Great, thanks Joseph. Btw, Arri Meta Extract seems to generate the .aml files, right? Or is the command line tool doing something different?
On Fri, Oct 28, 2016 at 12:34 PM, Joseph Goldstone <jgoldst...@arri.com> wrote: > > On Oct 28, 2016, at 12:14 PM, Gary Jaeger <g...@corestudio.com> wrote: > > That’s great info, thanks Joseph. I’ll ask the DP if any in camera grading > was done. And just so I’m clear, if they *were* done we could extract, > convert and use in nuke, correct? > > > You could, though at the present time, though the extraction can be done > with a command-line tool, the conversion from ARRI 3D LUT format to a > Nuke-ready format requires use of a GUI tool. Not so great if your DP used > a different in-camera grade on 2,500 effects shots. But mostly we find that > people develop a handful of looks in pre-production (“daylight exterior”, > “daylight interior”, “evening exterior”, “evening interior”, “flashback”) > and then select one of that menu of a handful for all the shots in a > sequence. Or just use one for the entire production. > > > And is that where we’d load the 3D LUT into an ocio file transform > downstream of the read but ahead of any FX work? > > > I wouldn’t think that you would be using the LUT until after you’d added > the VFX: > > 1. File comes in LogC. > 2. You extract the .aml with the command-line tool and use the GUI > tool to convert it to, say, IRIDAS .cube. > 3. Read the LogC file in and linearize to your working space (e.g. if > you are using ACES, linearize the LogC to ACEScg using the LogC IDT) > 4. bring in your dinosaur from your dinosaur’s encoding space to the > working space, and comp in working space > 5. transform the comp from working space to LogC (inverse of (3) > above). > 6. apply .cube file from (2) via OCIO FileTransform to get displayable > values. > > I don’t want to give you specific parameter settings since I’m using a > prototype ACES 1.0.3 OCIO config and am using Nuke 10.0v4. I think the > earliest Nuke 10.0 + ACES OCIO config support may have had some issues > (though HPD is the authority on this) and anyway, researchers shouldn’t > tell production people what to do. (And indeed you may not be working in > ACES at all, and I’ve totally forgotten what Nuke is like pre-ACES, which > is why 1-6 above are a bit vague.) > > Best— > —joseph > > > *Gary Jaeger */ 650.728.7957 direct / 415.518.1419 mobile > http://corestudio.com > > On Oct 28, 2016, at 11:58 AM, Joseph Goldstone <jgoldst...@arri.com> > wrote: > > The LogC is the same, and you can use a “standard LogC to Rec709 LUT” from > the ARRI website to match the traditional ALEXA look. > > Perhaps that’s enough of an answer for you right now. But for those of you > that are interested in the general question of LogC from the AMIRA… > > …there is a much more flexible color processing architecture built into > the AMIRA, the ALEXA MINI and the ALEXA SXT. In the AMIRA, the user has the > option of either extensively tweaking the standard processing, or loading > in a 3D LUT generated in another package (say, Resolve of Baselight) and > this 3D LUT will be carried in the .mov file as metadata. ASC CDL is also > applied differently to LogC data for those more recent cameras — bowing to > Hollywood tradition, on the AMIRA, MINI and SXT it’s the LogC data that > gets modified by the ASC CDL values, not the display-ready video values (as > is the case with older ALEXA models). > > There is not currently a command-line tool to extract these 3D LUTs > directly in a form that Nuke understands. There is a command to extract the > 3D LUT as a .aml file, and there is a GUI tool (the ARRI Color Tool) that > can be used to convert the .aml file into a variety of 3D LUT formats that > Nuke understands, but it’s not scriptable. > > To download the ARRI Color Tool, or to read more about it, you can look > here: > http://www.arri.com/camera/alexa/tools/arri_color_tool/ > > To learn more about metadata extraction (including pulling out the look > file) you can look here: > http://www.arri.com/camera/alexa/tools/arri_meta_extract/ > > I haven’t seen many features use this flexibility to date; many of them > are grading downstream and as Michael points out, LogC is LogC wherever. > But as you are on an AMIRA show, you should probably check to see if any > in-camera grading was done or any custom looks were loaded. > > Best— > —joseph > > On Oct 28, 2016, at 10:46 AM, Michael Garrett <michaeld...