I have a couple of questions for developers:
1. Is there any simple formula to recalculate bump map difference when switching from CPU to GPU in the middle of a work? I have two 2080 RTX and I wanted to test them, but bump looks more pronounced after switching to GPU.
2. When turning "resize textures on demand" option is set to "on demand mip-mapped" then what alghoritm is used to determine which map is to be original 16384x16384 and which will be reduced to say, 512x512. It is about camera FOV, look angle, or blur level from DOF?
3. Would it be possible to increase maximum resize GPU textures to 4096 or even 8192? Current 2048 is too small for closeups of the whole-body garnment and this forces me to use cpu for the sake of 16k bitmaps only
I hope these are very specific questions that can be adressed in a concrete way. Thanks for RTX support, by the way,
nl
nl![Click image for larger version Name: old clothing.jpg Views: 1 Size: 508.8 KB ID: 1056757](filedata/fetch?id=1056757&d=1605289174)
![Click image for larger version Name: new clothing.jpg Views: 1 Size: 507.0 KB ID: 1056758](filedata/fetch?id=1056758&d=1688719450)
1. Is there any simple formula to recalculate bump map difference when switching from CPU to GPU in the middle of a work? I have two 2080 RTX and I wanted to test them, but bump looks more pronounced after switching to GPU.
2. When turning "resize textures on demand" option is set to "on demand mip-mapped" then what alghoritm is used to determine which map is to be original 16384x16384 and which will be reduced to say, 512x512. It is about camera FOV, look angle, or blur level from DOF?
3. Would it be possible to increase maximum resize GPU textures to 4096 or even 8192? Current 2048 is too small for closeups of the whole-body garnment and this forces me to use cpu for the sake of 16k bitmaps only
I hope these are very specific questions that can be adressed in a concrete way. Thanks for RTX support, by the way,
nl
nl
Comment