I have to know how many starcraft game seconds make up a real second; lets say 10 game seconds = ??? seconds o_o

Thanks to the helpers.

16 Frames = 1 Game Second (on fastest)
1 HT = 2 Frame
12 HTs = 1 Real Second = 24 Frames

All we need is this... So..

16:24 -> 2:3

1 game second = 2/3 of a real second (1 real second = 3/2 of a game second)

If you want to know non-fastest game speeds, just search for them.

Oh forgot to add that's an approximation. A more accurate ratio is 8:11.9

The most accurate is that on Fastest, the game runs at 23.8095 frames per second (this has been experimentally derived - 1000/42).

1 game second is 16 frames. 16/23.8095 = 0.672 real seconds per game second, or the reciprocal is that 1.488 game seconds pass for every real second. This means 30 seconds in the real world will count as 44.64 game seconds. 10 game seconds count at 6.72 real seconds.

As hyper triggers occur every 2nd frame, and there are 23.8095 frames per second, we therefore get 23.8095/2 = 11.90475 hyper-trigger cycles per real second.

Finally.

I'm bookmarking this. Thanks.

Approximations -



That should work for any timings you really need.

Post has been edited 1 time(s), last time on Jun 20 2010, 7:26 pm by ImagoDeo.

It's actually 1.488 game seconds in one real second. Rounding up the thousandths would still only yield 1.49.

You might be thinking, "but it's so small, why does it matter?" It doesn't, short term. However, something called the race track principle comes into play. It states that a small differences in rates of change make large changes over time (this should be logical o/c).

Okay, after doing the math I come to the formulas x/11.9 for exact seconds and x/12 for approximation. With this we can calculate exactly what the difference will be between the two. Just take x/11.9 - x/12. This will yield 0.00070028 x. Hurray! This means that after 1000 trigger cycles (1.4 minutes), if you use 12 instead of 11.9 as your benchmark for real seconds there there will be a 0.7 second deviation (the 12 will have taken 0.7 seconds longer). At 10k trigger cycles (14 minutes) there is a 7 second deviation.

Not very much to be concerned about, but useful if you're looking for pinpoint accuracy.

Post has been edited 2 time(s), last time on Jun 14 2010, 2:32 am by Sacrieur.

Impossible with the given system. There are compensations, but we can't work with decimals, which causes big problems in counter arithmetic. Whole numbers are all we got, unfortunately. I do understand what you were getting at, though.