ARRI CEO: "We want to have the worlds best digital camera this year".

[Chris Nuzzaco]

My big concern with 8K are super tiny photosites and noise, less range, etc..

If Arri can somehow master a 3 chip (35mm size chips) that can take PL lenses already in use... That would be quite the achievement.

But I hear thats not very easily done.

 

[Kevin Lang]

There is always going to be something better out there. You have to just use the tools you have. This my dad can beat up your dad scenario is lame. All of these cameras will give you a good image.

 

[Poi Boy]

I wonder what red will have in 14 months ?
-A

 

[Dj Joofa]

Way to wake up sleeping giants, Jim! Too bad Red is already way ahead of them in technology having already released the world's greatest digital camera. And if they expect to release something in 12-14 months, I don't even want to know where Red will be.

One has to understand the development process in a technical company to realize that initially rapid development is not sustained as a company starts growing for several different reasons -- in fact it is not surprising that many times the whole marketing teams for the initial phase of the company and the phase once successful projects have been launched, are different.

Red has little obligation right now. But once in heavy production mode they may be slowed down as others do, and spend a lot of time in "maintenance" mode vs. "development" mode.

The others companies that you mention, such as Arri, are well-placed in terms of human capital and infrastructure. However, that may not necessarily translate into a commitment to release development resources from other projects and putting into a "Red-centric" project. However, if Arri's CEO has said that, then it may be taken to mean that he is willing to go that route.

 

[Dj Joofa]

Arri is not a threat for RED owners.

Hans

Actually, that may turn out to be very true. Red and Arri may co-exist easily as they might be targeting different markets.

Arri is an established company, well-entrenched in the film world, and I am sure they have their channels to pull some strings when it comes to a decision making time.

Bottom line is that Red may be suited for independent filmmakers/cinematographers/DPs who can directly buy from Red, where as Arri might like to target established production pipelines at premier places.

 

[peter roehsler]

All fine with me. I won´t be able to afford it, that´s for sure. Rental it will be anyway and I also don´t fancy a concept that I should end up the least expensive part of such a setup.

 

[Sanjin Jukic]

"ARRI Utilizes Cypress CMOS Image Sensor for Motion Picture Cinematography

Cypress Semiconductor recently announced that its OSCAR custom CMOS
image sensor is incorporated into the ARRIFLEX D-20 Digital Motion
Picture Camera and ARRISCAN Film Scanner from ARRI GmbH.

The 6-megapixel sensor is the same size as a Super 35-mm film aperture,
enabling the D-20 camera to use the same prime, zoom and specialty
lenses as 35-mm film cameras while shooting in High-Definition (HD).
The resulting digital image has the same depth of field as 35-mm film,
allowing film makers to direct a viewer's attention to a specific part
of an image, a feature that is a highly valued creative storytelling
tool.

The OSCAR CMOS image sensor gives the D-20 camera the ability to read
out high frame-rates despite a high active pixel-count of 2880 x 2160
pixels, as well as the ability to run speed ramps. Additionally, the
OSCAR sensor delivers excellent color fidelity and high dynamic range
for crystal-clear HD footage."

LINK>>>

"Cypress Semiconductor Corporation

We build the sensor for your image.

Cypress develops and markets innovative and high-performance CMOS image sensors,
based on patented technologies and know-how built up for more than 15 years.
Our CMOS active pixel sensors are based on innovative design approaches realized in
standard CMOS processing technologies.

We offer standard products as well as custom design of CMOS imager ASICs.
It is our mission to supply our customers with the CMOS image sensors they need,
when they need them, and as long as they need them.

A significant part of our business is the design and production of customized CMOS image sensors.
Cypress delivers custom image sensors according to the agreed specification and planning.
First silicon to spec is our commitment.

With more than 15 years of experience in the field of CMOS active pixel sensors,
we have proven solutions for multi-megapixel digital photography,
large area sensors (w/o stitching), ultra-high-speed, imaging for machine vision,
linear and 2D barcode imaging, medical XRAY imaging, single-chip camera integration,
and radiation-hardened CMOS image sensors for space and nuclear use."

LINK>>>

 

[Brent J. Craig]

I have googled all of the quotes in the original post and searched every way I could. Can you provide a link to this Arri interview?

 

[sparkhope]

I hate to agree with the guy, but I have always thought you needed a 6K sensor to really get beautiful 4K.

Pixel for pixel, I just really don't think a lot of 4K stuff, whether shot on a DLSR or the RED, will hold up at full res.

And yet aren't some film born DI's scanned on 4k scanners while the majority opt for the cheaper 2k variety?

 

[Dj Joofa]

And yet aren't some film born DI's scanned on 4k scanners while the majority opt for the cheaper 2k variety?

This is true. Because the condition of theater screens does not justify going for more than 2k scans. So why spend money when its not needed.

Also, please note that in a theater there is only one exact distance that a human would have the best response visually. Hence, different rows of a theater will have a different response, and there is no point in going over 2k these days, as this just means that you would have to sit closer to the screen for the best response. Just like the number quoted for a typical NTSC display CRT is that for the best visual response one should be seated away 6 times the picture height and for HDTV 3 times (I don't remember these numbers 6 and 3 exactly, and they are just from the top of my head).

 

[Tom Lowe]

My big concern with 8K are super tiny photosites and noise, less range, etc..

If Arri can somehow master a 3 chip (35mm size chips) that can take PL lenses already in use... That would be quite the achievement.

But I hear thats not very easily done.

You could probably look at 65mm sized sensors recording at 6K.

8K is not a real spec, whereas 4K actually could replace 1080p as the gold standard of consumer high-def at some point. 4K monitors are on their way.

 

[Tom Lowe]

And yet aren't some film born DI's scanned on 4k scanners while the majority opt for the cheaper 2k variety?

I suspect plenty of films will not hold up at 4k either. Hell, a lot of blurays look like crap at 1080p, especially stuff 15 years or older.

 

[martinnoweck]

Dear Mr. Dejan Ilic. -

I'd like to discuss the pricing of your matte boxes ...

LOL
Martin

 

[Cüneyt Kaya]

Everybody hurry up and try to catch up with the "vaporware" ....

all very PREDICTABLE.

on another note - here's a letter I just started -

Dear Mr. Dejan Ilic. -

I'd like to discuss the pricing of your matte boxes ...

...it is really cheap if you compare it to a NASA rocket.
Dear Mr. Ilic, if you can prove that the astronauts use your mattebox while making some nice pictures of the moon, I´ll understand, but if not...

 

[J.D. Frey]

my understanding is that there will one day be sensor upgrades, etc. for the red one as well. So really you are buying into a package that should last quite awhile.

 

[Radoslav Karapetkov]

For me it sounds pretty like a Balkan's macho story (Mr. Dejan Ilic),

it's about to make a big dust using a bit of salt to make a VERY LOUD and big bluff.

And then after all (fight) you could fast leave a battlefied defeated .

LOL.

Yeah. Dogs that bark usually don't bite... and viceversa.

 

[Dj Joofa]

my understanding is that there will one day be sensor upgrades, etc. for the red one as well. So really you are buying into a package that should last quite awhile.

That could be tricky. As I mentioned in an earlier post that before each camera ships an extensive calibration is done to reverse the effects of dark currents and non-uniform sensor behavior to uniform light exposures. Every sensor has different masks for these. This calibration data has to be some how applied to the signal coming out of the sensor, either in-camera before output, or off-line processing on your computer. I do not know how Red would handle this situation.

 

[Paris Remillard]

Now if Arri (or RED) could have a single 8K bayer filtered sensor with every other photosite NDed, and give us 4K with much more latitude, that would be okay by me.

 

[Dj Joofa]

Now if Arri (or RED) could have a single 8K bayer filtered sensor with every other photosite NDed, and give us 4K with much more latitude, that would be okay by me.

It is interesting to see a fixation with higher and higher resolution. I posted some comments on a similar topic on an other post and I repeat them here:

It is an unfortunate consequence of a misconception that is harbored many times that higher resolution say 4K, is necessarily better than a lower one, say 2k. Unfortunately, the dynamic range or the bit depth per pixel (say 12-bit per pixel) and camera sensor resolution (say 4k) are often independently quoted, without ever trying to ask the elusive question that given that particular dynamic range, is the advertised resolution the best one, or vice versa?

For a given micron technology size, the dynamic range increases with making the pixel size larger in terms of signal to noise ratio. However, on the other hand making the pixel size smaller results in higher spatial resolution and better modulation transfer function (MTF). Therefore, given a particular technology, there must exist a pixel size that strikes a compromise between high dynamic range and signal to noise ratio on the one hand and spatial resolution and MTF on the other hand. Though, it is not always immediately clear how to determine that optimal pixel number, and especially how to trade off dynamic range and signal to noise ratio with spatial size and MTF.

Image quality should be the ultimate objective of selecting "optimal" dynamic range and pixel count. Although, the connection is a little fuzzy, it is possible to derive that magical "optimal" pixel count and dynamic range for a given technology under certain constraints.

Marketing would like to throw in those numbers such as 4k, 12 bits per pixel, etc., without satisfying a curious user that for your particular technology:

(1) If I fix bit depth or dynamic range, is the pixel count advertised the most optimal one or a smaller pixel count would have worked better for greater image visual quality?

(2) If I fix pixel count, then is the advertised bit depth the most optimal one or a smaller one would suffice?

 

[Bruce Allen]

4k from RED has always held up well on the big screen whenever I've seen it. It's always nice to think that downsampling is the "solution", but there's no perfect downsampling filter, which sorta puts a spanner in that idea.

Graeme

So true on both counts...

Bruce Allen
www.boacinema.com