Any plans to fix RAYTRACED mode or is it now where its gonna stay?



If you can supply and model I will gladly test it out and see what I can do to speed that process up on your system.
By the way, have you tried rendering with Cycles on the CPU instead?

PS! Logging off here, past midnight now and beer is calling. Cheers!

(Nathan 'jesterKing' Letwory) #110

No, it is an INTERACTIVE solution, NOT realtime. That is what it says here, too:


Yes, EXCEPT NEON was intended to be RHINO’s approach to RTR, and was promoted as such, and Raytraced is its intended replacement.

I’d already encountered the info at the link you provided, and was LOOKING FORWARD to a GOOD RENDERER I could use INSIDE RHINO, with a possible option in addition to RHINO-RENDER, apparently the intended use for Raytraced. ALL of that works well as a concept, and offering it as an included (no additional cost: “FREE” ) internal feature is a WONDERFUL idea.

So- to my point: a “free” offering (included in the PRICE OF PURCHASE of the base package) requiring $350+ up to a THOUSAND DOLLARS or more to utilize … AND the stated intent of at SOME point REPLACING RHINO Render ? See where I’ve been TRYING to go with this, all along ?

Don’t get me wrong, I’m TRULY excited about the capability, and genuinely hope for success, but consider- let’s say we find that magic combination of settings that pulls the time back to something “reasonable” - say 3 to 5 minutes per frame, have we solved the NOISE problem ? Are we expecting to run Animations at whatever frame-time we DO get to ? Is there maybe a possibility of a truncated version where we get “usable” results earlier than the complete cycle-run (as in NEON) (fewer bonces, fewer cycles, something ) ? These are just questions, but I sincerely promise, they’re actually a real thing.

You guys have done something WONDERFUL with all of this, and with including A-O as a feature of the RENDERED viewport, with or WITHOUT “Shadows” on, but the result there is marginal too. I’ll check again for any additional improvement that may have been made since I last had to disable it out of desperation to get my project completed and delivered …

Ok, I think I’ve about covered it. Everything I’ve reported is EXACTLY as I’ve presented it, and these will remain barriers to implementation generally, as we (customer base) look to adopting the product going forward,. They aren’t insurmountable, but they DO have to be dealt with by folks on THIS side of the table, and I can assure you, While $350+ might be a reachable stretch (I’m looking to budget that with my next client invoice), $500, $650, $1000 + and MORE for MOST may NOT necessarily be. I mean no ill will, but I DO mean clear, open and honest communication on this important topic. I’ve been a RHINO fan and user for MANY years, and hope to be for many more to come ! I can’t thank you folks enough for such a POWERFUL tool !

Thanks again -



You know I’ve been using Rhino since 1997 or so, I think the first machine I ran it on might have been a 486 DX 50 overclocked to 83mhz? Is that possible? That or a Pentium 166. I remember back then, when most “3D accelerators” were frankly kinda garbage, people asking on the original Rhino 1 Beta newsgroup for Rhino to make more use of GPU hardware. Now finally with V6 they’re really doing that, with great results, but there is a price to pay. The GTX 550 was a low-end gaming card, with less performance than a card I bought in 2009.


Yeah, maybe not so much.
Anyone wondering can look it up and see My version is ALLOT closer than the one you’ve presented.

SO- you started at v-1. Well THAT beats me by a couple of versions ! I only started at -3 ! I guess I’m just a young whipper-snapper compared with YOU ! When I got my card, it was new out, and Mid-range. Yes, it was a gaming card, but that’s all that was available for less than a THOUSAND DOLLARS ! See the trend emerging here ?

Processor-shaming went out with the demise of the Pentium as THE power-house of choice , so really, I’m gonna let YOU have this one and bid you good night. Oh, by the way, EVERY SINGLE NUMBER I’ve reported STANDS, no matter HOW clever you may want to be side-stepping THAT inconvenient reality :wink: ! Mine runs Open-gl 4.1 - modern enough for ANYTHING writtten for nearly 200 cores of parallel computing power (RHINO, for instance), and if configured correctly, I can have a 15-minute animation RUN and FULLY RENDERED overnight, that YOU wouldn’t have until FRIDAY. How do I know ? I’ve_DONE_it .

LOTS & LOTS of CORES ? LOTS and LLOOTTSS of $$$$ :wink: ! Why I’m sure YOUR cores are the BIGGEST AND THE BEST ! Bet YOU got LOTS of money to make sure YOU got THE BIGGEST BADDEST video card bragging rights ANYWHERE ! Very COOL !!!

So, render-ON, my man ! RENDER ON !

Sincerely, this has been fun ! Everything I’ve put forth is EXACTLY as I’ve represented it, and MAYBE we’ll see a viable product SOON ? I think we’re ALL hoping so ! I know I am . ALL the BEST you you and yours -


(Andrew le Bihan) #114


Yes - but the point you are missing is that Neon isnt using your 550ti. At all.



