Greetings-
I have been playing with the idea of building my own real life Cornell Box/Radiosity Room. Kind of expensive and my wife was really rolling her eyes, not to mention, I don't have much time anyway so I gave that up......for now.
I walked into my hall bathroom a few weeks ago, closed the door and attempted to turn on the light. I missed. I thought, WOW its dark in here. Click! The overhead light was not the only one to turn on at that moment. My hall bathroom has no window and therefore is unaffected by exterior light. Hmmm. My own Radiosity Room....for free!
Here are the beginnings of my "bathroom adventure". Perhaps not as exciting as 'POPBOT' from the Showcase page (which I have to say is AWESOME!), however, a good chance for me to study light and how to model it in A:M. My results will come slow but in the end I hope to have shared something worth while. Comments are always welcome.
Full Version: My Own Radiosity Room
Hash, Inc. Forums > Technical Direction and Development (Learning Animation:Master) > A:M Rendering, Compositing and Special Effects > Texturing, Lighting and Effects > Lighting Effects
WOW!
Thats a really good start!
It beats my alley
Thats a really good start!
It beats my alley
Very good start. You will need to increase the both the sampling area and the photon sampling to get rid of that uneven illumination along the corner between the wall and the door frame.
You guy's experiments with radiosity really make me want to experiment myself.
You are really off to a great start.
Love the composition as well as the general lighting/look.
You are really off to a great start.
Love the composition as well as the general lighting/look.
Thanks! I liked that simple back alley look. As I said before nice composition.
Yves- I will change that on my next render. Currently my settings are:
Photons 20000
Sample Area 400
Photon Samples 100
Final Gathering on
Samples 200
Jittering 20%
Gamma Correction 2.2
I believe I used the calculation spreadsheet for setting the area. I may have miscalculated. I built my models to full size and then scaled them down to 25% in the choreography. I'm not sure what measurements I used for the calculations. I will guess that more is better than not enough.
I am assuming that the reason for scaling objects in radiosity shots is to keep from maxing out the sample area for really big scenes. Which reminds me, I believe once upon a time, there was some discussion concerning some adverse radiosity effects when scaling glass. I am using glass (with refractive properties) and other than the expected increased render times, I don't see anything wrong yet.
Rob
More to come!
Yves- I will change that on my next render. Currently my settings are:
Photons 20000
Sample Area 400
Photon Samples 100
Final Gathering on
Samples 200
Jittering 20%
Gamma Correction 2.2
I believe I used the calculation spreadsheet for setting the area. I may have miscalculated. I built my models to full size and then scaled them down to 25% in the choreography. I'm not sure what measurements I used for the calculations. I will guess that more is better than not enough.
I am assuming that the reason for scaling objects in radiosity shots is to keep from maxing out the sample area for really big scenes. Which reminds me, I believe once upon a time, there was some discussion concerning some adverse radiosity effects when scaling glass. I am using glass (with refractive properties) and other than the expected increased render times, I don't see anything wrong yet.
Rob
More to come!
Thanks Rodney!
The camera is pretty wide angle so I can really see how the lights are behaving. It is actually positioned where the bathtub should be. LOL
The lighting is six bulb lights embedded in the chrome plated Light Bar model. My intent is to be able to develop an architechural library of lights that mimic the way lights work in my own house. Coupled with a predertermined modelling style and scale I should be able to create interior design/architechural visualizations without rethinking lighting everytime I build a room (after years of alcohol abuse and three softball/baseball playing kids, I have precious few brain cells left). How's that for a line of bulls**t?!
Anyway, here are a few close-ups of the light bar. It kind of shows the evolution of the lights as well. The bottom left image is the most recent.
Peace,
Rob
Lots more to come!
The camera is pretty wide angle so I can really see how the lights are behaving. It is actually positioned where the bathtub should be. LOL
The lighting is six bulb lights embedded in the chrome plated Light Bar model. My intent is to be able to develop an architechural library of lights that mimic the way lights work in my own house. Coupled with a predertermined modelling style and scale I should be able to create interior design/architechural visualizations without rethinking lighting everytime I build a room (after years of alcohol abuse and three softball/baseball playing kids, I have precious few brain cells left). How's that for a line of bulls**t?!
