Panavision Genesis

[GregIrwin]

I have seen the Genesis and being a hardcore film guy... WOW! It will be great. It is approximately the same size and weight of the Millinium and seems very simple to use. I also saw the sample reel comparing film vs. Genesis. The handful of us in the theater could not tell the difference between the two mediums. Panavision and Sony have suddenly gotten this right.

 

It uses the same 35mm lenses that we are used to and has the same image size and depth of field as the 35mm format. It compares to Super 35. I was completely impressed with what I saw. It will be the big hit at Cinegear.

[Sam Wells]

That is pretty much the reason why they made different HD lens'. But that's a bit different from saying 35mm optics aren't good enough at all for HD which is what was being said back in 2000.

I think they made the digital primos for the reason they said they did -- to get the most sharpness possible given the limiting resolution that the F900 & HDCAM would present.

 

As for 35mm optics "not being as good" - in 3 chip this is a moot point (barring the P&S) because you can't use them as is... BTW there was a long thread on CML about this, starting from questions about the Angenieux-Zeiss adaptor, with Panavision arguing it out with engineers from Zeiss and later Cooke delurking for the occasion. Although it went over my head, guess what, there were three different points of view, are you shocked ? !!

 

Of course with the digiprimes Zeiss proved what you could do with a design dedicated to solve a specific problem i.e. how to match the state of the art in film lens quality whlie dealing with the prism issues head on. I think nobody can say they didn't succeed.

 

This is different, we don't know exactly what the issues are, but I suspect they are not primarily about "resolution." Panavision, Cooke, Zeiss, Leitz know how to make optics with high resolving power. Interesting that Panavision is talking about electronic compensation for lateral chromatic aberration...... suggesting there _are_ issues - with that particular single chip at least - and ray angles..

 

-Sam

[Nate Downes]

Mating up a lens to a chip is much more complex than people seem to get. At my day job, I see the results of this first-hand. We have a 4MP digital camera at half the price of a 2MP camera, and guess what, I'd pick the 2MP anyday for clarity of picture.

[Guest Pete Wright]

Hey guys,

 

I Think:

 

This is not a 3K camera. It's a 4K camera, just as their new projector.

 

They may use different pixel combination for diferent aspect ratios, like the Viper.

 

Pete

[Tenolian Bell]

I think a great part of this to is to really stop and with a critical eye analyze: are the HD production tools just as good or even superior to traditional film tools at what they are designed to do best.

 

At this point there is nothing better than the 435 at what it is designed to do best. And it keeps evolving and getting better.

 

I'm sure tomorrow we will look at the Genesis and say that it's really nice. But will we still be able to say "the Arricam is still better at X, Y, and Z"? If we are, then do we say the Genesis is good enough or do we say it is not totally to the task and rigours that have been developed with traditional film tools.

 

If we are able to come to that conclusion then it's not really grasping on to nostalgia and the past, it would be the need to work with tools designed and optimised for what we do as film makers.

 

I would wager when it comes to the point when we can look at the Genesis or whatever comes about, and no longer say the Arricam, Panaflex, or 435 is still better at X, Y, or Z then at that point I doubt many people will argue against it.

[Guest Pete Wright]

What is Super 35?

 

Pete

[Michael Nash]

I would wager when it comes to the point when we can look at the Genesis or whatever comes about, and no longer say the Arricam, Panaflex, or 435 is still better at X, Y, or Z then at that point I doubt many people will argue against it.

The needs of filmmakers is so varied though, that NO camera is ever perfect for everything. You choose the tool that's best suited for the job, and sometimes the job is to shoot or post more cheaply than film will allow. People are already choosing HD tools because they are superior at what THEY do best -- capture images electronically.

[Michael Nash]

What is Super 35?

It's the "full aperture" measure of the 35mm/4-perf negative, .980 x.735 inches. It uses pretty much the full width of the film between sprocket holes, so it's bigger than the traditional Academy aperture of .864 x .630 inches. Academy aperture is slightly smaller and off-center to leave room for the soundtrack upon contact printing.

