Chaos Corona Forum
Chaos Corona for 3ds Max => [Max] General Discussion => Topic started by: Dionysios.TS on 2016-10-10, 12:17:42
-
Hi all, I am working on a big international project and I am having some issues by using IES file information in Corona. I really need an accurate simulation but I have some strong doubts on how the engine treats the IES files and their intensity values.
In the past years I always used the native 3ds Max photometric lights where once loaded an IES file, the intensity values are driven automatically so the "cd" and "lm" values are written as in the file by the manufacturer.
In Corona of course this doesn't happen, so the user must manually do this and here is where I find a big difference or BUG?!?!?
Take a look at the IES Comparison image I uploaded here, the 3ds Max Photometric Light reads automatically and correctly the "lm" value. Now if we want to get the same result in Corona the "lm" value has to be set too way low?!? How this is possible? This kind of difference is very important for a natural based result!
The IES file has a 2.217,831lm value and Corona must be set to 47,123lm for the same intensity result!
Also in the 3ds Max Photometric Light's menu, if you change from "lm" to "cd" the values are correct but in Corona, if you change from "lm" to "cd" and you put manually the values from the IES data information, the result is different! O.o
I included a iGuzzini IES file here and from the IES info I read:
5.129.6cd which are 2.291,772lm
In Corona, 5.129.6cd has different intensity results if compared to 2.291,772lm. What a mess... :(
Looking forward for an answer.
Thanks,
Dionysios -
-
'Might' be because in CoronaLight, IES is still area light, which is super-odd, like, what kind of size do you create analytical point light :- ) ?
-
I just need to simulate the correct lighting effect in the space. The architects here need to see a "realistic" output and not my artistic point of view and is very understandable.
If I use also a Corona Sun element in the scene and my IES files are so wrong with their values, well, the result is NOT ok and everything becomes messy.
-
I don't know if this is a coincidence but Corona's 47,1 squared (47,1^2=2218) is very close to the 'real' value displayed the photometric light. Maybe test it with other values and if it's roughly the formula you have a rough estimate at least.
-
I don't know if this is a coincidence but Corona's 47,1 squared (47,1^2=2218) is very close to the 'real' value displayed the photometric light. Maybe test it with other values and if it's roughly the formula you have a rough estimate at least.
Didn't noticed that... I am curious to know what the dev team thinks about.
Thanks guys for your answers.
Dionysios -
-
It's not exactly the square so it really could be by coincidence, best would be to verify with different IES files.
-
Already tried and they work all the same way unfortunately.
-
Any news from dev team here?
Sorry for stressing you guys but I would like to know if you know anything about these issues.
Thanks,
Dionysios -
-
'Might' be because in CoronaLight, IES is still area light, which is super-odd, like, what kind of size do you create analytical point light :- ) ?
Maybe I'm missing something... but I thought it only became an area light when you uncheck 'keep sharp patterns? Otherwise it stays a point light regardless of the size of the shape, is that not the case?
-
'Might' be because in CoronaLight, IES is still area light, which is super-odd, like, what kind of size do you create analytical point light :- ) ?
Maybe I'm missing something... but I thought it only became an area light when you uncheck 'keep sharp patterns? Otherwise it stays a point light regardless of the size of the shape, is that not the case?
Now that you mention it, that might be the case ? I honestly didn't know. So if I check ON "keep sharp patterns" (it's stupid name then no?), the size of the light doesn't matter ? If yes that's wonderful, and I've been misled.
-
Yah, here are the differences between 2 long rectangle ies area lights with 'keep sharp patterns on' the left, and right with it off.
*EDIT
..... but as for the topic at hand. Corona and the 3ds max do treat the conversion from candelas to lumens differently.
in corona 1 cd = 3.142 lumens
in 3ds max photometric 1cd = 12.566 lumens
I also pulled this from the lumens wikipedia page
The lumen is defined in relation to the candela as
1 lm = 1 cd⋅sr.
A full sphere has a solid angle of 4π steradians,[1] so a light source that uniformly radiates one candela in all directions has a total luminous flux of 1 cd × 4π sr = 4π cd⋅sr ≈ 12.57 lumens.[2]
so something is happening differently between the two. Hopefully Ondra can illuminate what's going on ;P.
-
Glad i read this. i was about to go into work tomorrow and light a 50k sqft building with ies lights and play around with lightmixer. maybe ill postpone that project until one of the team chimes in :D
-
Yah, here are the differences between 2 long rectangle ies area lights with 'keep sharp patterns on' the left, and right with it off.