Anyway, here are a few close-ups of the light bar. It kind of shows the evolution of the lights as well. The bottom left image is the most recent.
Peace,
Rob
Lots more to come!
Is jittering the amount of light bounces?
looks good !
by the way . . . do scenes with radiosity render faster if you have more photon samples?
I'm not sure, but it seems for some scenes that this is true
Could somone verify?
I'm on the edge of my seat for more . . .
Ben
looks good !
by the way . . . do scenes with radiosity render faster if you have more photon samples?
I'm not sure, but it seems for some scenes that this is true
Could somone verify?
I'm on the edge of my seat for more . . .
Ben
| QUOTE (dreamin3d @ May 31 2005, 04:48 PM) |
| I believe I used the calculation spreadsheet for setting the area. I may have miscalculated. |
I see that the way I wrote my instruction is confusing.
If you increase the "photon samples" and reuse the spreadsheet, you will get a new larger sampling area. But it may help to further increase the sampling area anyway. The spreadsheet is used to compute a good starting baseline. Most of the time it is good enough but you can further play with the numbers.
| QUOTE |
| I built my models to full size and then scaled them down to 25% in the choreography. I'm not sure what measurements I used for the calculations. I will guess that more is better than not enough. |
You should always keep your models to full size. Photon mapping wil adapt. When you enter your dimensions in the speadsheet, the numbers you will get are adapted. Another important reason to keep all your model to full size scale is if you ever want to use hair, dynamics, particles, cloths, the dynamic calculations are adjusted to behave correctly at full size scale. You may not plan for that now but if you ever need to do that, you will be glad you already have all your model to proper scale.
Measurements you use for calculations are not important as long as they are consistent. That is you use the same unit of measurement for all your calculations.
| QUOTE (bentothemax @ May 31 2005, 07:23 PM) |
| Is jittering the amount of light bounces? |
Jittering adds randomness to final gathering samplings. It adds noise to the shading but get rid of sampling patterns.
| QUOTE |
| by the way . . . do scenes with radiosity render faster if you have more photon samples? |
No. About any number you increase in photon mapping settings will also increase the render time. However, the number that have the most impact on render time is the Final Gathering samples.
Thanks Yves,
I think I understand. I am constantly reading and rereading your tutorials. I will take a look at this when I get home.
Rob
I think I understand. I am constantly reading and rereading your tutorials. I will take a look at this when I get home.
Rob
Ok, thank you Mr. Poissant
srry, but one last question I had, (now that i know what jittering is) is how to set jittering?
Does AM automaticly do that, or do i have to set it.
In my radiosity properties panel there is no option for jittering
thanks
ben
srry, but one last question I had, (now that i know what jittering is) is how to set jittering?
Does AM automaticly do that, or do i have to set it.
In my radiosity properties panel there is no option for jittering
thanks
ben
Very nice!! I love the POV.
Every time I think I know a bit about Radiosity, then Yves comes along and I feel like I just fell off the turnip truck
Can't wait to see what else you have planned for this image. Love the chrome lightbar in the close-ups!
- pjc
Every time I think I know a bit about Radiosity, then Yves comes along and I feel like I just fell off the turnip truck
Can't wait to see what else you have planned for this image. Love the chrome lightbar in the close-ups!
- pjc
Ben-
Jittering is under the Final Gathering heading in the radiosity properties. Increasing it helps smooth out the geometric patterns in your image. A:M being so mathematically perfect, the ability to make perfectly sharp and flat objects, coupled with the abilility to exactly place objects(square, parallel, etc.) tends to focus photons into neat little patterns. In its extreme, there are times I have seen an obects outline on another, not a shadow or reflection just a color impression. Jittering helps smooth this out.
In my last posting, if you look at the lower right picture, there is a fan shaped shiny area near the ceiling above the light bar. In later pictures I increased the jittering to tone it down. I think I used 200 samples and 20% jittering.
Rob
Jittering is under the Final Gathering heading in the radiosity properties. Increasing it helps smooth out the geometric patterns in your image. A:M being so mathematically perfect, the ability to make perfectly sharp and flat objects, coupled with the abilility to exactly place objects(square, parallel, etc.) tends to focus photons into neat little patterns. In its extreme, there are times I have seen an obects outline on another, not a shadow or reflection just a color impression. Jittering helps smooth this out.