[Guest Pete Wright]

Thank you. So it is a 4K camera. It is 4Kx3K. I bet you you can use a smaller width of the CCD for 16:9 and Cinemascope. If you use the whole size, I'm sure Panavision has anamorphic lenses.

 

Do you know what is interesting. If it is film or CCD, 35 mm is 35 mm and 16mm film performance is about 2/3" CCD performance. So F950 type of cameras, when they come down in price will eventually compete with 16 mm film cameras.

[Filip Plesha]

Who ever said it is a 4K camera?

It has 12 million sensors but that does not mean that it records all that

data.

 

My cheap digital home-camera, for example, has 800 000 sensors and

records some 300 000 on tape.

[John Pytlak RIP]

The 12-megapixel analogy used in earlier Kodak ads was based on the general agreement that at least 4K scanning was needed to get all the information off a 35mm frame. In reality, if each grain of silver halide in film is considered a "sensor", each frame has much more than "12-megapixels".

 

Any single chip color sensor has to "divy up" the sensors among red, green, and blue, typically using a "Bayer" filtering scheme. (Bryce E. Bayer was the Kodak researcher who developed the technology). Plus, any regular array of pixels usually requires an "anti-aliasing" filter, decreasing resolution by a factor of at least 2 (Nyquist criteria).

 

The "film look" is much more than just resolution. Lots of R&D still being invested in new film-related programs, by companies like Kodak, Fuji, Thomson, Arriflex, and Panavision:

 

http://www.kodak.com/country/US/en/motion/...v2/sehlin.shtml

[Sam Wells]

Do you know what is interesting. If it is film or CCD, 35 mm is 35 mm and 16mm film performance is about 2/3" CCD performance. So F950 type of cameras, when they come down in price will eventually compete with 16 mm film cameras.

Sorry, but I think 16mm film performance = 2/3" CCD performance is a meaningless generalization.

 

F900's are already competing with 16/S16 in some areas and have been for a few years.

This is not news.

 

-Sam

[Jason Rodriguez]

Hi John,

 

A Kodak test I saw though showed that scanned at 4K resolution, 35mm film with enhancement was capable of a limiting resolution of 1800-1900 lp/mm. If that's so, then 8 megapixel cameras like the Canon 1D MkII, which is a bayer single-sensor camera is doing more than that, it's at over 2000 lp/mm. My Canon D60 has a limiting resoution of 1800 lp/mm, and that's a 6 megapixel camera.

 

BTW, what I'm about to say is not ment to be a disparaging remark, so please don't take it that way, but I'm hoping that as digital sensors get to where Kodak has officially been stating at numerous conventions and seminars the resolution performance of it's motion-picture film stock is at, that somehow the "specs" of film resolution don't mysteriously "shift" to higher megapixel numbers just do out-do the current digital state-of-the-art. Because let's be perfectly honest here: 35mm CN may have 12 megapixels of resolution, but I'm definitely not seeing 12 megapixels of resolution at my local cineplex. And yes, I've seen film contact-printed straight from the OCN (which is as good as it's going to get for a negative film), and oh my god, it looks beautiful-but again, I'm not seeing that performance when I go to the theatre to watch ANY movie. And you can't tell me you aren't loosing something in a contact print with another step of chemical processing, so even with a contact print you're not getting the total effect of what film could really do.

 

With digital 12 megapixels, sans compression, I shoot, and it's there. Or with the Dalsa 4K chip, I shoot and it's there-all of the 4K image, all of the dynamic range, all 16-bits of RAW data. Either of these cameras may or may not be better than film. But the fact is that with digital you can get everything that's there (with pluses and minuses of this being that you could clip information without enough dynamic range), and you get all the resolution that's there if you're not compressing it. So even with bayer interpolation, etc. you're at least getting 6 megapixels of clean digital data (no grain), and with OCN that must be contact printed to view, you have to be loosing some of those 12 megapixels in the contact print process, so I think we're coming out even in the end. And with fixed resolution digital projectors, even at 4K, you could have 18 megapixels, but after a certain point, you're not going to tell cause the projector just can't render the extra resolution.

