Retaining latitude and image quality from capture to delivery is key to VFX pipelines. Currently restricting to 8bits is limiting and making workflow a hassle, where it could be streamlined within RC. Is there any plan to support higher bit depth images? OR is there already support and we’re missing something? :?
Currently using Nikon D800 with 14bit uncompressed raw files. Transcoded to 16bit tiffs.
Thanks,
Joe.
Hi Joseph,
you are right, currently it is just 8bpc but internally we are ready for float per channel. It is in short plan to deliver this, but I do not have precise dates on that so far.
I’m raising priority. Everybody interested in this feature please add some comment. Thank you.
yes… + 1000
It’s a must . 
Hi Tomasb,
Thank you for your reply. Its great news that this is in the pipeline.
Let us know If you need beta testers! 
Greed if your importing and using tiff then exporting textures the same quality could be usfull
I’d like to see 16-bit tif too.
but having support for floating point exr dose sound interesting.
though i would assume bringing in pre-made hdr files would probably have issues with image sharpness and wouldn’t stitch very well.
Hi,
This feature would be an absolute must-have for us. Please let us know if you need beta testers.
Thanks!
I will join the list of those very interested in 16-bit. Agree it’s a “must have”.
Is there any update from the Capturing Reality team on this one?
This would be great, definitely interested!
A must have for the tillable textures creation and delighting process in photoshop.
I would love this. Currently We have to export all of the cameras from RC and import them into Agisoft Photoscan to bake down textures.
+1!
erikjameschristensen wrote:
I would love this. Currently We have to export all of the cameras from RC and import them into Agisoft Photoscan to bake down textures.
hello… I also need 16 bits textures… could you explain your workflow…
thank you
i can’t tell you how much +1 this gets from me
yes! 16bit support please…
Watch out with the terminology you use. There is a huge difference between 16bit and 16float. 16bit is integer. You want 16 and 32float support for color images. A Nikon D800 might have 14bit raw files but all of that information in the raw is not transcoded to 16bit tiff. The debayering process from Nikon is integers and not floating point. Nikon does not give out source code for the raw files either. Your best option for a 16float image would be to take multiple exposures then create an exr file from that which will be 16float color in linear colorspace.
What you really need in Capture Reality is to solve with an 8bit tonemapped jpg where the source of that is from a 16float exr. After the mesh is computed you would want to replace color only in RC with the 16float exr source files. Then have RC allow export of the pointcloud in a format that will keep the color data 16float with linear gamma.
They reason why I say solve with a lossless JPG is so you dont actually have RC load exr files for solving. The amount of Ram usage will be huge.
Joseph Steel wrote:
Retaining latitude and image quality from capture to delivery is key to VFX pipelines. Currently restricting to 8bits is limiting and making workflow a hassle, where it could be streamlined within RC. Is there any plan to support higher bit depth images? OR is there already support and we’re missing something? :?
Currently using Nikon D800 with 14bit uncompressed raw files. Transcoded to 16bit tiffs.
Thanks,
Joe.
Any update on this?
Thanks
Erik
Yes! Very much needing this for a true pipeline solution.
+1 for exr texture exports
Just adding my voice to this thread
Would be awesome to have 16bits textures export feature.