In my last posting, if you look at the lower right picture, there is a fan shaped shiny area near the ceiling above the light bar. In later pictures I increased the jittering to tone it down. I think I used 200 samples and 20% jittering.
Rob
Patrick-
Thanks for the comments! Like I said not as exciting as POPBOT, but anytime we get a chance to pick Yves' brain, its a good thing. Every time I get one of his responses, I spend hours reading his stuff again for more info. Its great having him around, he's like a walking physics encyclopedia.
Peace,
Rob
Thanks for the comments! Like I said not as exciting as POPBOT, but anytime we get a chance to pick Yves' brain, its a good thing. Every time I get one of his responses, I spend hours reading his stuff again for more info. Its great having him around, he's like a walking physics encyclopedia.
Peace,
Rob
Yves-
I went back and reran my calculations. I don't know where my brain went that day, but at the scale I am presently using I should have had a baseline sampling area of 600 not 400! I am rerenderg the basic scene with sample area 750 and samples of 150.
I started playing with the full scale models. I did some calculations using 20000 photons and came up with a value of 3150 for the sampling area. Without actually doing a render, I am guessing that at full size I should probably increase the number of photons for larger scenes to help in the estimation process.
Thanks as always,
Rob
I went back and reran my calculations. I don't know where my brain went that day, but at the scale I am presently using I should have had a baseline sampling area of 600 not 400! I am rerenderg the basic scene with sample area 750 and samples of 150.
I started playing with the full scale models. I did some calculations using 20000 photons and came up with a value of 3150 for the sampling area. Without actually doing a render, I am guessing that at full size I should probably increase the number of photons for larger scenes to help in the estimation process.
Thanks as always,
Rob
| QUOTE (dreamin3d @ Jun 2 2005, 06:16 AM) |
| I started playing with the full scale models. I did some calculations using 20000 photons and came up with a value of 3150 for the sampling area. Without actually doing a render, I am guessing that at full size I should probably increase the number of photons for larger scenes to help in the estimation process. |
No. The number of photons is not related to the scene size or to the object sizes. All the numbers are size relative. In other words, if you scale your scene up. and use the Excel sheet again with your new room dimensions, you wil get a new proportionately larger sample size. It is all relative.
Number of photons required for a scene is dependent on some scene complexity metrics. I don't know how to put that because it is all a question of selecting the best compromise between speed and quality but here are the basics:
Ideally, for any surface to contribute in indirect illumination, the photon sampling area should be at least as small as to fit in the surface. That is an ideal situation because in most real scene instance, this would be impractical as it would require way too many photons.
In practice, this means that you take the smallest surface in your scene, measure its smallest dimension and that would be your ideal sampling area.
Let's take your scene as example. Lets suppose that the smallest size surface in your scene is the side of the doorframe. If your room is in real size, that doorframe thickness would be around 1.5cm. that would mean that the photon sampling area would have to be no larger than 150 (photon sampling area is in 1/100th of cm). Well, to get that sample area, you would need 5M photons.
This is clearly not practical. You can increase the number of photons to the maximum of 1M but still the sampling area will be too large. So the solution is a paradoxical one. It is to actually increase the sampling area even if you don't change the photon samples. This will have the effect of bluring even more the photons contributions and help remove the uneven illumination that we see next to the doorframe.
The larger the sampling area, the more ambiance like the indirect illumination. You could use very few photons and very large sampling area to get almost the equivalent of setting an ambiance value on your surfaces. As you increase the number of photons in your scene, you also improve the exactness of the indirect illumination calculation.
The bottom line is if your goal is to simulate your lighting setups in your house, you would be better off by setting your number of photons to the maximum and then play with the other settings.
So after this long and contorted preamble, the answer to your question is...
yes.
So many variables, so little time.
As usual I have much to consider. This is a very clear explanation as always.
Rest assured, that every posting gets me a little closer to where I am trying to go.
Definetly more things to experiment with.
Peace,
Rob
Ouch! I fell off the turnip truck again.
As usual I have much to consider. This is a very clear explanation as always.
Rest assured, that every posting gets me a little closer to where I am trying to go.
Definetly more things to experiment with.
Peace,
Rob
Ouch! I fell off the turnip truck again.
