Houdini can't output 16bit EXR/PNGs?!

[Neikke]

hey guys, is there a way to output heightfield as a single grayscale channel exr or png in 16bit? I tried multiple variations in settings but no matter what I choose (RGBA, Single Channel, 8bit, 16bit, 32bit) for EXR it always gives me RGB 32bit and puts god damned height into alpha which is not convenient at all. For PNGs no matter what settings I have I can output only 8bit.. Is there some workaround for that?

[haggi]

You mean that photoshop is unable to import the file as it is. The files are 16bit:

Quote

file format version: 2, flags 0x0
capDate (type string): "2023:04:26 14:45:11"
channels (type chlist):
    A, 16-bit floating-point, sampling 1 1
comment (type string): "none"
comments (type string): "none"
compression (type compression): zip, individual scanlines
dataWindow (type box2i): (0 0) - (499 499)
displayWindow (type box2i): (0 0) - (499 499)
lineOrder (type lineOrder): increasing y
order (type string): "A"
pixelAspectRatio (type float): 1
screenWindowCenter (type v2f): (0 0)
screenWindowWidth (type float): 1
type (type string): "scanlineimage"

but photoshop is unable to handle them as they are maybe because the single channel is called "A" in the exr file what is usually uses for an alpha channel.

And the png files are 16bit as well:

C:\daten\test.png :  500 x  500, 1 channel, uint16 png

Oh, and this happens if I open the png file in photoshop:

 

Edited April 26, 2023 by haggi

[Neikke]

Hm Photoshop can read 16bit PNGs without any problem.. I still think there's an issue inside of Houdini as 32bit Floating point PNGs are way less in size than 16bit or even 8 bit ones - I think this is a clear indication that there's something wrong going on with conversion. 

[Atom]

I don't think PNG files support floating point values. 16b Fixed works for me, though.