*EDIT
..... but as for the topic at hand. Corona and the 3ds max do treat the conversion from candelas to lumens differently.
in corona 1 cd = 3.142 lumens
in 3ds max photometric 1cd = 12.566 lumens
I also pulled this from the lumens wikipedia page
The lumen is defined in relation to the candela as
1 lm = 1 cd⋅sr.
A full sphere has a solid angle of 4π steradians,[1] so a light source that uniformly radiates one candela in all directions has a total luminous flux of 1 cd × 4π sr = 4π cd⋅sr ≈ 12.57 lumens.[2]
so something is happening differently between the two. Hopefully Ondra can illuminate what's going on ;P.
Exactly, this is the point and I think is a mess.
Can you guys try to check if your IES files have the same differences? Would be nice.
Thanks in advance,
Dionysios -
-
yep It's not exclusive to ies lights though, corona lights and photometric lights behave differently at different lumens.
But here is the thing... I think the corona light intensity is actually correct using lumen values.
I think the issue is that corona interprets photometric lights incorrectly based on some quick tests I did using photographic exposures, but I could be wrong and overlooked something.
-
Didn't received any news from the dev team IES problem and neither another post about Kelvin colour problems (https://forum.corona-renderer.com/index.php/topic,13377.15.html)...
I guess guys you're busy these days but would be nice to know if there is something wrong with both issues or not.
Thanks!
Dionysios -
-
Any news? Still nothing... Houston?
-
I've noticed that 3dsmax photometric lights intensity doesn't change when size changes.
Corona lights do which I guess is correct behavior.
I've managed to get similar results if the size of corona lights is the same as 3dsmax light and I put the 3dsmax lumen value in the corona intensity in lux, not lumen or candela.
Not sure it works in all cases but for iGuzzini IES files, it's really close.
sorry for my english :p
-
Has the 1.5.1 hotfix addressed any of the problems you had with the IES?
-
I don't think so, unfortunately...
-
Hi. Don't know if this works with IES, but in my tests a photometric light set to a preset lightbulb matches a Corona light (sphere mode) exact when then "cd" values in the photometric light is the same number as the "lm" values in the Corona light, and everything else is set equal.
So if the photometric light bulb reads 161,0 cd and the Corona spare light reads 161,0 lm, then they will match.
Haven't tried it with IES, but maybe?
Hope it helps
-
Could it be that simple... I hope so!
-
Hi guys,
First of all, sorry for no replies. I don't know why it happened. I must have overlooked the topic, and also Ondra & co are very busy right now.
Could you please list all issues from this thread? It got a little chaotic, right now I wanted to report this to Ondra, but I am not even sure what should be checked. Is the original issue that 3ds Max native lights work fine (correct; as expected) with IES lights, and Corona lights seem to be incorrect?
-
Hi guys,
First of all, sorry for no replies. I don't know why it happened. I must have overlooked the topic, and also Ondra & co are very busy right now.
Could you please list all issues from this thread? It got a little chaotic, right now I wanted to report this to Ondra, but I am not even sure what should be checked. Is the original issue that 3ds Max native lights work fine (correct; as expected) with IES lights, and Corona lights seem to be incorrect?
No worries, we know you're busy guys... :)
In a few words, many of us checked this behavior in Corona:
- Get an IES file.
- Use its lumen, lux or cd value with a 3ds Max native photometric light = intensity result is physically correct.
- Same file, use its lumen, lux or cd value with a Corona light = intensity is very low.
Result: To get the same lighting intensity we need to multiplier a lot the IES value which is not a good thing as you can imagine.
So, why Corona does this? Why it doesn't read the REAL WORLD values in a 1:1 correct mode?
We found out that also Vray has the same sort of problem and they add an option to scale the IES values for it. Probably is a 3ds Max scale bug or defect? Who knows.
Dionysios -
-
Hi maru, here my understanding of the problem.
A. They are assuming that what they see when they render a Photometric Light (IES) in Corona is correct.
B. They want to match the same result with a Corona Light (IES), but in order to do that, they have to reduce the intensity multiplier by a lot (see the image in the first post (https://forum.corona-renderer.com/index.php/topic,13397.msg86385.html#msg86385)).
To further sum this up, the problems are two:
1. There is no consistency between Photometric Light (IES) and Corona Light (IES).
2. What is the correct result between the two?
Regarding point 1, I think there is a different interpretation of the numeric intensity values.
If that is the case, a simple solution would be scaling one of the two, like VRay does when you check "Use 3ds Max photometric scale".
