Rumour: Arri wants to strike back

[Eddie]

Would be nice to get one stop more and we would get it with Mysterium-X soon.

But most important thing is to get a fast lens.

Download and watch those examples (outdoors and indoor) that were shot on RED1 with available night light all @ wide open

Leica Noctilux-M 50mm f/1 (Made in Canada)

Leica Noctilux-M 50mm f/0.95 ASPH (Demo lens, Made in Germany)

Canon 50mm FD L f/1.2

Voigthlaender-Leica-M 35mm f/1.2

Sanjin.. fast lenses does nothing to increase dynamic range... And as I have pointed out before, some people prefer to stay above f/2 simply BECAUSE of the shallow dof. This leaves them wanting good iso/gain performance and optical image stabilisation, instead of speed.

 

[Steve Sherrick]

Sanjin.. fast lenses does nothing to increase dynamic range... And as I have pointed out before, some people prefer to stay above f/2 simply BECAUSE of the shallow dof. This leaves them wanting good iso/gain performance and optical image stabilisation, instead of speed.

I agree, fast lenses are only part of the story. It's a combo of lens, lighting, desired depth of field, shutter speed, etc that all come into play and affect dynamic range. Mysterium-X should improve things, and Monstro should take it to the next level.

 

[stephan orlandic]

I would not.
RED can offer uncompressed anytime they want. In fact, the boys from Codex showed me a unit with a connector they installed specifically for RED's RAW PORT. RED cancelled the RAW PORT because the demand was not there - not because they couldn't make it. When the workflow is considered - uncompressed 4K in long form is VERY inefficient. That might change, sure. And has anyone heard RED say that you can't get uncompressed out of the EPIC?? I didn't hear that Mitch ... did you??

This might be Your perspective only because it seems that You have never worked in a 4k production/postproduction which does not envolve 36MB/s 4K RED ONE files. Your Scratch system is only efficient in 4K worflow in hd/2k resolution with half quality, and only with 4K Red files. But try loading 4k 16bit tiff or 4K 10bit dpx and see what's happening.
Or work with RED 4K files, in 4K full quality.

Also, compare a compression level, especialy in blue channel with some Northlight 4k 35mm scans or Dalsa 4k images, and You will see why there is serious need for 4k uncompressed materials.

Also, anyone thinking about 4K post, knows that it takes a lot of time, and a lot of money. So if it is inefficient as You say, at least it offers a picture with no compromise. And in 100M+ production nobody cares about speed in 4K, when there is a big sacrifice in quality.

Also, any TITLE, VFX and CG work even from RED ONE's 36MB/s compression, will be done from 4K tiff or 4K dpx, so it will be still inefficient as if it have been shot with some 4K uncompressed camera/scan.

As for 4K uncompressed grading being inefficient, Baselight 8 and Resolve 350, can grade uncompressed 4k in real time, with 30-40 layers of CC, where Scratch can't even grade a full quality 4K RED.

You were saying something about inefficent ?

 

[Mitch Gross]

I didn't put it out there as a challenge to what RED is offering, and frankly maybe they will offer it and show up everyone else. I'm just saying that there are certain things that clients want -- right or wrong, they want 'em -- and if someone else delivers then they'll go there. And everybody is trying to move down that path.

 

[Chris Kenny]

As for 4K uncompressed grading being inefficient, Baselight 8 and Resolve 350, can grade uncompressed 4k in real time, with 30-40 layers of CC, where Scratch can't even grade a full quality 4K RED.

That's not just because it's uncompressed, it's also because there's no debayer required.

You were saying something about inefficent ?

Even most major Hollywood features aren't presently posted in 4K. So presumably at the moment it's considered too inefficient to justify the quality benefits even on those $100M+ projects.

Anyway, this is largely a moot point, with the data rates offered by the new Epic models. With images fed through a JPEG 2000 derived codec like Redcode at those kinds of compression ratios, you're probably going to need to do statistical analysis of noise patterns to tell the original apart from the compressed version at those compression ratios.

 

[stephan orlandic]

That's not just because it's uncompressed, it's also because there's no debayer required.

Even most major Hollywood features aren't presently posted in 4K. So presumably at the moment it's considered too inefficient to justify the quality benefits even on those $100M+ projects.

Anyway, this is largely a moot point, with the data rates offered by the new Epic models.

You have completely missed the point of the hole post.

I was telling Mark, that his vision of efficient post workflow (Scratch) can't playback neither REDCODE in 4k nor 4K uncompressed.

And that You have to work at least a good part of the film in uncompressed data.

And that there is an objective and real need for uncompressed/raw data from camera.

We are talikng here why uncompressed vs. lossy compression and not why 4K at all.

And FYI, REDcode 225 (whatever it will mean) in 6k 16bit camera and especialy in 9k 16bit camera, means LOSSY compression.

 

[Joe G.]

