 ppi to ~200 ppi (mostly to get better text rendering). A 24" 16:9 display at 200 ppi is about 4.2K.

[Mike Smith]

The tone of some of the anti-film folks here might be a bit out of line, but it's hard to reasonably argue with a lot of what they're saying.

If one looks at how far digital has come in the last ten years, it's nearly impossible to avoid the conclusion that the digital acquisition systems available ten years from now will vastly exceed the quality that film can deliver on every technical characteristic -- and provide enough resolution and dynamic range that any desired aesthetic characteristic of film can be emulated in post. Film will continue to improve as well, of course, but, as has been the case over the last 10 years, at a much slower pace than digital.

If one brings discussion of workflow issues into the picture, the conclusion that digital will largely replace film becomes even more inescapable. You'll probably be able to set up a 4K uncompressed workflow on a laptop in 10 years. Without external storage. No joke. Just project out what has happened over the last ten years in the storage and graphics processing markets.

It's even likely that 4K displays will be widely available in 10 years -- probably not because video content will be regularly delivered at 4K, but simply because there's going to be a big push in the next few years to move computer displays from ~100 ppi to ~200 ppi (mostly to get better text rendering). A 24" 16:9 display at 200 ppi is about 4.2K.
All very good points Chris. It is these sorts of intelligent discussions - pro and con - that I visit this forum for, not immature, hateful and ill-informed comments that clutter up the thread.

[Greg Voevodsky]

I don't know Vmax..... Mayby he's right. Looks like I got out of the motion picture (film) camera business JUST in time! Although that last camera I assembled (65mm shooting at a standard 60 f.p.s.) did look kinda nice on the big screen. Who knows, mayby Doug Trumbull was just an idiot for even coming up with the concept. Oh well guess I'll just go catch me a salmon for lunch.

The problem with Showscan is or was the same as 1080p 60fps without changing the shutter now - the pictures were artificially sharp without motion blur (they called it hyper reality) The demo I saw, yes in film school, ART CENTER, in the mid 90s - I was not amazed at all... the rollercoaster had the front wooden rungs all sharp - just take your video camera and crank the shutter up 200-500 and you will have the same unnatural ultra sharp look.

Sorry, but I can see why I-Max won. However, with RED, i-max shooting might really be in trouble. Wait till their is a digital IMAX projector and RED will be there!

[tj williams]

Chris... Whether population growth, or butter consumption the rate of change today will alter in the future and most of it will take the shape of a bell curve. Prediction that takes the form of prognostication thru straight line projection, never gets good results. Today it's true that digital is progressing faster than chemical. Does this show that it will always be so, no it doesn't

There will always though in my opinion be folks on the internet like JRB true believers no matter what they believe are almost always offensive. This may be the one assurred truth in a quickly changing world

[Chris Kenny]

Chris... Whether population growth, or butter consumption the rate of change today will alter in the future and most of it will take the shape of a bell curve. Prediction that takes the form of prognostication thru straight line projection, never gets good results. Today it's true that digital is progressing faster than chemical. Does this show that it will always be so, no it doesn't

In the digital world, straight-line projection (or, rather, exponential projection, which gives you a straight line on a graph with an exponential scale) provides extremely good results over sustained time periods in many cases. For instance, hard drive capacity:

http://upload.wikimedia.org/wikipedia/commons/a/a1/Hard_drive_capacity_over_time.png

Pixels per dollar in digital cameras (based on retail prices of Kodak digicams):

http://upload.wikimedia.org/wikipedia/commons/9/94/Hendys_Law.jpg

And everyone's favorite, computational speed:

http://upload.wikimedia.org/wikipedia/commons/thumb/c/c5/PPTMooresLawai.jpg/596px-PPTMooresLawai.jpg

That last one is particularly interesting, because it extends across five completely different processing mechanisms (and a full 100 years). It also shows double exponential growth. In other words, not only does computational capacity per dollar double on a predictable schedule, but the rate of doubling itself accelerates predictably.

Perhaps there will be some radical unprecedented breakthrough in chemical engineering that will throw the lead to film. Or perhaps, after 100 years, information processing advances will suddenly hit a wall.

But it's far more likely that film, like so many other alternatives to digital technology, will be crushed by the inexorable advance of Moore's Law.

[Stephen Williams]

But it's far more likely that film, like so many other alternatives to digital technology, will be crushed by the inexorable advance of Moore's Law.

Hi Chris,

I am staggered that has not over the last 50+ years where both technologies existed. Interestingly film has also improved over this time, the goal posts have always been moving.