Here is the room rendered again 750 sample area and 150 samples. I haven't had a chance to render a full scale choreography yet. I also turned of the mirror for now. It adds 1 hour to each pass(9 passes). Anyway, it does look better.
I do seem to notice that at these settings the lights appear to be much hotter near the ceiling.
Rob
I do seem to notice that at these settings the lights appear to be much hotter near the ceiling.
Rob
It is likely that I will be out of circulation for a bit. Softball, plus I am going to continue modelling and experimenting. Yves has given me much to contemplate.
However I wanted to show one more series of pictures of what I have to date.
Here is a collage of the beginnings of the sink, the cabinet and the plumbing. The bottom right is a pretty good representatation of the fixtures and countertop.
Keep in mind it hasn't been that pretty since I built the house!
Peace,
Rob
Radiosity Rules!
However I wanted to show one more series of pictures of what I have to date.
Here is a collage of the beginnings of the sink, the cabinet and the plumbing. The bottom right is a pretty good representatation of the fixtures and countertop.
Keep in mind it hasn't been that pretty since I built the house!
Peace,
Rob
Radiosity Rules!
Something I have noticed is, the more stuff in the room the better the light looks.
And Yves is dead on. Increasing the sample area and samples has definitely improved not just the smaller areas of concern, but the overall quality of the pic.
Thanks again,
Rob
And Yves is dead on. Increasing the sample area and samples has definitely improved not just the smaller areas of concern, but the overall quality of the pic.
Thanks again,
Rob
Wow. I can't wait for you to finish modeling the rest of the bathroom!
The sink detail picture is so nice, it could stand on it's own as an image!
I don't know about the 2.2 gamma tho...somewhere in between feels better to me, but that's just me...
- pjc
The sink detail picture is so nice, it could stand on it's own as an image!
I don't know about the 2.2 gamma tho...somewhere in between feels better to me, but that's just me...
- pjc
Sheesh!
Who needs Maxwell, when A:M radiosity does such a good job?
Who needs Maxwell, when A:M radiosity does such a good job?
Thanks guys,
I agree, the gamma does seem to be a bit harsh, whitewashed. My hope is, at least it holds true in real life, that as I add models and color, the harshness will tone down. Kind of like an empty room in a new house looks different when you put furniture in it. We shall see.
Peace,
Rob
I agree, the gamma does seem to be a bit harsh, whitewashed. My hope is, at least it holds true in real life, that as I add models and color, the harshness will tone down. Kind of like an empty room in a new house looks different when you put furniture in it. We shall see.
Peace,
Rob
The sink and counter looks really good. Good work.
You will no doubt receive comments about the gamma being too washed out. Some people, and in particular all Mac users, already have an already cranked up gamma on their monitor. But still the vast majority of users don't have any gamma correction on their computer. That is why the sRGB standard requires a gamma of 2.2.
You will no doubt receive comments about the gamma being too washed out. Some people, and in particular all Mac users, already have an already cranked up gamma on their monitor. But still the vast majority of users don't have any gamma correction on their computer. That is why the sRGB standard requires a gamma of 2.2.
Hey Hash Folks,
I told you I'd be slow here. I don't have a lot of time right now.
But I did have a couple of hours tonight and I started working on cabinet doors. I know its not lighting but I wanted to let everyone know I am still working on the project. Here is the frame for the cabinet doors.
Was thinking, if this needs to go into the WIP I can continue over there.
Peace,
Rob
I told you I'd be slow here. I don't have a lot of time right now.
But I did have a couple of hours tonight and I started working on cabinet doors. I know its not lighting but I wanted to let everyone know I am still working on the project. Here is the frame for the cabinet doors.
Was thinking, if this needs to go into the WIP I can continue over there.
Peace,
Rob
OOPS!
Thats actually the fit detail. Here is the exploded frame and the wire
Once I get a few more things modelled I'll shoot another render.
Rob
Thats actually the fit detail. Here is the exploded frame and the wire
Once I get a few more things modelled I'll shoot another render.
Rob
wow. excellent detail in the model! Can't wait to see more rad(iosity) renders!
- pjc
- pjc
You got enough details to animate an exploded view. Unless that is your intent. you should realize that that many details will just slow down your render and your working with your scene. I usually don't go into that many details when they don't need to show on final renders.