"With images fed through a JPEG 2000 derived codec like Redcode at those kinds of compression ratios, you're probably going to need to do statistical analysis of noise patterns to tell the original apart from the compressed version at those compression ratios."

You have hard data about the compression ratios?

Will we be able to minimize the compression somehow and maximize the data rate for a given frame size? Will I be able to get the listed 100MB/s at 6K, 4K, 3K and 2K? Would a 100MB/s 3K be able to be compared side by side to the 100MB/s 6K?

I don't believe the compression ratios have been determined yet. So you may be premature with your answer.

 

[Chris Kenny]

You have completely missed the point of the hole post.

I was telling Mark, that his vision of efficient post workflow (Scratch) can't playback neither REDCODE in 4k nor 4K uncompressed.

Which doesn't make it particularly inefficient, because you don't actually need real-time 4K playback even for those very rare projects that are actually finished in 4K.

But you appear to have missed my point, which is that uncompressed 4K is sufficiently unwieldy that even the big-budget guys don't seem to want to work with it. Don't you think this suggests it's rather a bad format to be directly acquiring all your footage in? (Possibly dozens or even hundreds of hours, for a feature.)

And that You have to work at least a good part of the film in uncompressed data.

There's a huge difference between working on a few shots uncompressed, and having the camera capture uncompressed. For starters, you probably have at least a 10:1 shooting ratio, so even if every shot in a film requires an uncompressed workflow to put titles on it or to composite in CG, you're still dealing with 90% less uncompressed data if you use compressed capture and only create uncompressed versions of the shots that actually end up in the movie.

And that there is an objective and real need for uncompressed/raw data from camera.

There's really not. There might be a need for somewhat less compression in some instances. But something close to 2:1 compression can be mathematically lossless, and you can get image compression of 4:1 or maybe even a bit better with no losses other than essentially rearranging the noise a bit.

We are talikng here why uncompressed vs. lossy compression and not why 4K at all.

And FYI, REDcode 225 (whatever it will mean) in 6k 16bit camera and especialy in 9k 16bit camera, means LOSSY compression.

24 6144*4608 (4:3 6K) 16-bit raw frames compressed to 225 MB/s gives you about a 6:1 compression ratio. And remember, larger images tend to compress better. The benefits of uncompressed vs. 6:1 compression are going to be extremely marginal.

More to the point, though, a 6144*4608 16 bit frame compressed to 225 MB/s is going to still look better than (say) a 4096*3072 (4K 4:3) 16-bit uncompressed frame (576 MB/s). As I've pointed out several times in the past, of course uncompressed results in better quality if all else is equal. But it always results in worse quality at a given data rate. And in the real world, there are always some constraints on your data rate.

 

[Chris Kenny]

I don't believe the compression ratios have been determined yet. So you may be premature with your answer.

I'm just basing this off of the raw frame sizes and the maximum data rates Red has announced. It's true that actual data rates may end up being a bit lower. But if this is the case, it will presumably be because Red has determined lower data rates yield perfectly acceptable images. (Remember, it doesn't require more processor power to compress data to lower ratios. If anything, it's generally the other way around.)

 

[Robert Berger]

Whatever,

but i would say that companies like ARRI or PANAVISION will still owe the market

for making the main stream of feature films until the companies like a RED will take them over.

But the companies like a RED should finally realised that the main weapon is not the size of the sensor in the camera

and also its resolution, fps speed and whatever in that way.

Simply it is called the set of FAST LENSES (aspherical and anamorphic) that would make you the next king on the market.

In a year or two everybody would have a powerful sensor but all that depends on the second stage in overall acquisition

and there is simply the market for the (FAST) LENSES.

So if you are looking at the future of that thing you should INVEST in a development of fast spherical or anamorphic

lenses and then you could owe the market anyway even if it is a FILM or a DIGITAL acquisition under the question.

Cameras are not owing the market.

Its about overall system and workflow.

If you are missing something or you are in a "scissors" or your partners and also if you are counting only on one thing then

you are the looser of the all game very soon.

I don't think fast lenzen wil win, because the littele dept of filed you wil have and a digital camera with later on the digigtal projection where 4,5 or 6K still stay 4,5, or 6K you need dept of field .So the market wil be won by digital camara's with more tha 640 ASA !! Mark my words, faster ASA camera's not faster lenzen!!

Robert

Robert

 

[Sanjin Jukic]

Sanjin.. fast lenses does nothing to increase dynamic range... And as I have pointed out before, some people prefer to stay above f/2 simply BECAUSE of the shallow dof. This leaves them wanting good iso/gain performance and optical image stabilisation, instead of speed.

Eddie,

See the shallow DOF @ f/1.2 wide open at below picture

The picture is from the test I shot last night.

THE LAST NIGHT TEST FOOTAGE>>>

thanx for putting "THE END" in your clip - it totally transformed these snippets into an epic story :clown2:

Thanx for your comment told in a kind of Swiss humor :clown2: .

