REDWideGamutRGB and Log3G10

[Graeme Nattress]

Yeah, that process is what I've been using with the output LUT. But I'm a bit confused about the preview builds workings. Is it supposed to calculate for a color space automatically? Because I still seem to need an output LUT in order to see a proper image.
Also, did I miss some link to a IPP2 Preview cube? Is that the latest in the Cube development we were talking about earlier?

In REDCine-X IPP2 Preview build you'd select colour space: Rec709 gamma space: BT1886 to get it to do the calculations for a proper output image. That will be the equivalent of doing a raw development to RWG/Log3G10 and then applying the cube lut, but done as a direct process rather.

When you fire up the REDCine-X IPP2 preview build, you'll get those as the default settings.

Graeme

 

[David Battistella]

This is great. Nice to see it implements Graeme. congratulations!

David

 

[Graeme Nattress]

This is great. Nice to see it implements Graeme. congratulations!

David

Thanks David. First step out of the door and now heads down for the full-on implementation!

Graeme

 

[Gavin Greenwalt]

Why bt1886 by default when software and hardware adoption is nearly nonexistent?

 

[Graeme Nattress]

BT1886 is the standardized EOTF of a HD monitor, and what any HD monitor should be calibrated to. "Reference electro-optical transfer function for flat panel displays used in HDTV studio production". I don't think there's anything more appropriate to use.

Graeme

 

[Ryan Glover]

Anyone have a link to just the preview lut so I can try it out on some RED One MX footage in Resolve? I looked through the threads but didn't see one -- thanks in advance!

 

[Alex Stolpakov]

I second that ---How to get the lut (widegamutRGB Log3G10) for use in Resolve

 

[Björn Benckert]

How does it work with playback in camera?
Did some quick tests today and playback looks different from capture or preview, correct?

 

[Christoffer Glans]

In REDCine-X IPP2 Preview build you'd select colour space: Rec709 gamma space: BT1886 to get it to do the calculations for a proper output image. That will be the equivalent of doing a raw development to RWG/Log3G10 and then applying the cube lut, but done as a direct process rather.

When you fire up the REDCine-X IPP2 preview build, you'll get those as the default settings.

Graeme

Gottcha, though I was supposed to select LOG3G10 as gamma, which would require a LUT output or grade.

On top of all this, do you know if Blackmagic is working on implementing all of this in their color management for Resolve? We have support for RWG/LOG3G10 in there, but the timeline color space and other management settings do not. But maybe that's not needed with RWG/LOG3G10? If set to a timeline color space of REC709, then that will work according to the output LUT choosen, right? Say I have a RWG/LOG3G10 to 709 Lut and color management settings set to REC709 2.4 Gamma and my screen calibrated to 2.4, that should be everything needed for proper 709 grading outputs? If later I'm changing a grade to HDR and UHD standards, I just choose a corresponding color space within Resolves color management and change the output LUT to a RGW/LOG3G10 to "whatever HDR standard may be" together with a screen supporting it.

That would be it, right?

 

[Graeme Nattress]

How does it work with playback in camera?
Did some quick tests today and playback looks different from capture or preview, correct?

Camera playback does support the LUT.

Graeme

 

[Stacey Spears]

Why bt1886 by default when software and hardware adoption is nearly nonexistent?

BT.1886 is what Hollywood is using. Consumers are also calibrating their displays to BT.1886.

If you are not using the black offset, this basically means 2.4 gamma, which has been the unofficial standard for years as that is the gamma of a Sony BVM. It's never ever been 2.2, which people often incorrectly state. Though, there are some using 2.2. We are only in this situation because 709, and 2020, never defined the EOTF. They only defined the useless OETF. At least HDR defined the EOTF and inverse EOTF on day one. It's funny that we use inverse EOTF and not OETF.

 

[Graeme Nattress]

BT.1886 is what Hollywood is using. Consumers are also calibrating their displays to BT.1886.

If you are not using the black offset, this basically means 2.4 gamma, which has been the unofficial standard for years as that is the gamma of a Sony BVM. It's never ever been 2.2, which people often incorrectly state.

Thanks Stacey! It's also (again without the black offset) what ACES are using for their REC709 output.

Graeme

 

[Jacobo Martinez]

Graeme, I downloaded the PREVIEW RCX VERSION and when opened a Helium file in a mac book pro this is what i see. In the official RCX the frame looks fine, no problem.

 

[Jacobo Martinez]

I just figured out what it was, for some reason the file opened with the GAIN set to 2:

 

[Graeme Nattress]

Yeah, that'll do it. IPP2 does do "gain" - that functionality will be taken over as part of the grading block.

Graeme

 

[Phil Holland]

Something that is a bit vague at the moment, but is being hinted at if you look at these versions is a new and simpler way to look at, work with, and process REDCODE RAW. For instance FLUT is MIA in the preview build. In fact, there is an overall simplifying effort of Color Space and Gamma that I'm sure we'll hear a lot about shortly.

 

[Graeme Nattress]

Yes Phil! One of the goals of IPP2 was to simplify things. Although there's a lot of extra power in the full IPP2, it would be no good if it was also too complex to use so rather than duplicate everything the original processing did, I actively worked to strip out unnecessary complexity. Gains are gone to be replaced in functionality by the grading section. FLUT, exposure and brightness are gone to be replaced by a single "exposure adjust" control. I don't think there's much you can't do in IPP2 that you could do before, but the name or the place where it's done will be a little different.

Other simplifications include use of industry standard colour spaces and output gamma functions where applicable, so you know that if you see something that looks like a proper spec, what you get actually adheres to that spec and works just as you'd expect.

More of this will become aparrent as we work towards the full IPP2 implementation, and I'll be here to help you all get the best out of it.

Graeme

 

[Jacobo Martinez]

It would be really helpful if all the new Ipp2 would be showcased in a video.

 

[Graeme Nattress]

It would be really helpful if all the new Ipp2 would be showcased in a video.

One of the great things about this preview release is that it helps us know what kinds of questions to cover and how to explain things to users.

Graeme

 

[Björn Benckert]

Been playing with this today.

Think it works very well with the new firmware and the IPP2 stuff. Love it. Here are my findings, cats at high iso and odd light temperatures... thats what happens when trying to be social at home while being eager to test new camera stuff

Cat @ 12800ISO developed with Graemes preview lut.

[/url]PastedGraphic-3 2 by Björn Benckert, on Flickr[/IMG]

Another one with more underexpo.

[/url]PastedGraphic-2 by Björn Benckert, on Flickr[/IMG]

And a third cat same low iso on all of them.

[/url]A001_C007_1218JA by Björn Benckert, on Flickr[/IMG]

Then here is some motion stuff. first just the RWGRGB + Log3G10 and nothing else, simply this is how my proxies looked. I shot them at 3400 ISO flut1 (which even if they where only 2k 422 was really nice to grade from together with the preview lut or other sauce.

https://vimeo.com/196174128

Then this one is rendered from the 8k r3d´s with the preview lut applied in redcinexpro at 6400ISO.

https://vimeo.com/196192729

And then this one REDWIDEGAMUT+LOG3G10 logexpocor + kodak sauce 3D lut from the proxies @ iso 6400:

https://vimeo.com/196203079

And then yet another one with a bit more punch in the expocorr to see how the highlights curved off.

https://vimeo.com/196198538

Nothing scientific in any of this but love to see how the colors behave even though it´s such low light levels and a mix of all kinds of light, my iPhone tungsten and some cold spotlights and even when iso is brought up to nuts levels color stay in check very well. Would love to try this stuff on an old redone mx just to see how it behaves.