Other than that, I like the care you take in modeling the details. But at the same time I observe that you did not model the bevels on the edges.
Other than that, I like the care you take in modeling the details. But at the same time I observe that you did not model the bevels on the edges.
Hey,
not much activity here lately, any new updates?
This was looking really really good, but i havent seen/heard anything for a while ?
Looking forward to it!
Ben
not much activity here lately, any new updates?
This was looking really really good, but i havent seen/heard anything for a while ?
Looking forward to it!
Ben
Ben,
Time is my enemy. I say that everyday.
Work, softball, migrating to a new computer and migrating my wife to a new computer have been bogging me down. It seems every time I get near my machine all of my kids showup(bright eyed, grinning, and logging on to their machines) ready to play a couple of rounds of network Battlefield 1942, the best video game on the planet.
I am currenty working on the toilet model for the bathroom. Of course for me, its like reinventing the wheel. I have so much going on, that I easily suffer from " Mental Spline Blockage". I'll get there, I promise.
I can't wait to get back to rendering.
Have I mentioned that time is my enemy,
Rob
Time is my enemy. I say that everyday.
Work, softball, migrating to a new computer and migrating my wife to a new computer have been bogging me down. It seems every time I get near my machine all of my kids showup(bright eyed, grinning, and logging on to their machines) ready to play a couple of rounds of network Battlefield 1942, the best video game on the planet.
I am currenty working on the toilet model for the bathroom. Of course for me, its like reinventing the wheel. I have so much going on, that I easily suffer from " Mental Spline Blockage". I'll get there, I promise.
I can't wait to get back to rendering.
Have I mentioned that time is my enemy,
Rob
| QUOTE (dreamin3d @ Aug 12 2005, 09:31 AM) |
| I am currenty working on the toilet model for the bathroom. Of course for me, its like reinventing the wheel. I have so much going on, that I easily suffer from " Mental Spline Blockage". I'll get there, I promise. |
I don't know if you're going for a special look for this toliet - but I believe there's a toliet model on the new extra CD. (as well as many other bathroom accoutrements - thanks I believe to Zach BG)
Thanks Nancy!
I am actually still using version 10. I intend to update to 12 very soon. What I have seen of it is awesome.
So for now I'll just keep plugging along.
Peace,
Rob
I am actually still using version 10. I intend to update to 12 very soon. What I have seen of it is awesome.
So for now I'll just keep plugging along.
Peace,
Rob
Yves,
I wasn't ignoring you. Been a little busy (of course).
Here is my thinking on this.
I am currently a designer (I'm actually several other things that I am far better at, but for now in my life's journey I play the part of a marine electrical designer),hence I see things in high detail. This is coupled with the fact that I'm not real good at drawing decals. By modelling in high detail I can apply wood grained materials(that I happily learned from your tutorial) with the grain going in the right direction and the end grain showing properly in the corners like stained cabinets do. This will also allow me to use the model in any situation, near or far.
Which brings me to the contradiction you have pointed out. You're right, it does seem silly to model in such detail and not bevel sharp edges. However, I have noticed in rad renders (this neat term coined by Patrick Clarke), that since light bounces all over the place, knife edgees seem to not show much different from very slightly beveled edges. I did some tests once upon a time, but I haven't located the pictures I came up with.
When I start beveling everything, the time it takes for me to model something goes up exponentially. So for me the trade off is okay ....I guess. So for the purposes of my Bathroom Experiment I decided to not 'relieve' sharp edges and see what happens.
I must say you don't miss anything! And I thank you very much for that. I will try to get my butt in gear and put out something worth looking at real soon.
And thanks to all for their comments, its great to be even a small part of such a talented and positive community of HASH folks!
Hereby ends another view into my dented mind.
Peace,
Rob
I wasn't ignoring you. Been a little busy (of course).
Here is my thinking on this.
I am currently a designer (I'm actually several other things that I am far better at, but for now in my life's journey I play the part of a marine electrical designer),hence I see things in high detail. This is coupled with the fact that I'm not real good at drawing decals. By modelling in high detail I can apply wood grained materials(that I happily learned from your tutorial) with the grain going in the right direction and the end grain showing properly in the corners like stained cabinets do. This will also allow me to use the model in any situation, near or far.