Regarding point 2, I think the correct result is with Corona Light (IES) with proper intensity values.
I have attached a test I made a while ago.
On the left there is a real world reference (2 cheap light sources).
On the right a simulated environment with 2 Corona Light that have correct light intensity and position in space relative to the subject.
The settings and exposure of the virtual camera also match the real camera.
-
Hi maru, here my understanding of the problem.
A. They are assuming that what they see when they render a Photometric Light (IES) in Corona is correct.
B. They want to match the same result with a Corona Light (IES), but in order to do that, they have to reduce the intensity multiplier by a lot (see the image in the first post (https://forum.corona-renderer.com/index.php/topic,13397.msg86385.html#msg86385)).
To further sum this up, the problems are two:
1. There is no consistency between Photometric Light (IES) and Corona Light (IES).
2. What is the correct result between the two?
Regarding point 1, I think there is a different interpretation of the numeric intensity values.
If that is the case, a simple solution would be scaling one of the two, like VRay does when you check "Use 3ds Max photometric scale".
Regarding poin 2, I think the correct result is with Corona Light (IES) with proper intensity values.
I attach a test I made a while ago.
On the left there is a real world reference (2 cheap light sources).
On the right a simulated environment with 2 Corona Light that has correct light intensity and position in space relative to the subject.
The settings and exposure of the virtual camera also match the real camera.
ps. Hi Dionysios, I'm Matteo :)
Ciao Matteo!!! :D
Thanks for writing here in more detail everything.
I really have the doubt who has the correct results. The problem came out while I was using the Corona Sun system and an IES file in the same time. I couldn't believe that a light with 5.065cd (as noted from the manufacturer) was practically almost turned off in the scene. So I had to crunch up the values to start seeing it's effect. Then I compared both lights together, 3ds Max IES and Corona IES and I found the big difference. I work with IES files since Lighscape and I know how they should work. Something is wrong...
PS. Te come va???
-
As you can see in my images, the V-Ray parameter "3ds max photometric scale", doesn't do that.
V-Ray IES and 3ds Max IES have, always, the same intensity!
From what I can understand thanks to my test, is that Corona doesn't handle 3ds Max photometric intensity very well.
(http://www.francescolegrenzi.com/Temp/Corona/107_IES_VRay.jpg)
(http://www.francescolegrenzi.com/Temp/Corona/108_IES_Corona.jpg)
-
What I see is that the intensity value inside a VRayLight and the intensity value inside a Photometric Light are interpreted in the same way by the renderer (VRay). They are consistent with each other. I also see that "3ds max photometric scale" is internally scaling both values, so that the renderer can interpret the numbers in two different ways: VRay way and 3dsMax way.
What I am saying is that the intensity value inside a CoronaLight and the intensity value inside a Photometric Light are not interpreted in the same way by the renderer (Corona). They are not consistent.
If you want them to be consistent you have to internally scale one of them and match the other.
I hope this explains better my point.
-
What I see is that the intensity value inside a VRayLight and the intensity value inside a Photometric Light are interpreted in the same way by the renderer (VRay). They are consistent with each other. I also see that "3ds max photometric scale" is internally scaling both values, so that the renderer can interpret the numbers in two different ways: VRay way and 3dsMax way.
What I am saying is that the intensity value inside a CoronaLight and the intensity value inside a Photometric Light are not interpreted in the same way by the renderer (Corona). They are not consistent.
If you want them to be consistent you have to internally scale one of them and match the other.
I hope this explains better my point.
Yes but we shouldn't scale anything in reality. The SW should do this automatically. What happens if one day I have to use Max with system setup in Inches and another day in meters? Does this influence the scaling too? If yes, then is crazy! :)
-
internally scaling = the SW should do this automatically
I'm speaking code, you are speaking artist. :D
Anyway, we are discussing on my speculations, the important part of my post was this:
Hi maru, here my understanding of the problem.
A. They are assuming that what they see when they render a Photometric Light (IES) in Corona is correct.
B. They want to match the same result with a Corona Light (IES), but in order to do that, they have to reduce the intensity multiplier by a lot (see the image in the first post (https://forum.corona-renderer.com/index.php/topic,13397.msg86385.html#msg86385)).
To further sum this up, the problems are two:
1. There is no consistency between Photometric Light (IES) and Corona Light (IES).
2. What is the correct result between the two?
I was just trying to help maru understanding what the problems are. :)
-
internally scaling = the SW should do this automatically
I'm speaking code, you are speaking artist. :D
Anyway, we are discussing on my speculations, the important part of my post was this:
Hi maru, here my understanding of the problem.