@gmail.com> > wrote: > > Hey Gary, > > Amira LogC is the same isn't it? It's possible Premiere is applying a LogC > to Rec709 LUT to roll the highlights off and get it looking presentable. I > would imagine from what you're saying that the footage looks blown out in > Nuke because it's not doing the additional s-curve in the above LUT to get > a reasonable image. Check out the Arri web site to download a LogC to > Rec709 LUT. > > Cheers, > Michael > > On 28 October 2016 at 13:02, Gary Jaeger <g...@corestudio.com> wrote: > >> I have a question about LogC as well. Do we need a different LUT to work >> with Amira LogC footage? I’m pretty sure the camera tags the clips, and for >> instance Premiere reads that and displays the footage with that lut so the >> editor (and client) see something reasonable i.e. not flat LogC like it is >> looking at the raw clips. But Nuke displays the footage quite blown out. I >> know the data is there, but is the AlexaV3LogC not the right transform? >> >> >> *Gary Jaeger */ 650.728.7957 direct / 415.518.1419 mobile >> http://corestudio.com >> >> On Oct 26, 2016, at 7:07 AM, Stepan Z <motionarti...@gmail.com> wrote: >> >> Hello i didn't think about the bitrate! Thank you Andrew! I'll read about >> the compare node! >> >> All the best >> >> Stepan >> >> >> >> >> On 25 Oct 2016, at 22:19, Andrew Mumford <a_mumf...@mac.com> wrote: >> >> Are your dpx's 10 bit ? - That would make it different for sure ! >> >> There's also a "hidden" but wonderful node called"Compare" that is great >> for checking these things - gives you a visual and error based output. >> >> --- >> Andrew Mumford >> >> >> On Oct 25, 2016, at 09:56 AM, Stepan Z <motionarti...@gmail.com> wrote: >> >> Hello >> >> I'm using nuke studio to conform a few short edits, publish a source dpx >> sequence and a nuke script. >> >> So i'm bringing in original alexa logc encoded prores files, and >> publishing dpx sequence with logC set as the colorspace in the export >> dialog. When i then bring over that dpx sequence back into NS and drop it >> on top of the prores and disable all colour transforms (viewer to none >> instead of sRGB and the per file transform to linear) perceptually the >> files look identical. But when you sample a pixel or an area with the >> viewer the rgb values after two decimal places seem to change when your >> flicking between dpx and prores. >> >> Is this normal and is just the difference in encodings or should they be >> exactly the same given that they come from the same file and are in the >> same colourspace? >> >> >> >> >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> > > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, https://protect-us. > mimecast.com/s/arGYBbiNp60UG?domain=forums.thefoundry.co.uk > <http://forums.thefoundry.co.uk/> > https://protect-us.mimecast.com/s/27lmBXIbrzlTx?domain= > support.thefoundry.co.uk > <http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users> > > > > > ------------------------------ > This email has been scanned for email related threats and delivered safely > by Mimecast. > For more information please visit http://www.mimecast.com > ------------------------------ > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, https://protect-us.mimecast. > com/s/K2R4BwcEbNOc1?domain=forums.thefoundry.co.uk > <http://forums.thefoundry.co.uk/> > https://protect-us.mimecast.com/s/enY4BbCGbV7ik?domain= > support.thefoundry.co.uk > <http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users> > > > > > ------------------------------ > This email has been scanned for email related threats and delivered safely > by Mimecast. > For more information please visit http://www.mimecast.com > ------------------------------ > > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > -- Gary Jaeger // Core Studio 249 Princeton Avenue Half Moon Bay, CA 94019 650.728.7957 (direct) • 650.728.7060 (main) http://corestudio.com
_______________________________________________ Nuke-users mailing list Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