What I find curious, a client who I have shot exclusively with the Viper over the last year wants our next project to be on film. It's actually a very small budget, we cant possibly afford a Viper for 6 - 8 days required.

Stephen

[Billy Summers]

Neither film or digital has ever equaled the massive quality of endless resolution that resides in my own mind.

That's why, when clients hire me, they know they're not just getting the best, they're getting 'better than the best...

[Chris Kenny]

Interestingly film has also improved over this time, the goal posts have always been moving.


Computational capacity per dollar is somewhere between one and ten billion times what it was in 1950.

What characteristic of film has improved by even 0.1% of lower of those two numbers, i.e. a mere 1,000,000x improvement? Or is film, perhaps, a million times cheaper? I think you'll find that if you graphed the improvement of any characteristic of film on those same exponential scales, the line would be practically flat.

Consider also that digital got good enough to replace SD analog electronic recording only about 12 years ago... and today there's already serious debate about it rivaling 35mm film (some people even say 65mm film) for quality. Unless the best film available in 1995 was an 8mm stock with 6 or 7 stops of dynamic range (which somehow I don't recall being the case), film clearly hasn't come nearly as far as digital over that time period.



What I find curious, a client who I have shot exclusively with the Viper over the last year wants our next project to be on film. It's actually a very small budget, we cant possibly afford a Viper for 6 - 8 days required.


Sure. Right now, digital is reaching the point where it's competitive with film, and cheaper for some projects, but not the clear all-around winner in every case. In five or ten years, the situation will be very different.

[Stephen Williams]

. and today there's already serious debate about it rivaling 35mm film (some people even say 65mm film)

Hi Chris,

There was a serious debate about 10 years ago from a Film Director who said HDCAM was equal to 65mm, another wealthy film director said it was better than 35mm.

I would be interested to see Red Back to Back with 65mm films from 50 Years ago, hopefully I will soon get the chance. I wish to see some 65mm films displayed in 4K and also Red as a film out V 65mm then I will have an idea where we are today digitally speaking.

Stephen

[Chris Kenny]

There was a serious debate about 10 years ago from a Film Director who said HDCAM was equal to 65mm, another wealthy film director said it was better than 35mm.


This actually does raise the interesting question of why high-end digital acquisition hasn't been advancing faster than it has. I mean, HDCAM to HDCAM SR, at about the same price point, isn't all that big a jump by the standards of the information technology world.

I think Red answers that question very nicely. The reason we weren't seeing the sort of development in the digital cinematography market that is to be expected from an information technology market is because the market segmentation practices of major vendors were obscuring technological advances.

In other words, the first HDCAM cameras and decks were probably priced pretty reasonably -- with the technology of 1997, it really cost outrageous amounts of money to do what they did. Sony just didn't bother to lower prices as the underlying technology got vastly cheaper!

Now we've got Red, which delivers significantly better quality than the first HDCAM systems, in a much smaller package, for a tenth of the price. That's more like what the graphs I posted above tell us we should be seeing.



I would be interested to see Red Back to Back with 65mm films from 50 Years ago, hopefully I will soon get the chance. I wish to see some 65mm films displayed in 4K and also Red as a film out V 65mm then I will have an idea where we are today digitally speaking.


While this will definitely be interesting, it's not so important in terms of the larger discussion here. Because if digital isn't there yet, it will be in five years.

[Stephen Williams]

Hi Chris,

Sony probably did not reduce the price that much, as very few units have been sold, so I don't think Digital cinematography has made them much money.

Well I think the Red camera in 5 years time will be improved over the Red one of today, many people will still make a film out for archive purposes, I know you will say that's lossy but it has worked as the best long term storage to date.

Stephen


This actually does raise the interesting question of why high-end digital acquisition hasn't been advancing faster than it has. I mean, HDCAM to HDCAM SR, at about the same price point, isn't all that big a jump by the standards of the information technology world.

I think Red answers that question very nicely. The reason we weren't seeing the sort of development in the digital cinematography market that is to be expected from an information technology market is because the market segmentation practices of major vendors were obscuring technological advances.

In other words, the first HDCAM cameras and decks were probably priced pretty reasonably -- with the technology of 1997, it really cost outrageous amounts of money to do what they did. Sony just didn't bother to lower prices as the underlying technology got vastly cheaper!

Now we've got Red, which delivers significantly better quality than the first HDCAM systems, in a much smaller package, for a tenth of the price. That's more like what the graphs I posted above tell us we should be seeing.



