tube lights

General disscusion about Kray
Locked
tiktane358
Posts: 90
Joined: Thu Nov 16, 2006 9:11 am
Location: France

tube lights

Post by tiktane358 »

Hello
I made a test with an old file I do a long time ago. I just wanted to see if I can render tubes lights that are on the ceiling.
Anybody have a great tip ?
Attachments
test.jpg
test.jpg (145.5 KiB) Viewed 6180 times
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

Hmm yes on quality tab switch to Direct mode and if you get grainy lights turn up lumi light max value.
- Jure
silverlw
Posts: 453
Joined: Thu Jun 02, 2005 7:05 pm
Location: Sweden
Contact:

Post by silverlw »

I would use a different approach and use inderict/photonmapping but hard to tell since we dont have the specific scene to test with. Do a quick test and see how what's faster. Direct mode is very accurate but can take long time comparable to Fprime/montecarlo.
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

Nah don't listen to silver. ;)

I use direct mode alot and sometimes it's even faster than indirect.
Anyways in scenes with lots of small lights you're better off with direct. Also you might want to switch to photon map since it seems to work better in scenes with lamps.
- Jure
Mario
Posts: 135
Joined: Sat Dec 23, 2006 9:42 pm

Post by Mario »

ies. light would do the work here.
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

Well IES lights themselves wouldn't help the render look better. IES is just a standardized format for describing light intensity of a light fixture.
- Jure
Mario
Posts: 135
Joined: Sat Dec 23, 2006 9:42 pm

Post by Mario »

Agree, but this king of tube lamps have a specific distribution that you could only imitate with ies. files.
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

Mario wrote:Agree, but this king of tube lamps have a specific distribution that you could only imitate with ies. files.
I think lighting&surfacing skills of an artist are far more important than IES files....
- Jure
Mario
Posts: 135
Joined: Sat Dec 23, 2006 9:42 pm

Post by Mario »

Depends if your looking for accurate lightning distribution or not.
tiktane358
Posts: 90
Joined: Thu Nov 16, 2006 9:11 am
Location: France

Post by tiktane358 »

What would you use Jure for tubes lights : 3d tubes with % of luminosity, or linear lights ? For linear light what is the better : 100000% with 1cm faloff or 50% with 3m ?
What are the linear light threshold ?
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

tiktane358 wrote:What would you use Jure for tubes lights : 3d tubes with % of luminosity, or linear lights ?
Hmm.. well geometry lights give you more freedom in shaping and are easier to place. They may render slower if they are not simple 4 point polys. You control quality (noise) of this lights with "lumi lights threshold - min - max".
On the other hand linear lights might be a bit faster to render, but you'll have more work placing them...
tiktane358 wrote:For linear light what is the better : 100000% with 1cm faloff or 50% with 3m ?
I think it shouldn't make any difference...
tiktane358 wrote:What are the linear light threshold ?
Threshold is a difference between neighbouring pixels at wich Kray will use more rays. Kray will use at least min rays at each pixels and will shoot more if difference between pixels exceeds "thershold" value. But even if difference between pixels is still more than "threshold" it won't use more than "max" ray.
This method is called importance sampling and you can find it throught Kray and it always works the same way.
- Jure
tiktane358
Posts: 90
Joined: Thu Nov 16, 2006 9:11 am
Location: France

Post by tiktane358 »

I tried playing with these values and I found something that works when the tube light is far of the closest surface about 1 meter.
When the tube light is 5cm close to the wall for example I have a lot of grain and I think that the light burn a lot at source and don't "light" enough around.
Can you show us an example of "Jure" neon in a black room ?
silverlw
Posts: 453
Joined: Thu Jun 02, 2005 7:05 pm
Location: Sweden
Contact:

Post by silverlw »

tiktane358 wrote:What would you use Jure for tubes lights : 3d tubes with % of luminosity, or linear lights ? For linear light what is the better : 100000% with 1cm faloff or 50% with 3m ?
What are the linear light threshold ?
It doesnt matter. Lights with inverse square falloff is just light with a natural and physicaly correct falloff. If When you change it's falloff radius OR change it's luminous power, the outcome will be exactly the same thing. Stronger or weaker, only that.
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

tiktane358 wrote:I tried playing with these values and I found something that works when the tube light is far of the closest surface about 1 meter.
When the tube light is 5cm close to the wall for example I have a lot of grain and I think that the light burn a lot at source and don't "light" enough around.
Can you show us an example of "Jure" neon in a black room ?
I've played with it some and it seems that you're better off with area or linear lights than with lumi polys. Lumi polys seem to produce too much noise...
- Jure
silverlw
Posts: 453
Joined: Thu Jun 02, 2005 7:05 pm
Location: Sweden
Contact:

Post by silverlw »

Im bound to agree with jure after some tests. This example is photonmapping + caysics and "auto" mode. Auto mean everything stronger than 100% luminosity will be rendered in direct mode ,the rest in indirect mode. Custom lights is possible with just polygons but if we use only polygons like i did in this example we start to enter into Maxwell domain rendertimes ;)

and i also did one with arealights only.
Attachments
Area00000.jpg
Area00000.jpg (203.81 KiB) Viewed 6095 times
CustomLights.jpg
CustomLights.jpg (206.38 KiB) Viewed 6091 times
Locked