But the picture you posted yourself says the release price was $150. In 2011. That’s “low-end,” and it was not the only thing that wasn’t $1000, please stop that. Please post an actual file that illustrates this speed difference you’re talking about instead of asserting what sounds to everyone else who uses and develops these products like nonsense.

(Andrew le Bihan) #116


What GPUs do these school computers have?

It sounds like you’re running in CPU mode.


(Des) #117

As it is a public school, most systems don’t have a GPU, they just have the basic on board video card. This was not a problem with V5 & Neon as the systems could easily accommodate the ray tracing. The trouble is that the students use the lab for 1 hour then a new class comes in, so if the ray trace takes over 30mins it is not going to happen. Likewise having 20 students starting at a screen tracing for that long is asking for trouble in the classroom.


As Andy says, forget about the 550, it just isn’t good enough. It’s like complaining that Neon is shit because it doesn’t render faster on an old Core2Duo CPU…

So on a final attempt to help you: Answer my questions already instead of repeating the price of beastly 1080 :wink:

1- What CPU do you have?
2- Did you even try to render on the CPU?

Use: RhinoCycles_SelectDevice and set the value to 0 to toggle the CPU.
Here my laptop with a four core i7 renders almost twice as fast as the 750m, so rendering on the GPU doesn’t make that much sense.

CPU, 300 samples: 2.45 minutes
GPU, 300 samples: 4:10

This is all reflected light, something that you can not achieve with Neon UNLESS Brazil is installed and knew how to set up a scene so Neon could benefit from it efficiently.

(Des) #119

Thanks all of you who had constructive input. I cannot test this on the random “Bunky” machines at school since it is a weekend, but I have found a way to make it for for my situation.

Changing from CPU /GPU had very little effect (I’m using my home PC which has a GPU)

What makes it possible is to change the samples to 128, this rendered in 7 Mins.

Still a problem with teh black being a mirror but I can live with that. Knowing McNeel, they will tweak a few things down the track.
My main reason to find solutions is autodesk & sketchup are dominating the educational market and I know Rhino is better for the students.

Once again, thank you all for your help


(Des) #120

Oh Yeah, the output


Two ways of tweaking the black material are:

  • set the reflection COLOR to dark gray

  • turn on fresnell reflection, but then you can’t control the amount

(Nathan 'jesterKing' Letwory) #122

eh? Turning on fresnel you definitely can still control the amount. With fresnel on 0% reflection means no reflection through to 100% reflection meaning full reflection.

edit addition: once you turn on fresnel the reflection color works like a ‘filter’. White means reflect all colors, i.e. mirror like. A non-white color means reflections are coloured like that, this is like the metallic workflow.


Blame the shinny colors and the price.

It’s like asking youngsters to grow, pick and EAT their vegetables, all while there is a jar of jelly beans on the table.

What would you expect them to go for…


Meh… test it out for your self and convince me if you can :wink:
Here 5 red balls with varying reflection and Fresnell ON:
0% 25% 50% 75% and 100%
For some reason 50 and 75 has more reflection than 100.

Edit: But good thing: on latest build if I turn off fresnell on all five materials in this file then the reflection amound works fine:

(Nathan 'jesterKing' Letwory) #125

What is the Rhino version used in the first image, and what in the second image? Anyway, the second is what I see here as well - but then again, I am working with the latest dev build I cranked out myself :wink:


Ok, I see that I wasn’t so clear on that.
Both images are latest build.
(6.3.18090.471, 31.03.2018)

First image has all “Fresnell” materials and the second image is with out fresnell.
Values for reflection is 0,25,50,75 and 100 in both images.


That’s the price NOW. When I bought it it was more than $250, I’m TRYING to think I’m remembering I went for the “ti” adding more than a hundred to THAT price, but that’s only recollection. No, BACK THEN it was EXACTLY as I’ve presented. Maybe a shift from PROVING I’m wrong and you’re right, we could shift the conversation to STILL with a card at $1000 we’re still getting NOISE at the END of a render that took the best part of an HOUR on someone’s card here that claims to be the hi-end version YOU’d mentioned ? That’d be nice.

With ALL of that said, I STILL get the results I mentioned before on a $150 CARD (!) :slight_smile:

  • STD RTR Mode: ZERO POINT ZERO seconds per animation frame
  • 30 seconds per animation frame in NEON mode.
  • unuseable time to an unusable result “Raytraced” :slight_smile:
    NOISE in an “IMPROVED” image that “LOOKS FINE” -

In this context, dealing with Card-shaming is actually pretty entertaining ! Thanks !



150 passes, 12 seconds: