Hi,
First of all, we really want to stay away from any "speed comparisons". It is extremely hard to reliably compare two different render engines. Maybe Vray would render faster with different render settings? Maybe Corona would render faster with some tweaks after the automatic conversion?
As far as I understand you are surprised that Corona A6 rendered 100 passes faster than newer versions. Also in your tests it seems that the daily build is the slowest - right? Well, rendering 100 passes does not say anything about image quality. Version X of Corona could render 100 passes in 10 minutes, but maybe the quality would be worse. Version Y could render 100 passes in 2 hours, and maybe the quality would be far better than after 100 passes in some other version... Maybe in A6 rendering 100 passes was just fine, and in newer versions it is an overkill and 50 would be enough for comparable image quality?
The main question is - is the image quality good enough for you? So rather than comparing numbers of passes, you should compare image quality after given time. We even noted this in our 1.4 release post some time ago:
"A note – with the introduction of Adaptive Image Sampling, it is now more useful to use the new Noise Limit for your renders than the old Pass Limit, to ensure consistent quality for every image or frame in the least amount of time."
https://corona-renderer.com/blog/corona-renderer-1-4-for-3ds-max-released/So comparing images from the same scene rendered with different versions of Corona with fixed time would be a good idea. If the quality is worse in newer versions - that would be strange indeed.
Another option is fixing the noise amount, and making comparisons how much time it takes for different versions of Corona to reach that noise threshold. But this can be done only for 1.4+ versions.
By the way - can you share your comparison pictures?
