Epic X, Epic FF35 and IMAX...

[Tom Lowe]

Yes, but remember that's 5.6K RGB, not 5.6K RAW. You'd need nearly an 8K RAW sensor to get 6K RGB resolution.

Also remember "Baraka" was shot in 5-perf 65mm, 1/3 of the size of the 15-perf 65mm IMAX frame.

I was careful to say that FF35's 6K RAW was only "in the same ballpark" compared to 5.6K RGB. The difference between 5.6K RGB and 6K RAW is significant, to be sure, but it's not a HUGE, insurmountable difference, IMO.

I posted the 5.6K RAW comparison to the 5-perf 8K BARAKA master just because it's the most convenient, and the scan was so recent from a negative I have to assume has been cared for. The 5.6K RAW images are so much better than the scanned 65mm 8K five-perf that I have to assume FF35 6K RAW will not be totally blown out of the water by 15-perf. Maybe I'm wrong... but I really think people are going to be stunned once they get a taste of 6K FF35 RAW motion. And I mean.... STUNNED.

 

[Mohammed El Sharqawy]

I'm with you Tom.. its really difficult to compare these huge sizes and resolution.. in fact you might get more res from an Imax 15 perf shooting.. but not to forget that processing 6K Raw might be much easier than 8K RGB .. let's say.. in realtime.. RED Rocket™™.. the upcoming version of premiere and ae etc.. easier and more creative workflow will add more aspects to the image quality that cannot be done with 8K RGB.. simply less practical to do it which renders it less creative..
one of the things that makes RED one image is unique is that it is dealt with better workflow than 35mm.. its not like film.. but its unique.. might be better as the future technology RED one gets 3.2K effective res and 35mm gets 4K theoretical res ..
645 should be better in case you have better lenses.. which is another addition to the FF35 format..
I mean in the end you'll have a better chance to fill the quality of FF35 than to fill a 15 perf IMAX and its workflow..
and of course frame rates and weight is another big issue.. makes FF epic more creative..

btw Tom you said epic 645 or FF35 TATTOO .. will be there a TATTOO program for 645 and 617? or that TATTOO program will be only tight to Epic-X and all other DSMC camera's will benefit from that TATTOO testing period? I think there is no TATTOO scarlets announced yet.. wrong?

 

[Jeff Kilgroe]

Tom, I think you make a good case here for 6K FF35 to be positioned for IMAX under certain conditions. Call it bare minimum or entry-level, it is possible. Many IMAX productions these days are undergoing a 4K DI process, some only a 2K process on occasion. The 6K RAW definitely lends itself to either of these applications. However, I think the EPIC 645 @ 9K is the REAL IMAX contender here and we must not forget the EPIC 617, which equates to roughly 2.5X 15/70 frame.

As digital projection becomes more commonplace, especially with the proliferation of 4K cinema projectors, FF35 at 6K RAW is not going to provide what is needed to present an IMAX image that is superior to 4K projection in a typical cinema.

For me, EPIC-X seems like a logical stepping stone, but it's EPIC 645 that I want and the 617 as well. Scarlet FF35 would be attractive if RED does indeed produce it. However I think most everyone is going to push for the 645 for the extra resolution and imager size combined with the higher frame rates that have been mentioned. Of course we will have to see what actually is possible with the upcoming cameras. I'm fully expecting all these announced cameras and specs to change as they come closer to reality.

 

[Tom Lowe]

Good summary, Jeff.

You have to believe that "True Digital IMAX" will be 8K projection (after a 4K interim, transitional period), and actually the 9.3K 645 does offer at least some downsampling to 8K. As David said, 10K or 12K would be more ideal for an 8K finish, but 9.3K is within reason and should be very stunning to viewers' eyes! I know nothing about the glass at that format size, though.

 

[Pawel Achtel]

It's all nice and good, but what glass will resolve this at any decent MTF and how to shoot the damn thing in focus when DOF is 1mm?

 

[Michael Hastings]

I know nothing about the glass at that format size, though.

It's all nice and good, but what glass will resolve this at any decent MTF and how to shoot the damn thing in focus when DOF is 1mm?

Well I'm just learning but there is a lot of 645 glass Hasselblad, Pentax, Zeiss, Mamiya - mostly excellent manual lenses from the mid 90s that cost $800 to $2000 in the 90s so $1500 - $3000 in today's dollars. Many available used for $100 to $300. Superwide and long tele's $600-$1200. Typically all metal construction and build style and quality similar to Zeiss ZFs and most seem to get very high ratings for optical quality relative to 35FF lenses. (Also nice camera bodies for $200 to $500 - depending on manual or autofocus - if you want to play with some stills)

The Flange focal distance is much longer than FF35 or PL so should be easy to make custom mounts for any of them.