[Nate Downes]

If I were to design an HD camera, I'd be designing it closer in nature to a Technicolor or 3CCD camera, a sensor for each color. And, for added effect, I'd even follow Technicolor's system for the basic "alpha" channel, a half-strength off of green. A friend of mine built himself such a camera using 3 B&W sensors and the results look absolutely remarkable. And the B&W sensors cost less than the color ones by an order of magnitude. So why isn't anyone doing this?

[Filip Plesha]

Jason. The tests done at 4K only show how much resolution does film have

when scanned at 4K vs. HD files uprezzed to 4K.

 

If the film was scanned at 8K it would show more than specified resolving power.

 

You have to be aware that a scan is never a perfect replica of one frame. (It gets close to that when you scan at double resolution and downres) A scanner will never

show films resolution so precisely to "describe" one resolved line on film with one pixel. If you want such precise scanning you have to use heavy oversampling.

 

The results you mention are showing the resolving power of film images in 4K scans, not the actual resolving power of film. The actual resolving power of film can be determined using a microscope.

 

The whole idea of 4K is based on the assumption that there is no difference in an optical print made out of original negative and digital 4K negative. That does not mean that the original negative does not hold more than 4K. If that was so, then there would be no need for 8 000 dpi scanning in still photography.

[Sam Wells]

:huh:

[Sam Wells]

Huh ? in response to 3 CCD, Am I misreading something ?

 

-Sam

[Nate Downes]

Huh ? in response to 3 CCD, Am I misreading something ?

 

-Sam

Don't mind me, I'm just discussing something I've noticed with the lower-end cameras, the move to single sensors for all functions and comparing the result to higher-end solutions with 3 sensors.

[Jason Rodriguez]

8000 dpi scanning is not typically done for 35mm, typically it's 4000dpi unless it's with the highest end drum scanners. Also there isn't much more information past 6000dpi according to this site.

 

Call me a skeptic, but it always seems as though the numbers for film keep rising like boats on the tide, no matter how good digital gets, when it reaches the numbers that the experts keep quoting film is capable of, all of a sudden, film has "so much more" information-from the same film stocks! Also, 8K film scanning is not practical right now for DI purposes. And many cinematographers I've heard from have stated that there isn't that big a difference between film scanned at 4K and film at 6K. So when members of the ASC, etc. are saying "there isn't that big a difference", I'd trust them, I figure they know what they're talking about. Now I didn't say there's no difference, just that after 6K, there's really not a whole lot more you're capturing, but you sure are taxing your computer equipment and spending a ton of cash (on top of the expensive film stock/processing).

[Balazs Rozsa]

Anyone wanna try and compare 4k to 65mm/70mm film?

The 65/70mm films have more resolution of course, but I think this is exactly why digital will be much better in a few years than film.

 

Digital resolution and storage will increase steadily and at some point the lenses will not be good enough to take advantage of all the resolution a 35mm CCD will give you. Then the CCD (or some other imaging device) size will start increase and quite soon 100mm will be reached for example. It is true such a chip will be expensive, but not very expensive. As opposed to how expensive it would be to shoot and work with 100mm moving film. With digital you only need to manufacture this chip only once. The output of these large CCDs will be incredible. Maybe the only limiting factor will be that you will not need more resolution after a point.

[Filip Plesha]

Jason...

 

It is a bit of a paradox to say that 8000 dpi is not used for 35mm. If it is not used for 35mm it is not used at all.

DPI is a relative value, it does not depend on the format. 4000 dpi will result in the exact same qualitty for every format from APS to large format (same grain size, sharpness etc.), the only thing that will change is the size of the digital image.

It is per inch, not per image width. If you need 6000 dpi for large format, you need it for 35mm too.

 

Anyway, I do agree with you that the difference after 6K is very insignificant.

(that is what I did say before). I am not talking about practice here, just theory.

