AccuRender nXt

advanced rendering for AutoCAD

On a project, after some quite minor changes of design, renderings started to turn out very different - no changes in lighting etc. Trying to remedy it, I found I am not familiar enough with controls, affecting the final result:

  1. What does exactly "Fresnel" do? On glazing, I keep it usually 0.7-0.8, what if I decrease it?
  2. How about IOR? I keep it between 1.5-17 on glazing... does 1.7 mean I will see more of the interior than on 1.5 or vice versa?
  3. What should be proper color of glass? Should I keep it very light, or not? How will darkening it affect the final result(with conjunction of other pramters?

I apologize again, I remember this must have been discussed already, but just now (deadline approaching and renderings keep looking ugly) I would mor than welcome some comprehensive explanation.

Views: 1548

Attachments:

Reply to This

Replies to This Discussion

From what I can see indirect is on in the second rendering and would need quite some passes to resolve. 

 

What I suggest is to have it off or render first at a low resolutionof about 640 x 480, to see the result quickly at about 20 or passes. When satify then go the whole log.

both renderings use THE SAME lighting setup. Pathtracer=1.

Not 100% sure what's going on-- but the simple act of enclosing the space may have blocked a lot of the direct light you were seeing behind the glass.  It may be legitimate in that the lighting that remains is largely indirect and may take a while to resolve.

 

Make sure Sharpness is set to 1.0, caustics are not on, and the thick/thin business is handled as best you can. 

 

It's unlikely these settings will matter much-- you may need to change the lighting a bit-- maybe add some direct sun.  You've got decent values for them already.  Fresnel values shouldn't affect the Path Tracer at all if your Transparency is set to 1.f.  Ligher IOR values make glass less transparent and more reflective.  Colors should be kept light.

 

 

 

To me it seems you are right about "enclosing the space" - however, I would not have expected that this could make such a difference - the narrow corridor to allow the light to enter the building (from the left to the right) is some 25m long...

  • caustic is not on
  • thick/thin to be adressed separately: If you have ACA installed, could you please tell me, how nXt handles solid infill in the curtain wall? From the real-world logic point of view, it should be handled as thin...my infills are 10mm thick to resemble two sheets of glass. There are two layers of curtain wall - see edges.
  • Sharpness is set to 1
  • Transparency was at 0.99 - how did the "Fresnel" parameter come to play there?
  • On HSV scale, V was at 99
  • IOR was at 1.55, I did increase it to 1.8 (this bit about "less transparent and more reflective" is really valuable, thanks)
  • There is not much I could do with lighting, the facade is facing north, so no direct sun at all...the rendering must be realistic.
  • I increased no. of indirect bounces to 4 on both renderings.

My goal is to see more of the inside of the building. From looking at real world world buildings I derive my belief I should to - if there is no direct sunlight hitting the facade, you see through it...

 

It seems that the increase of IOR did help, thanks... but still, I wold expect more transparency.

Attachments:

These look pretty convincing to me for north light-- really nice work, actually.  I'm not exactly sure what you mean by solid infill-- if you mean you're using a solid to represent the glass in the curtain wall then you should be all set.

 

If you want to send an nXtImage I'll be able to do a little more fact checking....

nXtImage attached.

Glad you like it, but still, I would like to see more of the interior. The only thing that comes to my mind is to increase IOR even more than 1.85, which sounds strange - even 1.85 is too much.

That is why I ask regarding curtainWall - I am not sure how nXt handles its infills (glazing panes on the rendering) - see attached: once I explode curtainwall, I get 3D faces, but prior exploding an infill is one entity....

Glazing in nowadays usually doubled, so I should get more reflections, should I not?

Attachments:

You may have misunderstood the original message-- there was a typo.  The higher the IOR, the less transparent and the more reflective the glass is.  Lower values of IOR will result in more light on the inside.

It's going to treat the glass as one thick sheet, exploded or not.  This won't matter very much.

 

Adding artificial light to the interior would obviously help matters-- very expensive, though, in terms of rendering and the effect would be subtle.

Yes, I misunderstood, thanks for the correction. Here is what I get on IOR decreased to  1.3 ... looks better even after only 90 passes...it is interesting, though, that change of IOR from 1.8 to 1.3 brings quite subtle difference - I would have expected more dramatic change
Attachments:

You'll notice it more at glancing angles.  For reference:

Diamonds: 2.417

High Index eyeglasses (very popular now) 1.87

Most architectural glass-- 1.52

High index glass (laminated or w/ thin film coating)-- up to 2.3

Set transparency to 1.0 and set reflectivity to something like 0.6 or less.

I'd keep IOR at 1.5 for glass.

With transparency set to 1.0 the reflective intensity should have no effect.  All of the reflection will be due to the IOR.
it seems Roy is right:-) - please find attached image rendered with settings, recommended by Peter
Attachments:

RSS

Search

Translate

Latest Activity

Maciej posted a discussion

nXt + AutoCAD 2025

Gentlemen,The last file I used to load, was ArAcad_24_64.arx, and, unfortunately, it doesn't work with Acad 2025.Any idea how to get the proper nXt-App which suits Acad 2025?Any help appreciated...Regards, Maciej MarkowskiSee More
23 hours ago
OYEBANJI EMMANUEL replied to OYEBANJI EMMANUEL's discussion CAN WE GET PETER MILNER TO TRAIN SOME OF US FOR A TOKEN ?
"I meant Chris Fox, not David."
Thursday
OYEBANJI EMMANUEL posted a discussion

CAN WE GET PETER MILNER TO TRAIN SOME OF US FOR A TOKEN ?

Good day, forum members,I was thinking that we could each, if interested, contribute a significant token to give Peter Milner and David to train us virtually in nxt render for Autocad. Am not sure he still uses it but the experience he has spanning over 20 years cannot be underestimated. I wish the forum co-ordinators could look into this. So that more of us will produce realistic renders with the paid version of nxt render and we will be able to encourage other users of other more advanced…See More
Feb 12
OYEBANJI EMMANUEL commented on OYEBANJI EMMANUEL's photo
Thumbnail

V3x copy

"Thanks"
Feb 11
Maciej commented on OYEBANJI EMMANUEL's photo
Thumbnail

V3x copy

"Soft & nice light scheme.  :)"
Feb 11
OYEBANJI EMMANUEL posted photos
Feb 7
OYEBANJI EMMANUEL posted a discussion

LIGHT STREAKS IN ACCURENDER STUDIO

Good day everyone.Does anyone here know how to create light streaks in ACCURENDER studio or in nxt render ? Please share your knowledge with me. Light streaks for interior ceilings in the p.o.p works and under, a t.v console. See More
Feb 2
OYEBANJI EMMANUEL posted a photo

3D_Replication final

I created this replication of one of Peter Milner's renders, not with AI but with sketchup (with the matchnewphoto tool), autocad and nxtrender.Critics and comments are welcome.
Dec 31, 2024

© 2025   Header image courtesy Peter Milner   Powered by

Badges  |  Report an Issue  |  Terms of Service