questions regarding overexposure in beauty-canvas (almost only white)

chrom

Hi,
I have following questions regarding overexposure in beauty-canvas:
1.) How can you avoid overexposure in beauty-canvas?

2.) How can you detect it in DAZ before render (histogram in DAZ, visible in viewport, ...?)?

3.) Why are non-canvas render (render window) fine but beauty-canvas almost white?
It looks like there are some auto detection and auto processing. Can it be simple disabled to have in worst case a short render for overexposure detection (only necessary if there is no answer for 2.)?

What I only find in forum are some open discussions. Perhaps I missed the right one.
I know you can readjust it in postprocessing but not in all programs and with almost white beauty-canvas you lost mass of quality because of using only few bits of the available bits. So the adjustment should be done in DAZ before rendering.

Thanks & wish you a fine weekend

 

Sevrin

If you are not planning to do any post work, then you should render your scenes without canvasses and with tone mapping on.  EXR files you get from Iray aren't ready-to-view, and they aren't meant to be.

Canvasses are high dynamic range EXR files meant to be used only compositing, whether in Nuke, Ps, The Gimp or some other raster image editing software suitable for editing high dynamic range images in layers.  In fact, far from losing details or quality, what you get is a 32-bit image containing more information than the standard 8-bit images resulting from standard rendering.  The details in the unaltered image are all there, but they fall outside the dynamic range of the human eye and beyond the capability of your monitor to display unless you go out and buy a monitor designed for HDR output.

You have a nice weekend, too.

 

chrom

I want to use it in a photo software or fusion 9.

But instead of readust it in a photo software because it seems to be totally overexposed I would like to render it not overexposed to use more of the value range. Somewhere in forum I read the hint to use exr for best quality regarding its value range.

When I look at a histogram before / after post processing an overexposed exr I would assume it is not a problem of my eyes or monitor but an "incorrect usage" of dynamic range space.

Perhaps I misunderstand you.

Matt_Castle

As EXR format stores the values in 32-bit floating point, there isn't really much concept of under or overexposure; the format can store a dynamic range of 255 exposure steps, enough to store brightnesses up to (roughly) a billion billion billion billion times brighter or dimmer than 8 bpc format can support. (In practical terms, it uses a binary form of scientific notation).

Yes, with default settings, the canvases are about 13 or 14 exposure steps too bright to correctly display on most monitors, but this is a trifle to what the format can actually store; The format will never clip under any normal circumstances.

Sevrin

chrom said:

I want to use it in a photo software or fusion 9.

But instead of readust it in a photo software because it seems to be totally overexposed I would like to render it not overexposed to use more of the value range. Somewhere in forum I read the hint to use exr for best quality regarding its value range.

When I look at a histogram before / after post processing an overexposed exr I would assume it is not a problem of my eyes or monitor but an "incorrect usage" of dynamic range space.

Perhaps I misunderstand you.

When you render using canvasses, you get to save the image, let's say "Picturename.png" in your Render Library.  You also get a folder in the same directory as your saved PNG or whatever called "Picturename_canvases" containing the actual canvasses in EXR format.  The Picture.png is mostly just something DS will save whether you want it or not.  You can't save your canvasses without saving it, so give it a name, save it and then discard it.  The actual images to do your compositing with are the EXR files in the Picturename_canvases folder.   They will look just as overexposed, but they contain all the information of your image, and you just need to adjust exposure to reveal it.