Small Bath MR exploration

[William Alexander]

A very small bathroom. Max6 rendered with mental ray. It's a work in progress...

 

http://www.cgarchitect.com/forum/filepush.asp?file=Clydes4.jpg

 

Looking for some feedback.

 

Thanks

WDA

[jucaro]

WDA,

 

can you post the settings you used?

[skogskalle]

at first I thought the lighting was a bit flat, but since you have such a big window and mirrors I guess there wont be any really dark shadows... so its probably good as it is.

you have a dark spot on the floor to the right which I dont understand where its coming from...

 

The modeling looks good to me.

 

The only real problem I see is the materials. The bathtub and the toilet looks soft, like they where made of cotton... which would be kinda weird And the shower-thing (pardon my english) on the wall doesnt look like chrome yet - it needs more bright highlights.

 

it can be improved, but its allready a nice pic.

 

cheers!

Kalle

[salf]

nice scene, let a newbie like me point 2 things:

 

1- The union between the toilet and the floor looks strange, it looks like the toilet is just resting there and its not attached to the floor, so you can move it, maybe is too bright, it should have some shadow along the shape of the toilet base where it meets the floor, and if you go further, remember that the toilet is pasted to the floor with "white concrete" (dunno the term in english), so you should see something similar to the union between the tiles.

 

2- The space underneat the sink looks strange, its a mirrow?, and if its an empty space it should be darker, or it will look better if you cover it with a small cabinet door, cuz it looks strange.

[graphix]

wda,

 

very nice work...... aside from what has already been pointed out I must say the grout on the tile floor looks about 1/4" deep.

 

As mentioned before there seems to be a total lack of shadows, even with the mirrors and polished tile there should be some more definition to the shadows, like behind the toilet.

 

Good work...look forward to seeing the end result.

 

graphix

[William Alexander]

Settings

 

http://www.cgarchitect.com/forum/filepush.asp?file=MRsets.jpg

 

All settings were progressive from minimal to current. Sun-Spotlight is global controlled GI and the fpiont area ies- tub lighting is local controlled GI energy,photons. Tryed final gather... very unhappy with darkening / flattening of image. Basically it seems to kill the dynamic range. jucaro any thoughts?

 

The mirrored door panel is there to -open up- the space and create some interest. Without it any vanity cabinet really closes in the space.

 

The tile is 1/4 inch above the grout. Needs edge refinement!!!!

 

I did brighten the image in PS. Should have adjusted using the dropper tools. New work station monitor-cg & print, old one lost 15 percent of grey scale to black. Still adjusting settings and thinking for web/graphics oriented monitor. So it washed out a bit.

 

Raw max image

http://www.cgarchitect.com/forum/filepush.asp?file=RAW.jpg

 

Better levels adjusment-high compression

http://www.cgarchitect.com/forum/filepush.asp?file=RAWlvl.jpg

 

 

Salf,

Did you mean to say... the base of a toliet has eased or rounded edges that cast a shadow and ground the object. Because I initially took what you stated as toliets are always caulked and this caulk is visible as the grounding element in an image?

 

The chrome is physically correct. Look at the WC handle it's the same material. Most of the chrome is in the ambient range due to lighting and does not shown much reflection and specular. Suggestions?

 

The stray unknown shadows- Wall and door and not rendered but used in GI calcs. Shadows may or may not help the image? This also illuminates the toliet - GI - due to the sunlight blasting that wall.

 

Thanks.. more wip to follow

WDA

 

[ January 24, 2004, 03:53 PM: Message edited by: wda ]

[jucaro]

WDA,

 

It seems the mirror isnt generating any photons, so you might just better turn off caustics GI as this only contributes to long render times. You can turn it on again after you can achieve the intended illumination levels.

 

The ceramic fittings lack the specular reflections.

 

If I may suggest:

 

1. Turn off GI

2. Turn on FG

--start with a few samples, say (100)

from that amount of rays, you can start building up the rays cast value geometrically.

 

If you want, you can email me the raw model.

[ch83575]

WDA,

 

What MR calls GI is not what you are thinking of as GI. In MR FG is much more like the radiosity type GI you are probably used to with Max and things like v-ray (i know that v-ray can supliment its GI with photon iradiance maps, but that is beside the point). The GI you are using is strictly a photon tracing method. The problems you are having are due to a few things:

 

1. Not enough photons

2. Not enough photon accuracy

3. Too many bounces (turn down bounces to 2 or 3)

4. Most of your materials are specular and therefor default to reflecting almost 100% of photon energy.

 

Having said that you can probabily see why you are not getting any shadows (the dark spots on your floor are not really shadows, they are the gaps between photons, thats why you need more).