As far as focusing, for narrative it is obviously difficult but my guess it is more likely nature/adventure stuff that most of us are talking about so F5.6, f8, F11 would be common - particularly with an effective 800 ASA or more monstro sensor.

 

[Jeff Kilgroe]

It's all nice and good, but what glass will resolve this at any decent MTF and how to shoot the damn thing in focus when DOF is 1mm?

No different than shooting medium format stills or 70mm film / IMAX now really. Not saying that it's easy or anything like that, but I'm sure RED will have a few tricks up their sleeve for those larger format sensors and optics.

My Mamiya 645 lenses are very nice optically, very sharp, should resolve just wonderfully with the 9K 645 sensor. These lenses are beautiful with a 28MP digital back. Mechanically speaking, the lenses are crap with short focus throws, rough gear motion, noisy motor system, etc... As Michael points out, the older mechanical lenses are typically a lot better constructed. And there are a lot out there for relatively cheap. I'm still unsuccessful in finding a decently priced wide prime though for 645 though. I'm talking about ~28mm or wider. I already have a 35mm in my kit. The 28mm Mamiya is going for $3600 for the AF model and $2000 for the older manual version. I found a Hasselblad V mount manual one a while back for about $900 and didn't buy it since I would have to use a mount adapter with it. Dumbass me. I should've snatched it up.

 

[Michael Hastings]

Ha, Jeff. You're like me: wider baby, wider.

Of course even that $2000 28mm isn't that expensive when you consider what a PL mount 10 or 12mm (S35 equivalent) goes for.

 

[Peter Osinski]

Hey, I just threw this together so that I could tell more closely how these sensors and film compare. I know there is one on the announcement page but it helps me having things on top of each other.

Either way I included some larger format stuff so I thought you guys might like it/ maybe it will help your discussion.

And on that note, I really am not sure after looking at this that I believe FF35 will be really able to compete with IMAX film, which according to what I've read is 15 perf 70mm film. The 645 for sure will be on its level and the 617 will be just insane.

The ff35 will be able to compete with Standard 65mm (5 perf 70mm) without question. None of this really matters right now though because we are projecting at 4k max, and with the "multiplex design" versions of IMAX that a popping up around the country, I doubt we will see too many more full size IMAX screens.

 

[JD Holloway]

I doubt we will see too many more full size IMAX screens

Good point, think mini-imax.
Imax-home theatre or Mall-max. As long as view angles are relatively the same, things get much simpler. You don't need 25000 lumens or 8+K resolution to make a nice 12-20 foot Imax'ish screen at home...

From Imax wiki,
"The screens used by digital IMAX installations are also much smaller (28x58 feet) than those found in traditional IMAX cinemas...
IMAX digital currently uses two 2K-resolution Christie projectors with Texas Instruments Digital Light Processing technology alongside parts of IMAX's proprietary system. The two 2K images are projected over each other, producing an image that is potentially of a slightly higher resolution than common 2K digital cinema. Originally, IMAX had been considering using two Sony 4K projectors. Some reviewers note that this approach may not produce image quality higher than using one 4K projector, which are available for some non-IMAX theaters, including AMC's own."

So 4K worth of projection is "good enough" for Digi Imax, even with edge blending. Now this is very different from film Imax but is still Imax by name.
I'm thinking any 4K+ projector with enough lumens to cover 28x58 feet with 4k of measured resolution is good enough for Digi-Imax. 6K FF Epic is definitely in that ballpark.
You may not make people happy though with this "small screen Imax".

 

[Tom Lowe]

As Yoda said, "Clear your mind... you must unlearn what you have learned."

We have to be careful not get too caught up looking only at physical sensor size, rather than understanding the huge underlying technical differences between massive film negatives and high-resolution digital RAW.

If you asked most people, "Which has better, cleaner resolution, a single frame from a 6K 24-megapixel Nikon D3x, or a single frame scanned from 65mm 5-perf film?" Most would probably guess the Nikon D3x. And they'd be right. If you ask most cine professionals right now which will have more resolution, the Red Epic FF35 6K Raw camera, or 5-perf 65mm, most would probably choose 65mm out of instinct.

Why? We have already established that the 6K RAW has better resolution (dynamic range issues aside). But people are not used to thinking about large RAW frames as a motion-picture format.

Take a look at these Nikon D3x frames, and tell me honestly whether you think you could scan a frame from 65mm 5-perf and beat it?

http://chsvimg.nikon.com/products/imaging/lineup/digitalcamera/slr/d3x/img/pic_001b.jpg

http://chsvimg.nikon.com/products/imaging/lineup/digitalcamera/slr/d3x/img/pic_007b.jpg

http://chsvimg.nikon.com/products/imaging/lineup/digitalcamera/slr/d3x/img/pic_004b.jpg

I have seen 65mm projected, and it can't touch that kind of image fidelity. Not even close.