A. They are assuming that what they see when they render a Photometric Light (IES) in Corona is correct.
B. They want to match the same result with a Corona Light (IES), but in order to do that, they have to reduce the intensity multiplier by a lot (see the image in the first post (https://forum.corona-renderer.com/index.php/topic,13397.msg86385.html#msg86385)).
To further sum this up, the problems are two:
1. There is no consistency between Photometric Light (IES) and Corona Light (IES).
2. What is the correct result between the two?
I was just trying to help maru understanding what the problems are. :)
:D :D :D
As you wrote seemed each of us should do manually this. No problem with math but I just don't want to. :)
I think that now he understood everything perfectly!
-
I have attached a test I made a while ago.
On the left there is a real world reference (2 cheap light sources).
On the right a simulated environment with 2 Corona Light that have correct light intensity and position in space relative to the subject.
The settings and exposure of the virtual camera also match the real camera.
Karnak, this test is done with IES light profile ? (And if so, IES left as area light or as point light (keep sharp patterns).
-
Hi. The test was done with IES "Keep sharp patterns" on.
The IES file was not of the same lamp, I searched for a similar one, but lumens output was correct.
It is not an accurate test by all means. What are your thoughts?
-
Great :- ) Just wanted to know when the IES is accurate (keep sharp patterns or not). This gave me the answer :- )
-
Im confused. So is there a way in max using corona of getting accurate IES Intensity or not?
-
So here are the problems:
1. When you load an ies file in a corona light, it doesn't automatically use the manufacturers lumen value that's stored in the file.
(using a 3ds photometric light, loads the correct lumen value)
2. Corona lights and photometric lights are inconsistent from each other when they share the same dimensions and lumen value.
(corona lights are significantly brighter in this case.) The confusion is arising from not knowing which light is physically accurate.
3. When switching from candela units to lumens they produce inconsistent conversion values.
(Corona lights don't take into account the shape of the light when converting from candelas to lumens.)
a fully exposed sphere of light should be 1 cd = 12.566 lumens (as it is in 3ds photometric lights)
all corona lights regardless of the shape of the light (from dropdown menu or light distribution data from ies files) convert 1cd = 3.142 lumens
-
So if i use photometric lights in corona i can get a somewhat accurate simulation still?
-
No, I don't think so. I could be wrong but based on some quick tests I've done, I think corona is interpreting the lumen/candela values from 3ds max photometric lights incorrectly because while 3ds max takes the shape of the light into account, corona does not.
If you want the most accurate light for a pot light as of this moment this is probably the most accurate; load up your ies files in a corona light, make sure it's on disc or rectangle, make sure keep sharp patterns is turned on, look up the lumen values from the light manufacturers spec sheets and put that in the lumen value of your lights. That should be the most accurate.
If you have free floating hanging lights, it's a bit more obtuse.
-
No, I don't think so. I could be wrong but based on some quick tests I've done, I think corona is interpreting the lumen/candela values from 3ds max photometric lights incorrectly because while 3ds max takes the shape of the light into account, corona does not.
If you want the most accurate light for a pot light as of this moment this is probably the most accurate; load up your ies files in a corona light, make sure it's on disc or rectangle, make sure keep sharp patterns is turned on, look up the lumen values from the light manufacturers spec sheets and put that in the lumen value of your lights. That should be the most accurate.
If you have free floating hanging lights, it's a bit more obtuse.
Is the shape and size of the light in this 'particular' case then truly irrelevant as it should ?
-
Yes, as long as your using lumens or candelas as your unit, the light shape between rectangle and disc are irrellevant, as long as it's not sphere or cylinder.
Based on the 1cd = 3.14 lumens that's present in corona's unit conversion. this equals to an 120 degree apex angle which is roughly the same as most ies files for pot lights.
If you want to see this in action load up an ies file, switch to lumens, click on keep sharp patterns and cycle through the different shapes.
(rectangle and disc stay the same, because they are acting like a point light)
next
switch it to the default lighting units and then cycle through the shapes they will all act differently.
-
Good morning all, excuse me for turning back on this but I am curious to know if the dev team is working on the IES mess so they can put some order on this. :)
Thanks!
Dionysios -
-
sorry, with everything else going on, this is on pretty low priority right now
-
sorry, with everything else going on, this is on pretty low priority right now
No worries, is important at least to know is in your To Do list... :)