You can set a minimum size reduction (eg. 0.1%) and if in a pass the gain less than that the optimization for that file is halted.
Output: C:\DOCUME~1\AMMINI~1\IMPOST~1\Temp\Homunculus v01p261 copy.png coltype:6 (RGB+Alpha) filter:5 (mixed)
Input size: 356248 bytes
Output size: 421819 bytes - 80533.125 -> 341285.875
Output size: 330685 bytes - 3886.875 -> 326798.125
Output size: 325630 bytes - 93.125 -> 325536.875 <---almost useless
Output size: 325492 bytes - 21.750 -> 325470.250 <---almost useless
Output size: 325462 bytes - 2.000 -> 325460.000 <---useless
Output size: 325452 bytes - 0.750 -> 325451.250 <---very useless!
Output size: 325448 bytes
Change: -30800 bytes ( 91% of original)
As you can see the last 4 passes makes the png only 120bytes smaller, stopping at the third pass result would have cut the processing by an half, stopping at the second by 66%.
Sorry for my english and keep up the good work!
not a support question