Are we approaching the limit of what you can actually see on screen? What next?

[Roberto Lequeux]

I agree with you David. I think he might have been talking about where he would personally have focused RED's R&D resources at the time when that video was shot. I think it is pre M-X, if I am not mistaken. Besides it is a pretty safe assumption that he really appreciates RED's high MFT at 2k, which comes from a properly down-converted 4k image --especially knowing he was considering shooting portions of Che on 65mm film. This is my point. Red One Mysterium is a camera that can deliver footage that feels a bit like 65mm in terms of detail. Now RED makes a camera that can deliver even higher MTF than 65mm, without any additional costs.

It is still up to the artists to know when and how to use resolution, as well as all the other layers of the medium. There is no excuse when you can drop a filter, apply additional softening in post, use soft lenses... etc. What matters most is the people behind, not the camera; and oddly enough that is exactly why RED's versatility is so important to me. RED gets that, evidenced by 120fps 5k, 300fps 2k, heaps of single-exposure DR, HDRx, tracking from HDRx, size, weight... cost!

Thank you RED. Can't say it enough. And thank you Steven for motivating RED so much (more) with "Che".

[Graeme Nattress]

If you have a very good dynamic range, you can have bright saturated colours. If you don't alter the primaries of the system and have a low dynamic range scene then that scene will get darker, or you'll be sending those bright saturated colours towards white. That is why dynamic range will show you a larger colour gamut in usefully visible ways than altering the primaries. The primaries of 709 are not the best that can be achieved today, but I think DR is still more the limiting factor...

Graeme

[Bruce Allen]

If you have a very good dynamic range, you can have bright saturated colours. If you don't alter the primaries of the system and have a low dynamic range scene then that scene will get darker, or you'll be sending those bright saturated colours towards white. That is why dynamic range will show you a larger colour gamut in usefully visible ways than altering the primaries. The primaries of 709 are not the best that can be achieved today, but I think DR is still more the limiting factor...

Graeme, 709's red primary looks like a brick-like color.

If you make it very bright, you will just get the impression of very brightly-lit bricks. Even if they are against inky black.

I am sitting here with my DreamColor monitor - the saturated reds you can get out of it are just stunning if you have it in full gamut mode. Good but not as great in P3 mode. And in 709 mode... just sad. No amount of luminance contrast can change this.

I sincerely hope that RedRay isn't just 4K with 709 colors. Likewise, that Red's 4K projection system isn't limited in color gamut to 709. That would be a real shame. Please give us a P3 mode or a beyond-P3 mode for both RedRay and projector. If there could be some kind of token sent from the player to the display to tell it "hey, shift into P3 mode, you have a real movie to show" that would be ideal.

If you build the most wonderful laser projection system on earth, great. Just please let us author content for your beautiful laser projector that takes full advantage of the colors it can create!

Sorry if I am making assumptions.

Bruce Allen
www.boacinema.com

[Blair S. Paulsen]

I would love to see packaged media like Blu-Ray (and formats to come), as well as downloadable media files adopt the P3 color space from DCI as the modern standard. For backwards compatibility a rec 709 LUT would be a necessary evil, but with the right flags that would only be applied when the display device required it. Sony, Panasonic and others have been offering "wide gamut" variants in their displays for several years now but as noted by others virtually all material authored for the home is in rec 709 so the rubber isn't really hitting the road.

I trust that Graeme understands the impact of higher DR on color rendition better than I do. Other than the Dolby reference display which claims to offer 11 stops of DR I am not seeing much industry movement toward genuinely higher dynamic range, contrast ratios or maximum brightness.

While I am all geeked up over the Sony 4K VPL1000ES consumer projector I hardly think it's specced 2,000 lumen output will offer anything close to the DR of Jim's Sony T-420 SXRD's 22,000 lumen rig at Red Studios which I consider a 10 stop device. Would love to hear from Pat at IGI or others what they consider the real world DR of the roughly $200K T-420. Anyone have a DR spec for the big Barco they have been using for the 4K Fest - Justin?

