Add 16f option to render_bit_depth command

This change adds an option to let sway render onto XBGR16161616F
buffers, assuming hardware and drivers support this. Such half-float
formats may be used in the future for applications (including nested
Sway) to submit high-dynamic-range buffers.

Using this option is even more likely to reveal bugs in graphics
drivers, monitors, and screenshot tools than 'render_bit_depth 10'.
This commit is contained in:
Manuel Stoeckl 2022-07-20 20:11:37 -04:00
parent 6b97c4fa71
commit 1351c4e505
4 changed files with 13 additions and 3 deletions

View file

@ -165,9 +165,10 @@ must be separated by one space. For example:
adaptive sync can improve latency, but can cause flickering on some
hardware.
*output* <name> render_bit_depth 8|10
*output* <name> render_bit_depth 8|10|16f
Controls the color channel bit depth at which frames are rendered; the
default is currently 8 bits per channel.
default is currently 8 bits per channel. The "16f" option requests
color channels encoded in half-precision floating point format.
Setting higher values will not have an effect if hardware and software lack
support for such bit depths. Successfully increasing the render bit depth