While this will definitely be interesting, it's not so important in terms of the larger discussion here. Because if digital isn't there yet, it will be in five years.

[Chris Kenny]

Hi Chris,
Sony probably did not reduce the price that much, as very few units have been sold, so I don't think Digital cinematography has made them much money.


Sure. Sony seems to believe that the digital cinema camera market doesn't exhibit much price elasticity. In other words, that demand is relatively constant regardless of price. I think they're very wrong about this, but if this is what you believe, high prices make perfect sense.

Sony also has to worry about all its mid-range products. They can't lower prices on their high-end digital cinema cameras that much unless they also push down prices on their mid-range products, which would depress prices on their prosumer products. In other words, Sony cutting the price of the F23 to, say, $30K would ripple down their entire video product line.



Well I think the Red camera in 5 years time will be improved over the Red one of today, many people will still make a film out for archive purposes, I know you will say that's lossy but it has worked as the best long term storage to date.


A better approach might be to print high-resolution 2D barcodes to black and white film, storing the original digital data.

[Gavin Greenwalt]

A better approach might be to print high-resolution 2D barcodes to black and white film, storing the original digital data.

... brilliant! Let's take it a step further. Not just barcodes but 16 color barcodes at 1k resolution printed at 4k.

Edit: Sorry I keep changing my post as I think about the math involved. 16 color bar codes would be 4 bit. So 3 color pixels would represent one RAW pixel. Printing 1k of data (at 4k resolution) that's .25k of data per frame. That means your film would be 256 times longer than a standard feature. Not as clever as I initially thought unless you could extract 4k worth of perfect data from a film print. Then it would be only 8x as long.

Maybe if you printed it twice. Then did an optical comparison to enhance the useability allow you to print at 4k in which case it would only require 4x as much footage as a feature.

Of course that's just RAW you would probably want an RGB since you've already graded and what not and don't want to take another RAW resolution hit. So make that 12-24x the reels you would need to store.

To the question of "how has film improved".

By several stops of lattitude. Better color reproduction. Far more than .01%. I would say more like 20%. For one only students now shoot on reversal stocks unless a specific effect is desired. And that's just from the 60s. Look at 8mm and super 8mm films and compare them to a modern 8mm film shot on a Kodak Vision stock.

[Stephen Williams]

.

By several stops of lattitude. Better color reproduction. Far more than .01%. I would say more like 20%. For one only students now shoot on reversal stocks unless a specific effect is desired. And that's just from the 60s. Look at 8mm and super 8mm films and compare them to a modern 8mm film shot on a Kodak Vision stock.

Hi,

I don't think Chris wanted you to answer that question, its far easier to believe that film died in 1960 as predicted 5 years earlier.

Stephen

[ColinSmith]

The problem with Showscan is or was the same as 1080p 60fps without changing the shutter now - the pictures were artificially sharp without motion blur (they called it hyper reality) The demo I saw, yes in film school, ART CENTER, in the mid 90s - I was not amazed at all... the rollercoaster had the front wooden rungs all sharp - just take your video camera and crank the shutter up 200-500 and you will have the same unnatural ultra sharp look.


I'm not going to argue about the aesthetics of fast shutters and deep DoF, but you have to remember that those are stylistic choices, part of the language of story telling on film, and "reality", at least in terms of your perception, if not in terms of the optics of your eye, is that everything IS in focus all the time ;-)

[Gavin Greenwalt]

Hi,

I don't think Chris wanted you to answer that question, its far easier to believe that film died in 1960 as predicted 5 years earlier.

Stephen

In all fairness 1960 35mm film quality has died and been surpassed by high end digital cinema cameras. :D

[Chris Kenny]

... brilliant! Let's take it a step further. Not just barcodes but 16 color barcodes at 1k resolution printed at 4k.

Edit: Sorry I keep changing my post as I think about the math involved. 16 color bar codes would be 4 bit. So 3 color pixels would represent one RAW pixel. Printing 1k of data (at 4k resolution) that's .25k of data per frame. That means your film would be 256 times longer than a standard feature. Not as clever as I initially thought unless you could extract 4k worth of perfect data from a film print. Then it would be only 8x as long.


Full aperture 4K is what? 4096x3112 or thereabouts (I find a couple of different numbers in a quick search). If you're just using black and white dots (and B&W film stocks are more stable and higher resolution, so it's probably a good idea), that's 12746752 bits, or about 1.5 MB/frame.

This is a little impractical if you're storing uncompressed data, but not so bad for compressed. REDCODE RAW runs about 27.5 MB/sec at 24 frames/sec, so you could actually store a frame of REDOCE RAW on less than one frame of film.

Even if you assume you can only actually store 2K full aperture recoverably in each frame, you only need four times as much film. Which is about the same amount you'd need to make a traditional RGB separation. And even if you double or tripple the data rate (since compressed RGB will take up more space than), it's still pretty reasonable.

Distribute parity information throughout the print (makes it a bit longer longer, of course), and you can recover all the data even if a substantial number of the bits are unreadable or totally gone.

Some people might scream about storing compressed data, but I bet you lose less information than you do turning your digital data into analog images.

If you're worried people won't be able to figure out the compression format in 50 years, just document it and put the documentation at the beginning of every reel (optically, so it's directly human-readable when magnified). Do it in a half dozen languages if you like.



To the question of "how has film improved".

By several stops of lattitude. Better color reproduction. Far more than .01%. I would say more like 20%.

Sure. I didn't say film hadn't improved more than .01% over the last 50 years. I said it hadn't improved more than .01% as much as information processing technology has. Information processing speed at a given dollar cost has improved upwards of a billion times since 1950, or 100,000,000,000%. (Yes, I know exactly how crazy that number looks, but it happens to be supported by the data.) Since you say film has improved about 20% (I'd actually give it a little more credit than that), I think you agree with me. :whistling:

(And yes, advances in computing don't necessarily generate directly into higher image quality for digital acquisition... but the graph I posted above shows exponential growth for sensor resolution as well. CMOS sensors are fabricated using largely the same technologies as microprocessors, after all.)

[Chris Kenny]

Hi,
I don't think Chris wanted you to answer that question


Actually, I don't mind at all. I'm curious what your figure would be for how much film has improved since 1950 and since ~12 years ago (the introduction of mass-market digital systems)

[Stephen Williams]

Actually, I don't mind at all. I'm curious what your figure would be for how much film has improved since 1950 and since ~12 years ago (the introduction of mass-market digital systems)

Hi Chris,

I think going from the extended range stocks - vision- vision 2 probably 3 extra usable stops. I have only been shooting Motion Picture Negative since 1979, one generally took more care as fixing errors in post was not an option.

In terms of speed Kodak only made 1 Color film stock 100asa 30 years ago, the previous emulsion was 50 asa, so there have been huge improvements in speed, grain reduction & latitude.

Sony set out in 1982 to develop a HD system to rival reversal film used in television. When they introduced the F900 it probably did outperform 7240 VNF, so they achieved their original goal.

Stephen

[Chris Kenny]

I think going from the extended range stocks - vision- vision 2 probably 3 extra usable stops. I have only been shooting Motion Picture Negative since 1979, one generally took more care as fixing errors in post was not an option.

In terms of speed Kodak only made 1 Color film stock 100asa 30 years ago, the previous emulsion was 50 asa, so there have been huge improvements in speed, grain reduction & latitude.

Sony set out in 1982 to develop a HD system to rival reversal film used in television. When they introduced the F900 it probably did outperform 7240 VNF, so they achieved their original goal.


OK, so let's say film today is five times as fast at a given level of grain vs. 1950, and has three more stops. Is that reasonable?

Would you say resolution has increased at all?

[Steve Sherrick]

I've expressed this in other threads and based on what I just read, still feel that people are kind of missing the point here. There is the temptation to want to say one is better than the other, and it is impossible to quantify because they are not the same medium. They capture images completely different. The results are subjective. Quite narrow thinking to want to say there is only one way to shoot moving forward because it is superior to everything else. I love film, and I love what I am seeing from Red. I might use either one depending on what the project calls for in terms of story, tone, budget, etc. Paint comes in acrylic, watercolor, oil, etc. What gives you the look you are after?

Jonathan makes outlandish statements all of the time. I don't mind them because quite honestly they are sometimes entertaining. He stepped out of line with the short bus comment. Easy to get offended by that. He can say all he wants about technology and that tripods should only cost $50, but that one statement is a mistake and he should retract it.

Seriously guys, film vs digital is kind of like Mac vs PC. It really does come down to what gets the job done. As long as they are both coexisting, why try to bannish it. No one is forcing you to shoot film or Red. They are choices, artistic choices.

Anyway, don't mean to hijack the thread, just had to throw my two cents in, especially about Jonathan's offending statement.

Steve

[Chris Kenny]

I've expressed this in other threads and based on what I just read, still feel that people are kind of missing the point here. There is the temptation to want to say one is better than the other, and it is impossible to quantify because they are not the same medium. They capture images completely different. The results are subjective. Quite narrow thinking to want to say there is only one way to shoot moving forward because it is superior to everything else. I love film, and I love what I am seeing from Red. I might use either one depending on what the project calls for in terms of story, tone, budget, etc. Paint comes in acrylic, watercolor, oil, etc. What gives you the look you are after?

Right now digital and film "looks" are complementary, as you say. If one is interested in the future of motion picture acquisition, though, it's a different story. In ten years the latter will be a fully enclosed subset of the former. In other words, digital systems will be able to reproduce any look you can produce on film, and a lot of others besides.

There's simply nothing that prevents a sufficiently advanced digital system from doing this, and the workflow and cost benefits of digital will be such that a lot of resources will be invested in making it happen.

[Steve Sherrick]

In other words, digital systems will be able to reproduce any look you can produce on film

Perhaps, although in the purist sense, one could argue there is no way to represent film because of the difference in aquisition processes. One chemical, the other 1s and 0s. However, I am very much in agreement that Digital will be the predominent format at some point in the future.

And there will be major advances in picture quality, both in terms of resolution and manipulation. As you suggest, there will be more advanced tools to help you achieve almost any look imaginable.

In the meantime, I still see film as being relevant as a medium.

Steve

[Chris Kenny]

Perhaps, although in the purist sense, one could argue there is no way to represent film because of the difference in aquisition processes. One chemical, the other 1s and 0s.


Hypothetically, if your sensor created a sufficiently accurate record of the photons that hit it, you could write software that would produce film looks by literally emulating the chemistry of film at the molecular level. Mind you, this is a sort of thought experiment to demonstrate the theoretical capabilities of digital processing. The computational capacity to do this practically is probably decades away. Fortunately, when employing film looks for creative purposes one only has to create simulations that are convincing to unaided human vision, which is a much easier problem.



In the meantime, I still see film as being relevant as a medium.


Sure. It will be relevant to people who can afford it.

And it will be relevant to people who are comfortable with it. Frankly, it's not particularly relevant to me. I don't "hate film" or anything. It's just that digital cameras came along early enough that I don't think I've shot more than a handful of frames of film in my adult life, compared with probably tens of thousands of digital images, and dozens or hundreds of hours of digital video. And at least I shot some film when I was a kid. The kids growing up now are going to think of film the way I think of black and white TV or vinal. (I'm 25. The first mass-produced CD hit the market a few weeks before I was born.)

[Hans von Sonntag]

Hi,

to add some thoughts on this.

I see this from a pure pratical view:

Any producer/film company will stay with their workflow unless there are vital reasons to change this:

1. Financial gain
2. Creative decision

1. Financial gain: Any production with a comparable high budget will continue to shoot on 35mm (stock, lab, etc. are just a fraction of the budget) unless the director/dp want it otherwise. Low budget productions are prone to new technologies not always for the better.

2. Creative decision: Yes, this is a matter of taste. 35mm and digital are different.

Personal experience: I shoot most commercials on 35mm. Some with F900, none with F23, Viper. The workflow with HDCAM SR is too expensive compared to 35mm considering that the final master will be still SD. The F900 is by far not "good" enough for commercials that need moody pictures.

Corporate films or docs that I used to shoot on S16 are now all shot with F900. My Arri S16 has not been used since 2 years. This is because there is a financial gain in this kind of genre when shooting digital and not a big visual lost.

Side note: Arri just released a brand new S16 camera. Many DPs like S16 for its looks, not for its technical performance, same with 35mm. Many shows are still shot on S16. For anything that will be released in SD S16 is great. SD will be still around for many years (at least in europe).

A pure digital workflow with a camera like RED will enable low to mid budget produktions to gain much better production value. Will Film die? Some time, yes but not yet.

Hans

[oldphart]

... He stepped out of line with the short bus comment. Easy to get offended by that.
...
Steve

Is there some cultural reference I do not get?

I had to walk to school, btw.

[Chris Pickle]

Generally speaking, handicapped children ride a school bus that is shorter than a typical school bus, since obviously there are fewer handicapped kids then non-handicap. "Short bus" has become a negative slang term to compare someone with a handicapped child, most often mentally handicapped.

