cai567890 / pcsx2

Automatically exported from code.google.com/p/pcsx2
1 stars 0 forks source link

[Feature request] FPS/performance graphs #137

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
AFAIK, that was discussed a long time ago but I duuno what decision was 
made >_>
Anyways, here is the idea: made a graphs representing current PCSX2 
performance. Quick example: FPS counter graph.

Visualized example (ugly): http://img23.imageshack.us/img23/2184/
fpsgraph.png

I'm not sure if wxWidgets allows that but I hope >_>

Or maybe such feature can be implemented in other way: just dump some 
parsable data which can be used by 3d party apps.

Original issue reported on code.google.com by zantezuken@gmail.com on 5 Apr 2009 at 11:57

GoogleCodeExporter commented 9 years ago
Oh, almost  forgot to mention that the main idea is to make it as built-in 
feature. 
Also, FRAPS (and similar analogs too) sucks, his data can be incorrect 
sometimes 
(which is the main reason why it is not used in many game tests).

Original comment by zantezuken@gmail.com on 6 Apr 2009 at 12:02

GoogleCodeExporter commented 9 years ago
for the data I'd suggest a file for every run which saves the settings you had 
and 
time and frames 
also an external or integrated program to see and compare runs to each other 
with
legend to see which line has what kind of configuration xD

Original comment by kaboo...@googlemail.com on 6 Apr 2009 at 2:36

GoogleCodeExporter commented 9 years ago
And if I may ask... what the use of this? I mean, well.. yeah you get a nice 
graph and?

Original comment by kraka...@gmail.com on 8 Apr 2009 at 8:27

GoogleCodeExporter commented 9 years ago
Good statistical (and standartized in PCSX2 env.) information for comprasion 
(not 
only between games but revisions as well)?

Original comment by zantezuken@gmail.com on 8 Apr 2009 at 10:13

GoogleCodeExporter commented 9 years ago
Fun, but generally not practical.  Benchmarking Pcsx2 optimizations is actually
pretty easy since as developers we usually know what to look for and when.  
That is,
the code being run at specific points in games is both traceable and usually
repetitive, and so it's easy to do benchmarking of specific optimizations.

The one advantage I could see here is if the benchmark supports disabling the GS
entirely, for pure benchmarking of an EE or VU component without GS/GPU 
bottlenecks
and such.  For that, and since graphs *are* fun, I'm placing this on the 
'backburner'
for later consideration.

Original comment by Jake.Stine on 8 Apr 2009 at 12:07

GoogleCodeExporter commented 9 years ago
Yeah, that is what exactly I'm talking about.
Backburner, indeed :P

Original comment by zantezuken@gmail.com on 8 Apr 2009 at 1:44