WOW! Sure am glad I don't have an AMD card....   ~V~

       Val RogueAnanda

Val,

There is that, but also I am not at all convinced still that iRay can be used for animation (at least practically).  I think it's VERY telling that the example file Miranda shows rendered as an animation was rendered at 320x240.  That tells me that trying to render a test at anything other than a ridiculously small size would take too long even for RL to do (my gut is telling me it took several hours -- and with multiple GPUs.  Multiply that by a factor of 8 or more and you have for just one avatar standing there perhaps 8-12 hours...).

I hate to be right all the time, but iRay just doesn't seem at all practical to me.  I'm sure there will be some (many?) here who are willing to put up with insanely long render times for it, but not this camper.

---

Alienware Aurora R12, Win 10, i9-119000KF, 3.5GHz CPU, 128GB RAM, RTX 3090 (24GB), Samsung 960 Pro 4TB M-2 SSD, TB+ Disk space
Mike "ex-genius" Kelley

That's what I read and it still was not clear if 3 is a free upgrade from 2 if I own the game dev pipeline already. I understand that the IClone users will get the base one but as for the full pipline owners is this free as well or is there an upgrade fee. That's my concern. 

---

iClone 7... Character Creator... Substance Designer/Painter... Blender... Audacity...
Desktop (homebuilt) - Windows 10, Ryzen 9 3900x CPU, GTX 1080 GPU (8GB), 32GB RAM, Asus X570 Pro motherboard, 2TB SSD, terabytes of disk space, dual  monitors.
Laptop - Windows 10, MSI GS63VR STEALTH-252, 16GB RAM, GTX 1060 (6GB), 256GB SSD and 1TB HDD

As Miranda posted here:
Some functions such as Auto-skin weighting, GoZ and LOD are supported in CC3 Pipeline Edition only, which implies that CC3 for IC plus 3DXchange Pipeline is not equal to CC3 Pipeline Edition

This means that the CC3 Pipeline version needs to be purchased to access all options. If you don't want any of the options listed above then the iClone Pipeline version with the free CC3 upgrade is all that you need.



Gerry

---

System: Win 10 Pro (21H2), Asus X99-E WS, CPU i7-5930K -3,5 GHz, 32 GB DDR4 2666-16 RAM, NVidia GTX 1080 Ti GPU - 11 GB VRAM(Driver Studio-536.99), Samsung 850 Pro 512 GB SSD, 6 TB HD storage.

Hmm...I know we saw 320 x 240, but I'm willing to bet 1080p will be within an expected 'norm' .  Any render, for example, that I've done on Octane takes about 6-10 seconds a frame @1080p, with 6 Pascal GPU.  that includes with denoiser.  Any notion that Octane denoiser is instantaneous, @1080p, is not the case.  I use it, forums speak of it, and developers confirm it is not the case.  
So, the idea that ANY *raytrace* renderer will do 1-2 seconds @1080p is just not going to be feasible now with current GPU technology, in any case of rendering.

HOWEVER - What is definitely very encouraging is the iterations setting + Denoiser Iterations settings shown.  This is very good news, because it means, like the 3DS Max Iray plugin, you can set a small sample/iteration target setting, say 10,  and then have the denoiser kick in at 10 iterations, for an 'upon completion' denoise mode.  This is where, with multi-GPU, I have experienced a 6-10 second render time per-frame, at 1080p.
And yes, if too long, use iClone realtime PBR, UE4, Unity....it's not like Iray will 'turn off' your ability to use it.  But yes, for RayTrace, this is the reality of where the technology is, now....
The reason I say all this is that 6 seconds for a 1080p iClone Iray render would not be a fail by ANY stretch of the imagination.  It would be, quite the contrary...A HUGE win.
I mean, do I want it to be 1 second? - YES.  Do I like that UE4 renders MULTIPLE 1080p frames per-second? HELL YES!...
But not until RTX will anything bring us that.  And unfortunately, at a cost of 3-4 Volta cards in your rig.

@Miranda(RL) - thanks for the feedback to comments in this thread.  I myself was initially disappointed by the smaller # sliders for tonemapping. 
I have since come to realize that what will actually be the biggest breakthrough is allowing us to use multiple LUTs on an Iray rendered image.  As that would negate the need for more image-shaping sliders.
Please consider LUTs as part of the initial Iray release.
Thanks!

Tony,

Let's be generous here and take your optimistic (which I think are, particularly for only one GPU) figures -- six seconds per frame.  10 frames a minute is our top speed and we'll even say 24fps is our goal (although nowadays it's FAR more realistic to use 60fps for any digital delivery platform, or at least 30fps).

Rendering 20 seconds an hour would have been GREAT about 20 years ago.   Nowadays to wait 10 hours for a a few minutes of animation is ridiculous, not when you can get great looking stuff in a fraction of that time.  As I pointed out MUCH earlier (when this was first announced) we're not even talking about a final render (because I don't care how close you get with a preview using iClone's native one, you'll still have to render at least once or twice more to get it right). 

But here's the big problem (and one that RL doesn't even seem to want to address, the elephant in the room): without a render queue you're really screwed.  Do you really want to be at the mercy of a single render pass when you might be at work (you poor slobs :> or asleep.  You will waste precious hours because the best iClone can do is that one render.  AND without the ability to use a render farm you are also tying up your machine (in many cases folks ONLY machine capable of doing this) because once you start your render you won't be able to do anything else.

Sorry -- but iRay sounds pretty crappy to me, and I knew all along the times would be this way (frankly, I think they'll even be a bit worse).  I predict now (and I'll be right again.  Sigh) that we'll see folks here put out a precious 10-20 second animation showing how wonderful it is and we'll NEVER see anyone using iRay for a full movie (one of at least three of four minutes, one that tells a complete story).  Not in my lifetime.

---

Alienware Aurora R12, Win 10, i9-119000KF, 3.5GHz CPU, 128GB RAM, RTX 3090 (24GB), Samsung 960 Pro 4TB M-2 SSD, TB+ Disk space
Mike "ex-genius" Kelley

I am curious about this point, can't you just render a certain range of frames, in different batches, on your own manually, such that you can utilize the PC as you wish?  I know what you are talking about, I use my own method in Octane.
(ie you have 5,000 total, so you render 1-1000 in one pass, then later 1001-3000, then later 3001-4000, etc.)