 

[David Mullen ASC]

Tom, just keep in mind the film images do benefit to some degree from the fact that the detail, i.e. grain, lies in a different position on each frame, so there is some increase in resolution when projecting versus looking at a still frame. I'm sure digital images also benefit, but just not as much since the pixel pattern is fixed. I've seen plenty of side-by-side still comparisons where the 35mm version doesn't look as sharp, yet in projected tests, the 35mm version looks much better due to averaging of grain over multiple frames.

On the flip side, projected digital images sometimes look less sharp because without a grain pattern over the image, your eye only has actual object detail to focus on, whereas with a grainier image, if the grain is sharp, we tend to assume that the image is sharp. Without grain, the image can therefore feel soft.

I agree though that physical sensor size is not really an indication of resolution. On the other hand, pixel count really isn't a measurement of resolution either.

Unless you've seen 65mm motion picture film projected compared to a digital cine camera footage projected, you're comparing apples and oranges if you are thinking of a digital still seen on a computer monitor versus motion pictures.

And since 6K RAW for moving images doesn't exist yet, you're speaking in hypotheticals.

 

[Tom Lowe]

I hear you, David. But 6K RAW for motion does exist. Timelapse.

I think the only way this will really be settled is by testing it and showing a side-by side, as you said. Even the Mysterium X 5K Epic might be able to go head to head with 65mm 5-perf.

Why not arrange a test. Rent a 65mm camera from Panavision, and shoot side-by-side with Epic X. Then project the film on a 70mm projector, and the Epic footage on the best 4K projector possible.

If I get to meet Fricke at this Lucas thing in February, I will ask him if he'd be game to help with such a test. Epic X vs 5-perf.

Edit: LOL, David you and I edit our posts so much that it's like having an ongoing conversation.. hehe.

 

[Peter Osinski]

You may not make people happy though with this "small screen Imax".

especially me

 

[Jeremy Torrie]

Huh. I'd love to have some Epic-X ready to go at the end of January but it ain't happening...we're doing an Imax test up in CHurchill on the Northern Lights for an IMAX feature that's already presold to a number of markets...we will be testing the Nikon, Canon, and Kodak is donating $12,000 worth of stock to do side by side for the timelapse elements...the film has to travel in nitrogen to get up there, but I would have loved to be able to line up my Epic-X and see how it all shakes down.

 

[Deanan]

Tom, just keep in mind the film images do benefit to some degree from the fact that the detail, i.e. grain, lies in a different position on each frame, so there is some increase in resolution when projecting versus looking at a still frame.

This effect is dependent on the grain size vs optical resolution. As you get to the larger film sizes where the grain size is smaller than the finest details resolved, a single still frame is about the same as frames in motion with the exception of the added perceived sharpness that grain adds.

 

[Jeff Kilgroe]

We also have the ability to add and manipulate noise and grain in post, whether the image was originally acquired on film or digitally. There are times where a little added grain goes a long way to give the appearance of more detail.

Also, printing images back to film introduces grain on that level so the only way to remain truly grain-free would be to have an all-digital pipeline from acquisition through projection or other delivery.

 

[David Mullen ASC]

You'd want to make comparisons to the same final display system or format, so a 5-perf 65mm comparison to a 6K RAW frame should either involve both being shown in 4K digital, or both in a 65mm print, or both blown-up to IMAX.

Though I'm not saying that it wouldn't be comparable, though 5-perf 65mm is not the same thing as 15-perf 65mm IMAX.

I find, practically-speaking, the 4K RAW RED ONE image to be on par with average Super-35 resolution and just below 35mm anamorphic resolution, especially in a film print. So 5K RAW would be better, and 6K RAW even better still, probably approaching 5-perf 65mm quality. But until a comparison test can be done and posted to the same format, it's an unanswerable question. I'm certainly not going to let the final word be an inaccurate comparison of a 65mm-to-1080P Blu-Ray copy of a movie to high-end digital stills. I find comparing still frames to be useful... but not definitive on what the results look like in motion on a large screen.

 

[Add Hawk]

Here is one of my scans of some 8perf slr film if it helps for some idea of detail, appologies in advance for the dirt on it, it was scanned just under 6k.

(http://www.flickr.com/photos/29501710@N02/3460934600/sizes/o/)

Oh it's reversal stock rated at 100asa for anybody that is interested.

 

[Tom Lowe]

You'd want to make comparisons to the same final display system or format, so a 5-perf 65mm comparison to a 6K RAW frame should either involve both being shown in 4K digital, or both in a 65mm print, or both blown-up to IMAX.

How about this. Side by side shoot with Panavision 65mm 5-perf and Epic X with Master Primes. 65mm scanned at 6K and finished at 4K. Epic X shot at 5K and finished at 4K. Both projected on the best 4K projector possible.