Which brings me to the contradiction you have pointed out. You're right, it does seem silly to model in such detail and not bevel sharp edges. However, I have noticed in rad renders (this neat term coined by Patrick Clarke), that since light bounces all over the place, knife edgees seem to not show much different from very slightly beveled edges. I did some tests once upon a time, but I haven't located the pictures I came up with.
When I start beveling everything, the time it takes for me to model something goes up exponentially. So for me the trade off is okay ....I guess. So for the purposes of my Bathroom Experiment I decided to not 'relieve' sharp edges and see what happens.
I must say you don't miss anything! And I thank you very much for that. I will try to get my butt in gear and put out something worth looking at real soon.
And thanks to all for their comments, its great to be even a small part of such a talented and positive community of HASH folks!
Hereby ends another view into my dented mind.
Peace,
Rob
[attachmentid=18151]Well, just like a bad check, I'm back.
I just purchased version 13 and I can't wait to dive in: however, I promised myself that I would at least finish what I started. I'll probably finish somewhere around A:M 20.
Here are the latest from the bathroom. I've added the toilet and the tub. All I really need now is the decor.
Seems that radiosity has really cooled down since the advent of HDRI, IBL and AO, but here it is anyway. I stopped the render at 8 passes (24 hours). I want to keep modelling, while I can.
I think I need to tone down my lights the gamma corrected image is a little washed out.
Peace,
Rob
I just purchased version 13 and I can't wait to dive in: however, I promised myself that I would at least finish what I started. I'll probably finish somewhere around A:M 20.
Here are the latest from the bathroom. I've added the toilet and the tub. All I really need now is the decor.
Seems that radiosity has really cooled down since the advent of HDRI, IBL and AO, but here it is anyway. I stopped the render at 8 passes (24 hours). I want to keep modelling, while I can.
I think I need to tone down my lights the gamma corrected image is a little washed out.
Peace,
Rob
Very realistic, Rob... I make that as a Delta Tub shower handle and a Kohler toilet? Nope maybe not a Kohler... many Kohlers have the flush handle on the side but the tank looks Kohler like.
Charlie<<< plumbing trivia expert wannabe
Charlie<<< plumbing trivia expert wannabe
I'm not even sure. I'll have to go downstairs and look.Thanks for the compliment,
Rob
I'm really glad to see some further expirimentation with this topic. The scene is also very nice looking! Radiosity is one of those lovely tidbits in AM I wish I had way more time to play with!
-Ethan
-Ethan
Thanks Ethan.
More soon.
Cheers,
Rob
More soon.
Cheers,
Rob
[attachmentid=18257][attachmentid=18256]Gotta love A:M, you can stick a camera just about anywhere. Here is a shot of my way over modelled toilet.
Peace,
Rob
Peace,
Rob
Incredibly realistic! I love it! One thing though: your bathroom doesnt seem to have any other lights than the ones above the sink. That said, i believe your radiosity intensity is too high. Lights like thoes above the sink would cast hard shadows, and I can't find any hard shadows anywhere in your renders. Lowering the radiosity intensity and raising the light intensities would be best. 
Also, 24 hours seems like a very long time to render such a small image. What settings are you using for the radiosity?
Amazing modeling, Rob.
Also, 24 hours seems like a very long time to render such a small image. What settings are you using for the radiosity?
Amazing modeling, Rob.
Well, the important thing here is: what kind of tub and shower faucet, what brand of toilet?
Sheesh, either go down and check, man, or make a bigger render so I can zoom in!
Sheesh, either go down and check, man, or make a bigger render so I can zoom in!
aren't they all generic? i thought they only made one of each type of plumbing utility, right???
QUOTE
I'm really glad to see some further expirimentation with this topic.
Well I am certainly glad everyone is still into this stuff. Believe me I wish I had more time as well. And thanks to all for the compliments.
QUOTE
your bathroom doesnt seem to have any other lights than the ones above the sink.
That's correct. It is as close a representation of the bathroom in my own home as I can make, which is lit by only six bulbs.
