using benchmark2 (benchmark1, IMO, should only ever be used for historical comparisons, as it doesn't have skins or many particle effects at all)
//flame's DX9 propaganda config
//(modified slightly to turn off shadows)
41.143 seconds 117.71 fps ( 8.50 ms/f) 16.261 fps variability
41.407 seconds 116.96 fps ( 8.55 ms/f) 16.212 fps variability
41.033 seconds 118.03 fps ( 8.47 ms/f) 16.217 fps variability
40.897 seconds 118.42 fps ( 8.44 ms/f) 16.047 fps variability
40.843 seconds 118.58 fps ( 8.43 ms/f) 16.526 fps variability
//AVG:
41.065 seconds 117.94 fps ( 8.48 ms/f) 16.253 fps variability
//LQGFX
39.559 seconds 122.43 fps ( 8.17 ms/f) 17.366 fps variability
39.295 seconds 123.25 fps ( 8.11 ms/f) 16.993 fps variability
39.443 seconds 122.78 fps ( 8.14 ms/f) 17.540 fps variability
39.612 seconds 122.26 fps ( 8.18 ms/f) 17.312 fps variability
39.344 seconds 123.09 fps ( 8.12 ms/f) 17.045 fps variability
//AVG:
39.451 seconds 122.76 fps ( 8.14 ms/f) 17.251 fps variability
not too surprising in itself, but then I had a bit of an idea, to check if any of these framerate gains were specifically because of the small config flame made:
//AND NOW, THE ULTIMATE TEST!!!!!
//(I shoved flame's stuff to the end of LQGFX)
38.639 seconds 125.34 fps ( 7.98 ms/f) 18.882 fps variability
38.628 seconds 125.37 fps ( 7.98 ms/f) 18.782 fps variability
38.611 seconds 125.43 fps ( 7.97 ms/f) 19.056 fps variability
38.238 seconds 126.66 fps ( 7.90 ms/f) 18.483 fps variability
38.803 seconds 124.81 fps ( 8.01 ms/f) 18.446 fps variability
//AVG:
38.584 seconds 125.52 fps ( 7.97 ms/f) 18.730 fps variability
...and I actually get a better framerate, albeit at the cost of slightly worse fps variability. how strange