My son rides the "Short Bus" to school where he is integrated into grade 11 and I'm very proud of him. The reason I find these types of negative terms offensive, is that most often, these wonderful kids can't speak for or defend themselves. While racial, fat, blonde, etc, motivated negative jokes or comments are equally offensive, most people in those groups could speak out against it.

With handicapped children, it's a pretty lopsided fight. I'm not sure how someone could feel good about themselves at the expense of this particular group. I do think that most of the time, it's not intended and just ignorance, but I think we can evolve. Especially within a group of so-called professionals.

Cheers,
Chris

[Stephen Williams]

Would you say resolution has increased at all?

Hi Chris,

The latest slow speed films from Kodak have enormous resolution, far exceeding film of 10 years ago. Probably out resolves at least 90% (or possibly all) of the lenses out there when not stopped down to say T4.

Higher speed films are very clearly lower in resolution.

Stephen

[Michael Brennan]

Based on last 10 years development I do not think that digital imaging will vastly outperform film within the next 10 years.

Although development of digital recording and compression are at a fast pace the dynamic range of sensors is not improving at a fast enough rate.

One can argue it is image processing where many improvements have been are made rather than the sensor itself and it seems to me that having reached 14bit ADs and 12bit recording RAW we are at the threshold of diminishing returns.


Dynamic range is number one on the wishlist that is the hard nut to crack.

Aside from cramming more pixels into a given space a wonder breakthough may happen that greatly impoves other aspects of picture quality, but the last time this happened in ccds was the micro lens which gave us an extra stop overnight, that was the early nineties.


Mike Brennan

[Chris Kenny]

Aside from cramming more pixels into a given space a wonder breakthough may happen that greatly impoves other aspects of picture quality, but the last time this happened in ccds was the micro lens which gave us an extra stop overnight, that was the early nineties.


There's more than one way to crack the dynamic range problem. For instance, you could shoot three frames for every frame in the final output, and combine them into a single HDR frame. With sufficiently advanced image processing, you'd be able to match everything up and convincingly fake the motion blur.

Alternatively, if you had a camera with a crazy number of photo sites, could could put ND over some of them, so they'd preserve highlight detail. A quick patent search turns up a similar idea: apply different amounts of analog gain at different photo sites. That one might have an edge in that it could be turned off (which would increase sensitivity).

[jbeale]

If film is the reference target, there's no need for increased overall sensitivity on the CCD/CMOS, because it already outperforms film as far as noise/grain levels in low light. It's the bright highlight handling that is the issue, and when you have plenty of photons there are plenty of approaches, apart from just extra ND on some pixels.

1) Image stray light reflecting from the sensor surface onto a second sensor.
2) Put a 10% or lower pellicle beamsplitter in front, going to sensor #2.
3) You could fab a second sensor layer below the normal one (a la Foveon, but the lower layer is just for the bright highlights).
3a) Could be fabbed as a single chip, or
3b) the front one is backside-thinned and another separate chip bonded to it. 4) You could make photosites with physically different sizes, say the "highlights" sensors are 1/10 or 1/100 the area of the normal sensors.
5) Split sensor electronically: capture "overflow" electrons after a full-well condition in the main sensor into a secondary capacitor, but with lower transfer efficiency on secondary.
6 Nonlinear element on individual pixels that leaks off charge increasingly as stored voltage level goes up (additional FET as voltage-controlled resistor?)

I'd consider all these methods "intuitively obvious to one skilled in the art" and I am certain they've all been considered before. No doubt we haven't seen them yet only due to cost issues. Actually I think idea (4) has already been done in Fuji "SuperCCD" cameras, see for example http://www.dpreview.com/news/0301/03012202fujisuperccdsr.asp
http://en.wikipedia.org/wiki/Super_CCD
http://www.dpreview.com/reviews/fujifilms5pro/page18.asp

[Kevin Halverson]

5) Split sensor electronically: capture "overflow" electrons after a full-well condition in the main sensor into a secondary capacitor, but with lower transfer efficiency on secondary.

This approach would seem to imply that the photosite capacitor's dynamic range is the limiting factor. Rather, I would suspect that it is the number of quantization steps of the converter that limits the upper boundary (at least in designs that have high frame rate and site count products). Perhaps with a massive increase in the number of quantizers (and perhaps a change to their conversion approach) this could be increased, but the impact to things like the physical constraints of getting the results off the die (pin count) would quickly become a real problem.

This is a fascinating line of thought but I suspect, highly irrelevant as there are so many other significant bottle necks in the signal chain between the capture and display extremes.