QUOTE
Lights like thoes above the sink would cast hard shadows, and I can't find any hard shadows anywhere in your renders
Surprisingly enough, the real life lights don't cast hard shadows. It has been over a year since I first built the light rig, but I'm pretty sure I took that into account. The bulbs are frosted white, as opposed to clear, and are a relatively low wattage. If you look ealier in the post at my early light bar test renders, the real life bulbs actually look pretty much like the A:M renders (I was kinda proud of that).
QUOTE
i believe your radiosity intensity is too high.
Patrick has pinged me for that too.
I agree that something is definitely to intense. For my recent renders, I dropped the light intesity from 50% to 45%, but I don't see much difference. I have been stopping the render when I get home from work, then I apply the Gamma correction with Paint Shop Pro. I've noticed that a Gamma lower than 2.2 looks better to me (1.4 for the toilet shot), so something must be to bright. But, I thought radiosity intensity should always be 100%? Any insight on this would be most welcome.QUOTE
24 hours seems like a very long time to render such a small image.
Yeah, and that's only between 4 and 8 passes out of 25!
750 sample area
150 samples
20000 photons
200 final gathering
20 jitter
I think that's right. I will look again when I get home.
Thanks again for the compliments.
Peace,
Rob
QUOTE
Well, the important thing here is: what kind of tub and shower faucet, what brand of toilet?
HA! HA! I forgot! The plumbing fixture are all MOEN (shower and sink). The toilet marking has worn off. It starts with a BRO.....Broan, Brower, Brown......No, I don't think BROWN would be a good name.
More soon!
Rob
Dan,
Something I forgot to add about the render times. There are (3) things that are jacking up the render times. I have ordered them from, what I believe, least effect to worst.
Number of patches - I'm not sitting at my machine at home, but I know that a couple of my models a just a little over modelled (the cabinets for instance - see my earlier posts). Thats my own fault.
Reflection - While I haven't done much texturing in this scene yet, I have played a little bit. Its easy to see that there are objects that have reflectivity. Reflectivity increases render time as does the number of reflections (currently set to 5 and I may have to bump it up to 6 or 7).
Refraction - Probably the big one. The mirror, the bathtub handle, and knob all have refraction. In some earlier test renders, removing just the mirror cut render time almost in half.
I suppose its the price I pay for trying to make something look real.
Peace,
Rob
Something I forgot to add about the render times. There are (3) things that are jacking up the render times. I have ordered them from, what I believe, least effect to worst.
Number of patches - I'm not sitting at my machine at home, but I know that a couple of my models a just a little over modelled (the cabinets for instance - see my earlier posts). Thats my own fault.
Reflection - While I haven't done much texturing in this scene yet, I have played a little bit. Its easy to see that there are objects that have reflectivity. Reflectivity increases render time as does the number of reflections (currently set to 5 and I may have to bump it up to 6 or 7).
Refraction - Probably the big one. The mirror, the bathtub handle, and knob all have refraction. In some earlier test renders, removing just the mirror cut render time almost in half.
I suppose its the price I pay for trying to make something look real.
Peace,
Rob
QUOTE(dreamin3d @ Jul 14 2006, 07:27 AM) 
QUOTE
i believe your radiosity intensity is too high.
Patrick has pinged me for that too.
I agree that something is definitely to intense. For my recent renders, I dropped the light intesity from 50% to 45%, but I don't see much difference. I have been stopping the render when I get home from work, then I apply the Gamma correction with Paint Shop Pro. I've noticed that a Gamma lower than 2.2 looks better to me (1.4 for the toilet shot), so something must be to bright.Make sure to take your own montor gamma adjustment into account. If your own monitor already had a gamma adjustment of 1.8, for instance, an additional gamma adjustment of 2.2 will definitely make the render look way too washed out.
For the same reason, you should always be cautious when you get feedbacks from other viewers about your render being too dark or too bright unless they can tell how their own monitor is gamma adjusted. Since a lot a people don't know anything about gamma adjustment let alone monitor calibration, those dark-bright feedbacks are just shots in the dark. Know your own monitor setup and trust only your own monitor setup.
QUOTE
I thought radiosity intensity should always be 100%? Any insight on this would be most welcome.
