View Full Version : 32-bit floating point


Marius Boruch
March 9th, 2008, 08:40 PM
what is it and when to use it???

Bill Ravens
March 9th, 2008, 08:58 PM
the devil incarnate

Glenn Chan
March 9th, 2008, 10:03 PM
It:

1- Allows you to render with higher bit depth. This reduces rounding error during rendering. When passing the values between filters, 32-bit values are passed instead of 8-bit values.

It also lets you do linear light compositing (though some of those things are possible with 8-bit):
http://glennchan.info/articles/vegas/linlight/linlight.htm

If you have banding issues with your renders, rendering in 32-bit might solve it.

2- There are some gotchas with 32-bit. Some codecs behave differently in 32-bit mode versus 8-bit mode.

http://glennchan.info/articles/vegas/v8color/v8color.htm