Default Render Options

I’ve merged in several of @perminder-17 new shaders, including fog (fast) and depth-of-field blur.

There’s been some questions that the ambient occlusion and other new rendering features slow down rendering, particularly on integrated graphics and older hardware.

So … what should be the default?

  • turn on new rendering & let users turn off as needed
  • default = fast & let users turn on as desired
0 voters

I figure some discussion is warranted here, so I’ll throw in my two cents.

My current visualization preferences are specifically because of my use cases, which is primarily visualizing and then putting graphics into a powerpoint or LaTeX document. Because of this, I usually need something where getting a transparent background is as easy as Win+Shift+S, take a screenshot, and then paste into a slide. After that I can just simply set any black color as transparent, and everything slides into place very quickly.

To make this process smoother, my current visual settings are the following:

  1. Ambient Occlusion - Disabled
  2. Fog - Disabled
  3. Edge Detection - Disabled
  4. Background Color - Black
  5. Projection - Orthographic (This is personal preference, not ease of use)

I think that a nice feature for Av2 would be a way to copy an exported image instead of having to save it, which would reduce disk usage and time spent faffing about with image importing. I’d also say that a good QOL feature would be changing the way graphics are exported, as the current method includes the entire field of view, with the reference axes. I would instead have some sort of way to have it export only the molecule in frame, with some padding on the margins to make it fit well.

Once I finish my quest to learn enough C++ to contribute and fix up the molecular properties dialogue, I may take some time working on this (along with the thousand of other things I desperately want to help fix but can’t because of my inexperience in C++).

Edit ⇒ Copy Graphics

You can, of course, hide the reference axes. The code currently just saves the framebuffer. What you’re describing sounds like a bunch of work. The code would have to disable the reference axes (and what if the user wants the axes in there?) figure out some sort of cropping / padding, etc.

Let me see if I can help by creating a branch for you.

Oh, I should mention, that my default had been to use a transparent background color in OpenGL. That way it’s easier to copy / paste into a slide, document, etc.

It turns out a lot of graphics drivers didn’t like this, so they made the whole window transparent. :man_shrugging: