Hi all! I''m trying to measure the frame rate of my realtime application and to also scale movement according to it, so it runs at the same speed on all processors. I found something here: http://www.gamedev.net/reference/programming/game/article753.asp At the bottom of the page there are some simple sample programs to try. There''s one DJGPP version and one for VC. However, I''m using Cygwin, and I don''t know if that''s the same as DJGPP. -The precompiled version of the DJGPP sample achieves an incredible >13000 frames per second -The precompiled version of the VC sample achieves a pretty good >~1300 frames per second -The REcompiled and modified version of the DJGPP sample achieves a pretty bad ~65 frames per second - I modified it to use the Cygwin equivalents clock() instead of DJGPPs uclock() et al. -The REcompiled and modified version of the VC sample also achieves a pretty good >~1000 frames per second But why in hell is the DJGPP version SO much faster than the Cygwin version??? I just changed a few characters/lines.... Can I somehow read an set the timer frequency for clock(), like I could read/set it with the Windows API functions QueryPerformanceFrequency et al.? What else can I do to get high resolution timing? I must be doing something wrong here, so help is very much appreciated! Karsten karsten.wutzke@starconcept.com Please reply to sender and group. Thanks!