This one adds a setting to tweak the workaround timing (which is now disabled by default). [...]Well, if I say I added a new setting, obviously it was not there before xd Let me know if that option in aggressive or 10 ms helps somehow.
There are various influences, so at least I tried to compensate for that with enough attempts.The helix code is pretty immune to additional compiler options that can speed up things for flac for example.
For me, good stable results with minimal differences.
On the settings menu, like everything else. I don't understand what's not clear, I even showed you a screenshot yesterday...Ah, your screenshot did not show the new settings therefore the confusion. I never saw the new options ...
Thanks for confirming.No great surprise, I guess, considering compiler advances in the intervening period.
Looks like the old ASM path is somewhat underwhelming.
Thanks John, just wondering on the Intel/MASM build: Did you test with -U2 to actually select the SSE code path? Without -U2, the C routines will be used.Yes, -U2 was used.