free download: archviz blender scene

Alain, don’t post images, post the .blend file, so people can test it.

Lighting with a HDR:

http://i.minus.com/ibo6AwFBOIbZ4H.jpg

Language? Little harsh to command him dont you think so?

Yes. Not intended to come across as bold type with exclamation marks. I guess it depends how you read it.

There does seem to to be some mis-communication happening though. Alain says his glass material looks better, emirage is unable to reproduce it using Alain’s suggested settings. Bao2 explains why not to clamp to 0.9, Alain says he sees no relevant difference. I think more information will be communicated if Alain shows his adjustments instead of only stating his disagreement.

Ha! I just noticed the intersecting leaves near the camera, on the pool

Organic,

the way I read it it came across a bit short and this way. I found it that with written text writing in a more open format prevents a lot of misinterpretation like on my side in this situation I guess.

Compare maybe

Dont post images, post the blend file so we can test them.
or
Hey could you also post the blend file so we can test them?

And think about how a stranger would read it.

The stranger knows english?

cekuhnen.

I already made my intention clear in my previous post. Also, I feel offended by your didactic, preachy, preceptive, moralising tone. If you want to continue this, take it up in pm, as it is not the topic of this thread.

thanks for sharing this great scene. may i ask how its done that instanced trees have different colors?

http://www.archiviz.ch/download/pabellon_barcelona_remake_windowsglass_by_alain.zip

Kind regards
Alain

Here are the Clamptests of the scene I’m working on (just raw renderings):
Modell of the Farnsworthhouse by Peter Guthrie.
Barcelona Chair from flyingarchitecture.com
Barcelona Couch from designconnected.com



Kind regards
Alain

@Alain
Well in your clamp tests it also depends on your lights. Try to add a light with strength of 150 or 1000 and then you’ll see the difference. If all your lights have lower strength than your clamp value it won’t make any difference. If you are using just a HDR for lighting check its values in an image editor because different images have different dynamic ranges so the clamp might not have any effect as well

I am lost now. In both images rendertime is worst when using clamp 0.99. I thought you were defending clamp 0.99 was the best option to choose.

With Sun yet on the sky:

http://i.minus.com/iIyFAEFuDoeMr.jpg

Seems to be a bug in Cycles I noticed several weeks ago. But I don’t understand yet what to do to expose the bug perfectly to be fixed.
I can see in some scenes I need to use high values for the lamps and in others I need to use very low values. Seems a scale problem but I can’t understand yet what is happening.
For example in my post above the Sun lamp is using a value of 3. Initially I used 5000 and it did the whole render white. Then lowered to 100 and it was too much yet. Then 10 and then 5 and 3 seems the best value to give the look of six o clock sun I was looking for.

In other scenes I remember entering values of 2000 or so for the lights. Even in the floating head scene the values are of 50 or so.

So, what is causing the need for so low values? I didn’t spend yet time to find it but probably there is a bug or with scale or with multiple importance sampling, is my guess.

I can see in some scenes I need to use high values for the lamps and in others I need to use very low values. Seems a scale problem but I can’t understand yet what is happening.

I’m with you. I suspect something similar. Noticed under many cases.
Discarding these nodes, applying new, same setup, may fixes it.

Actually the wiki says that normal lamps are in Watts http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Cycles/Lamps . The sun lamp has different units, because it would be too high (wikipedia says full daylight 32,000–130,000 lux)(as a test you can make a fully reflective plane, light it with a sun, render and then sample its values). Also mesh lights get more intensity the bigger they are whereas normal lights don’t (usually in other rendering engines you have an option "normalize"which causes lights not to be brigther when you scale them). Lastly light falloff which is inverse square by default. If your scene isn’t modeled in real world scale you’ll need higher values, so that might be the problem you’re having.

Another take on the sunlight:

http://i.minus.com/ibu0DXhjVuDitH.jpg

Yes defending against noise but not against the speed. The difference of speed in this case is not relevant for me.

Kind regards
Alain

I use a strength of 1.0 for the sun and 1.0 for the Skytexture.
1.0 Blender Units = 1.0 Meter in my Scene.

Kind regards
Alain