I don't think fast lenzen wil win, because the littele dept of filed you wil have and a digital camera with later on the digigtal projection where 4,5 or 6K still stay 4,5, or 6K you need dept of field .So the market wil be won by digital camara's with more tha 640 ASA !! Mark my words, faster ASA camera's not faster lenzen!!

Robert

Robert

Robert to see some DOF @ f/1.2 wide open follow the footage from the last night with the link above.

 

[Stephen Williams]

What did you set the focus on when you made that shot? The ear looks sharper than the front eye.

Eddie,

See the shallow DOF @ f/1.2 wide open at below picture

The picture is from the test I shot last night.

THE LAST NIGHT TEST FOOTAGE>>>



Thanx for your comment told in a kind of Swiss humor :clown2: .



Robert to see some DOF @ f/1.2 wide open follow the footage from the last night with the link above.

 

[Sanjin Jukic]

Stephen,

it's a handheld and documentary kind of shooting when you focus by eye in EVF.

For example this guy anyway didn't like that I was shooting him at all and

I moved camera a bit up to shoot the lights after just to avoid a contact.

Let's say a sort of "pararazzi" shot when you are not absolutely sure if you are in focus or not.

The lens was wide open at f/1.2 for sure.

Have a look at the footage again>>>

 

[Jay A. Kelley]

I will say that RED's Lack of open file format is a MAJOR downside to their product line. If another camera were to come out with a completely open format from the get-go, then they would get very serious consideration from me.

Jay

 

[Cüneyt Kaya]

Stephen,

it's a handheld and documentary kind of shooting when you focus by eye in EVF.

For example this guy anyway didn't like that I was shooting him at all and

I moved camera a bit up to shoot the lights after just to avoid a contact.

Let's say a sort of "pararazzi" shot when you are not absolutely sure if you are in focus or not.

The lens was wide open at f/1.2 for sure.

Have a look at the footage again>>>

sanjin dont you think that its lets say unpolite posting footage of a person without permission in the internet?

 

[Cüneyt Kaya]

I will say that RED's Lack of open file format is a MAJOR downside to their product line. If another camera were to come out with a completely open format from the get-go, then they would get very serious consideration from me.

Jay

yep...nobody will understand why it takes that long.


and lets say red will keep the status quo for a while----then they should provide the right Tools to get it to work.

but as long as properly working telecine software (Redcine) would eat into scratchs market we cant expect that the guys from assimilate will work hard for a solution.

pc redcine is a catastrophy.
the most annoying bugs are known.....loading more then 120 clips in redcine leads to what?


and honestly......that a skater from ny and a nz nerd saves the workflow with additional software and workarounds....is just weird..hey i paid more then 30 000 usd......ok i know its all beta...
but its now 15 months
maybe red should put some pressure on their partner assimilate:
something like
redcine should be as stable as scratch.

but the best....really the best would be if they publish the SDK.....RED is not a software company....right?
Right now the guys doing the software doesnt do RED a favour imho.
There would be much better software tools in 6 months than redcine...imho.
Let RED build the Camera and others the software.


--------------------------

just to clarify...


i dont want to bash red or hear something like, go get another cam.
i am one of the greatest supporters fo red in germany.
what i dont want is:
red falling behind

 

[Nick Wolf]

Kinda hard to put the squeeze on someone who has got you by the balls ( : Emoticom Optional : )


DogDay.

 

[Lucas Wilson]

I was telling Mark, that his vision of efficient post workflow (Scratch) can't playback neither REDCODE in 4k nor 4K uncompressed.

We have several customers in LA that are working on 4K uncompressed 10bit DPX in SCRATCH. It works just fine, as long as you have the appropriate hardware. SCRATCH cannot currently do multiple layers of realtime CC in 4K. That is currently reserved for a Baselight8, Resolve, and Pablo.

There is obviously a price/performance trade-off as well. It goes without saying that to get that kind of performance, you will be paying exponentially more for a complete system.

Best,

Lucas
-----
ASSIMILATE, inc.
LA, CA, USA

 

[Lucas Wilson]

but as long as properly working telecine software (Redcine) would eat into scratchs market we cant expect that the guys from assimilate will work hard for a solution.

REDCINE is a RED product. Feature development and feature requests are driven by RED, not ASSIMILATE.

Lucas
-----
ASSIMILATE, inc.
LA, CA, USA

 

[Cüneyt Kaya]

REDCINE is a RED product. Feature development and feature requests are driven by RED, not ASSIMILATE.

Lucas
-----
ASSIMILATE, inc.
LA, CA, USA

thank you for claryfing...
what are the feature requests you are working on...i guess its a colabiration or you can tell me who is doing the coding at RED...

btw. scratch really works as datamanager/telecine...i mean stable etc.
who is responsible for that redcine has problems