Yes. Radiosity intensity should always be 100% unless you are looking for a special effect. You could always drop the object surface radiance to 95% but this is basicaly irrelevant. What you really need to take care of are the surface colors. When using radiosity, the surface color represents reflectance. That is the percentage of light that is reflected off the object surface. In the real world, pure 100% reflectance does not exist, just as well as pure 0% reflectance does not exist. So that rules out white and black for surface colors. Even the purest and brightest white that you can get out of a manufactured object if maxed at about 95%. As a rule of thumb, surface colors should be restrained to between 10% and 90% on any of the Red, Green and Blue channels.
Hmm.....So that means my brightest white would be:
229
229
229
Interesting. I backed off of the 'pure' settings, obviously I didn't go far enough.
I will have to run through your monitor calibration tut as well. I have recently changed from a single CRT to dual LCDs and there is a big difference.
Thanks as always,
Rob
Oh...here are a couple more renders:
[attachmentid=18308][attachmentid=18309]
229
229
229
Interesting. I backed off of the 'pure' settings, obviously I didn't go far enough.
I will have to run through your monitor calibration tut as well. I have recently changed from a single CRT to dual LCDs and there is a big difference.
Thanks as always,
Rob
Oh...here are a couple more renders:
[attachmentid=18308][attachmentid=18309]
QUOTE(dreamin3d @ Jul 14 2006, 12:30 PM)
I will have to run through your monitor calibration tut as well. I have recently changed from a single CRT to dual LCDs and there is a big difference.
Just note that LCD monitors attached to a laptop are way too anisotropically lighted to be trusted in any ways when doing lighting and texturing. Don't even think of trying calibrating them. That would be a complete lost of time. But a good laptop LCD monitor can be trusted. I have a BenQ which is already calibrated at 1.8 gamma (the Mac standard) and I can rely on it for lighting and color work.
QUOTE
Oh...here are a couple more renders:
This is looking really nice.
QUOTE
HA! HA! I forgot! The plumbing fixture are all MOEN (shower and sink). The toilet marking has worn off. It starts with a BRO.....Broan, Brower, Brown......No, I don't think BROWN would be a good name.
Well then! Guess I won't be the Star for the program This Old Bathroom, Dang!
The renders are looking great, BTW. And thanks for checking, my trivia answer!
Ah, so there are the hard shadows. I guess your intensities were fine after all. For some reason i was thinking the lights were much further behind the camera, so i was looking for hard shadows on the wall next to the toilet. Little did i know they weren't there.
I apply the Gamma correction with Paint Shop Pro
Wont you lose color data if you do it this way? I thought animation:master kept track of floating point color data, which is why the exposure and gamma correction post effects work so well and look so clean.
yves: I still dont fully understand the need for gamma correction. Are you trying to make the image appear correctly on a gamma corrected moniter? If you dont have a corrected moniter, which I believe you said most people do not, than isnt it just a lost cause? To the majority of people, your images are going to appear washed out anyways. Also, don't gamma corrected moniters have a hard time reaching true black? The ones I've seen can get pure white pretty good, but black turns into a rather annoying 75% grey.
How does one go about calibrating their moniter anyways?
QUOTE(dreamin3d @ Jul 14 2006, 04:27 AM)
I apply the Gamma correction with Paint Shop Pro
Wont you lose color data if you do it this way? I thought animation:master kept track of floating point color data, which is why the exposure and gamma correction post effects work so well and look so clean.
yves: I still dont fully understand the need for gamma correction. Are you trying to make the image appear correctly on a gamma corrected moniter? If you dont have a corrected moniter, which I believe you said most people do not, than isnt it just a lost cause? To the majority of people, your images are going to appear washed out anyways. Also, don't gamma corrected moniters have a hard time reaching true black? The ones I've seen can get pure white pretty good, but black turns into a rather annoying 75% grey.
How does one go about calibrating their moniter anyways?
Dan,
This link isn't the one I was looking for but it does have some information on the subject.
I think Yves Poissant did one of his forum mini-tutorials on the subject. Its out there somewhere...
Talk of Gamma Settings
Read down about half way for the Gamma discussion.
I'll see if I can find the mini tutorial on Monitor adjustment.
This link isn't the one I was looking for but it does have some information on the subject.
I think Yves Poissant did one of his forum mini-tutorials on the subject. Its out there somewhere...
Talk of Gamma Settings
Read down about half way for the Gamma discussion.
I'll see if I can find the mini tutorial on Monitor adjustment.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.