irradiance map slower than BF?

[super gnu]

with the newer builds, im often finding that the irradiance map is slower (including the calc phase) than just doing a BF render.

while its great that BF has improved so much, this leads me to believe that the irradiance map could do with some love.

i understand that the focus has been on using brute force recently, since its more predictable and higher quality, however irradiance map is still extremely useful and appreciated, and im sure not only by me! in many cases the results are more than good enough.

while computers (and vray) have improved to the point where BF is now practical in many situations, there is no getting away from the fact that faster is *always* better. if i can render an image on 2 machines rather than 10, ill be very happy. and since imap is doing less calculations, shouldnt it be faster than BF in almost all situations?

so, my question is: since its fairly obvious the code in the BF engine is far more modern than that used in the imap, are there any optimisations that have been added to the BF engine that could be ported over to imap?

tldr: dont abandon imap!

[Lupaz]

There's a thread about this here too:

http://forums.chaosgroup.com/showthr...973#post668973

[super gnu]

yes and in fact i asked the same question, but it was missed by (hopefully just) being at the end of a thread, so i started a new thread to see if i could get some more feedback.

[vlado]

I didn't miss it; but there's nothing to say at the moment There are things that can be done to the irradiance map to make it somewhat faster, we just need to find time for it.

Best regards,
Vlado

[super gnu]

ok understood, i guess i just wanted clarification that there are things that can be done, and that they are maybe on that famous list somewhere, rather than the imap being consigned to the "photon mapping" part of vray's development schedule, because, imho, imap is still one of the best features of vray.