As an analogy many believe that a camera as promising and relatively affordable as the F65 would have been a 2017 product if not for RED. Somebody needs to disrupt the agonizingly slow advancement on the display side - too bad Jim and the Red Team are so busy making cameras... heh heh heh

Cheers - #19

[Blair S. Paulsen]

(snip) I am sitting here with my DreamColor monitor - the saturated reds you can get out of it are just stunning if you have it in full gamut mode. Good but not as great in P3 mode. And in 709 mode... just sad. No amount of luminance contrast can change this.
Bruce Allen
www.boacinema.com

I have seen exactly what Bruce is talking about on my DreamColor and have been roundly trashing rec 709 to anyone who'll listen for the last couple of years. Interlace is a kludge that lasted far too long, let's hope rec 709 doesn't follow suit. Is there an existing SMPTE spec we can move to? Can P3 spread outside of DCI land? Calling Mike Most...

Cheers - #19

[Graeme Nattress]

REC709 was a cludge on it's primaries to begin with, but at least monitors can display it without emulation. The dreamcolor can't do p3 and has to emulate, so even though it's native gamut is wide, it doesn't fully enclose p3.

Graeme

[Eric Rodemann]

I don't know what the exact limit is, but I"m convinced that there is a limit beyond which 99% of the general movie going public cannot discern a difference in quality, and the 1% is probably lying when they say they can.

It is not a matter of diminishing returns, but a matter of human limitation. There is a range above which the human ear cannot hear sound. It makes no sense to design a stereo speaker that can produce sounds outside the human range of hearing. You could say this speaker is the best, because it can go past 20KHZ! So what? No one can hear past 20KHZ. How does that make it better, when you cannot perceive the higher range? Once you construct a speaker that can produce 20khz, you're done. Higher khz is not an improvement. You can work on size, projection, depth, use of electricity, but making it 30khz won't do anything for you. And listening to a symphony with even 20khz sounds may not be pleasing anyway.

Vision is no different. There is a limit to human sight. We don't have microscopes for eyes. Hitting that limit is not a negative. It's a good thing. Once you reach that limit, you can focus on other features.

Do you need 90K? I'm fairly certain you don't. I think we're already getting close to what the limit is, when it comes to producing a pleasing image.

[Blair S. Paulsen]

Whatever percentage of the P3 spec the Dreamcolor can reproduce and whatever emulation it has to do, it still offers an opportunity to see the pathetic colorimetry of rec 709.

In any case, I am most interested in the next standard, be it P3 or something else, that offers greater fidelity.

In theory the IIF/ACES standard will offer us the chance to master to a color space beyond the discrimination level of most humans. Perhaps output transforms matched to current and future displays will allow us to see as much of that gamut as the device can put out without the necessity of a, by definition, limited standard.

Cheers - #19

[Graeme Nattress]

Thing is Blair, you could take the DCI route and go straight for XYZ, and you've been able to do that for quite a while.

Graeme

[Bruce Allen]

Thing is Blair, you could take the DCI route and go straight for XYZ, and you've been able to do that for quite a while.

Graeme

You can't do that at home though. You can only go to 709. That's the whole problem.

I think that Blair and I would just like to politely ask RED - eg the guys who have said they are making RedRay and 4K projectors - to consider either:
- supporting P3
- supporting xvYCC (it's a kludge... but it is enabled in displays now, we just need RedRay to be able to output xvYCC... and it's compatible with 709, though with some color clipping)
- support the creation a new color space standard if you don't like P3 (call it 809)
- doing something fancy (like XYZ or ACES combined with clever device-specific output transforms - seems hard to me to get something that automatically looks good without clipping or odd colors, but awesome if you could pull it off)

There is Ethernet on HDMI now. How hard can it be to communicate between RedRay and a display device and find out if the display device supports the wider color gamut standard? If it does, awesome - display the original content and signal the display device to go into wider color gamut mode. If it doesn't, do a transform to 709 with as little clipping as possible.

I'm sure the other display manufacturers want the same thing, because all of their current devices are hamstrung by the current 709 standard. Sony have wide gamut projectors, wide-gamut LCD screens, all of those OLED monitors... Mitsubishi have super wide gamut Laser projection RPTVs, etc.

I don't see why you guys are fighting Sony and Panasonic about active vs passive shutter when instead you could be working together on a better color standard that would improve things for everyone.

Bruce Allen
www.boacinema.com