 

Item #4 might actualy turn out to improve the scene the most. To change the photon behavior you can over-ride the Max defaults for photonic properties for your materials and apply a photon_basic shader. Make the properties more diffuse and less specular. You should be amazed with the results.

 

Also, any skylight or or dome light or environment light will not effect you scene at all without FG on.

 

If render times go up drasticly when #of photons and photon accuracy are increased, turn on FG at very low settings just to smooth out the effect.

 

Good luck!

 

By the way, i think you have a great start!

 

-Chad

[Joosti Ink]

Hi WDA,

 

Few things I would like to add:

 

- is't the hairbrush WAY too big?

- the window (is it a window?) doesnt look like a window to me, more like it is a poster glued to the tiles. Let some light come trough it, and let it reflect some.

- the mirror refelection is a bit too shart imo.

- it is not clear where the light is comming from, my suggestion would be to let some light come in trough the doorway(?) you are standing/looking from.

 

Anyway, looks like a lot of comment and the last post was much better then the first, but I think you can get a real nice render if you work on these little things.

 

gr,

Bart

[William Alexander]

Changed some settings

 

GI Radius from 3 to 1.5

 

Photons. ref 5 refr 5 sum 10 -due to clear glass refr depth

 

FG on, samples 500, max radius 3

 

 

FG only samples 200 Yuck

 

Mold Fungus really bad

 

The real Image.. PS levels shadow-midtone-highlight...enhanced glossy specular-paintbrush white luminosity blending... need more work with MR to achieve these effects raw render...

Cleaned Up Bath

 

rgrds

WDA

[jucaro]

Hi William,

 

I have checked your file and these are what I think about your setup:

 

1. Use a direct light instead of a spot light. Better yet, use the IES Daylight assembly+MR light shader (Infinite Light)

 

2. Use a bump material for your tiles instead of modeling them tile by tile. Although you can create a greater realism by modeling the tiles, the same effect can be achieved by using bumps and displacements.

 

3. Optionaly, you can use the Architectural Materials instead of creating your own. The Architectural materials is just a Mental Ray DGS Material which is predefined for architectural use.

 

For the rendering setup:

 

-Turn off Caustics.

-Turn off GI

-Turn on FG

-in FG, use 5000 samples

-in FG, use a Max value of 10 feet and a Min: value of 1 foot

-Adjust exposure controls as necessary

 

Best Regards

[William Alexander]

Juan Carlos,

 

1. Ah, yes... no light shader, thats why the day light did'nt work well and the spot light was substituted.

 

2. The high sample levels, long FG times. Map / displace works best. I don't model tile by tile, use pre-maid blocks in ADT. Allows for some design and material layout flexibility. + having accurate line drawings. But not clean CG viz models-I suppose with the long render times I could do elevations by hand, lol.

 

3. The arch materials, did not realize that.

 

4. The FG Min Max, you are refering to the radius and not the ray decay, correct?

 

5. One should not avoid seemingly large sampling, photon and energy amounts? While the radius amounts should be kept relative to the scene dimensions? Is a 1/10 rule of thumb a good standard or starting piont?

 

Heres one rendered with your settings. Working better, just need a bit more light upper corner.

FG Only

 

Thanks,

WDA

[William Alexander]

A revision,

 

FG samples 2500

max 10' min 1'

Fall off, more illumination higher in room,

start 6" stop 15'

 

http://www.cgarchitect.com/forum/filepush.asp?file=Clydes7dof.jpg

 

[ January 30, 2004, 12:45 PM: Message edited by: wda ]

[jucaro]

Do have a skylight in there?

[William Alexander]

yes,

 

Daylight system. Sky was turned down to .3 multiplier. And sulight up to 50. Propably a feeble effort to get the suns direct illumination to over power the effects of the sky burning an over intense light. Reference the diagonal cut aways of the walls. Color and relative photo percieved intensity visible there.

 

I maybe having perception problems actual vs CG of the illumination in the space. It is real, I studied the existing lighting. Eye squinting for f-stop film speed and varyiyng contrasts and how the light was distributed. May be puuting too much value on my perceptions and not enough on illustrating. However, this image should not have the dark shadowy kind of feel-that seems present with standard type settings, imho it makes it feel small- not desired. It's not a WC, or closet anyway.

 

wda