There is some information at 6K or in some cases beyond recorded in film.

And yes 8K DI is pointless unless you are doing it for some future 8K projection.

I belive 4K was chosen because it is enough to make no difference from an original negative when printed.

But if someone's goal is to digitally archive a film in all its subtle glory, then

he should use 6K because you never know what standards can come.

Later on you might regret limiting the resolution to 4K when some 6K or 8K

imaging standard appears.

 

And as for the link, when you are testing to find out what is the maximum resolution that you need to scan film at, then you can't really use 6000 dpi, scan it at that resolution and say that this is the maximum.

You'd have to use 8000 dpi or more, and compare it to 6000 dpi and then say:

ok there is no difference between the 8000 dpi scan and the 6000 dpi scan, meaning that 6000 is maximum.

 

And of course it all depends on the emulsion. Some of the last 50 ASA still films

(they don't make much 50 ISO films anymore) had maximum resolution up to 200lp/mm , that is more than 8K for standard super 35 frame. And then again, some of the modern color negative films (for example portra 160 ISO) have about 120 lp/mm maximum resolution at 1000:1 contrast ratio. I really don't want to go and search for references about these resolving powers of these specific emulsions, but feel free to search the net and find out for your self.

 

Anyway, the bottom line is this:

Yes, 4K digital camera would probably end up being indistingushable from film in a 35mm print, or even a 70mm print. But that does not mean that in theory 35mm film can not resolve more than 85lp/mm (this is 4K). Just look at the MTF curves of even faster films. Non of these curves will show zero response at anything less than 100 lp/mm.

[Jason Rodriguez]

Hey, I'll agree with you on all your points.

 

BTW, when I was saying 1800lp/mm for 35mm film/6 megapixel digital (bayer CMOS), I meant 1800l/ph (lines per picture height), big difference there!

[Mike Brennan]

If we are expecting Genesis to be 4:3 aspect ratio how many pixels would be used for a typical theatrical 1.85:1 presentation?

 

In comparing single chip Genesis Super 35mm format to 3 chip 1920x1080 formats we should be aware that for 16x9 or 1.85:1 transmission/release not all of the Genisis 12million pixels will end up on screen.

 

Anyone care to do the math?

 

Also it has been reported that it is a strip ccd sensor. As far as I know this is a sensor with red green and blue pixels.

 

This could mean that there is no bayer filter no bayer processing...

 

 

 

 

Mike Brennan

[Jason Rodriguez]

A what type of CCD?

 

How would that work if it's a strip of pixels? Line-scan style camera instead of area-scan? I didn't know those could work without weird motion artifacting.

[Eric Steelberg ASC]

If we are expecting Genesis to be 4:3 aspect ratio how many pixels would be used for a typical theatrical 1.85:1 presentation?

 

In comparing single chip Genesis Super 35mm format to 3 chip 1920x1080 formats we should be aware that for 16x9 or 1.85:1 transmission/release not all of the Genisis 12million pixels will end up on screen.

 

Anyone care to do the math?

 

Also it has been reported that it is a strip ccd sensor. As far as I know this is a sensor with red green and blue pixels.

 

This could mean that there is no bayer filter no bayer processing...

 

 

 

 

Mike Brennan

I believe you are correct in all of the above.

 

I just returned form Cinegear. The camera does have a 4:3 sensor and it does NOT use bayer filtering/processing. I assume that means it is a strip sensor. It is a 4k and by using a 12mp sensor it will be future proofed meaning it will be able to record higher resolutions when the recording technology catches to the rest of the camera. The tape format is HDCam but is a new type of tape stock.

 

Here's the most interesting think I learned:

When you "overcrank" the camera, the tape runs faster, and when you "undercrank" it runs slower...just like film. Right now the variable speeds are only possible with the Sony onboard deck and they had to refer to a Sony Engineer to see if it was possible to do it while running it out for uncompressed capture. The ability to ramp is expected in the future.

 

Oh and the viewfiinder is crazy sharp, but very big.