I did not affirm that at all.
We collectively poured a ton of work into v5, and surely not to make it slower, less flexible, or generally worse.
It may still be the case, but we’d need some -factual- convincing, as we’re measuring every little nook and cranny of the software to shave time.
A few times over.
I also don’t understand the part with the excuse that the products on the market and the trends are already like that.
They are not “already like that”.
Everyone who is anyone in the market is spending *a ton* of resources to refine the same approach.
They all do so because it is the most economical way towards a converged image.
It’s a fact, backed by very solid maths and extensive field testing, not an excuse.
But you are free to read it as you please.
I can’t achieve better or even the same results faster with Vray 5 compared to Vray Next and that’s a fact!
Then you should give it a bit more time.
It has already delivered a few times in the hands of our clients, and while no work is ever complete in this business (meaning, we did have bugs. we will have bugs.), it’s already feature packed and *much* quicker to delivery than any version that came before.
It just takes some getting used to.
Only different results with some relative time and image quality.
The difference in image quality and render time is due to the various techs having changed, and the way the defaults have been set up for V5.
So that is the reason why without changing anything a scene in pre-v5 may render slower in v5.
That the LC doesn’t bubble in animation anymore, at default settings, isn’t a minor upgrade.
That instead of three layered shaders you can now use one is no small speed improvement.
The list is nigh endless (well, no, it’s the size of a few changelogs.), i’m sure you get the gist.
As for “better” results in “less” time, if you do not measure noise in the resulting render, you’ll never know how much work was being done, and so couldn’t possibly figure out a metric for which version is quicker.
I could share a nuke gizmo with you to do that, or you could use any third party (possibly photographic) application to achieve the same goal.
Noise could well be below visibility level, and yet be half that of another image, hence the engine would have had to do twice the work.
Or it could be less in some more evident part, and much more in others, not readily observable. (the noise distribution profile i spoke about in my previous post.)
It’s better to trust numbers only, not one’s eyes, in this.
TL;DR:
V-Ray 5 is either identical in speed, or quicker, than any that came before when set up to do the same amount of work.
You should spend more time on it armed with this knowledge, and an open mind towards workflow changes, brought on by the new VFB tools, LPEs, the LightMix, the UberShader, and so on and so forth.
V5 is *not* a service pack of Next, or a minor update on v3.x.
It shouldn’t be used as such, for one will risk coming away dissatisfied.