@gorion103
Given the table above, how did you end up with that number? Incadescent and LEDs are clearly different and to me its pretty obvious they list the power wattage.
As a general starting point, I use sun strength 400 and sky strength 40, as they are easy to remember. That’s based on this info if you want to setup your own. They are close enough to the real thing and I can (with photographer addon) use the sunny 16 rule and have good exposure for front sunlit subjects - just like on my camera.
The wattage system for interior lights are so confusing that I have given up understanding them - lumen is easy to find usually but nowhere to enter, whereas wattage is always power usage which require efficacy/efficiency (and I don’t even know which to use) where is listed nowhere. Additionally, a 40W incadescent light appears to produce different amount of light on a 220V/50hz vs 110V/60hz power system. So on a household bulb you have lumens which cannot be entered anywhere, only power wattage (see table above even) which is useless without efficacy/efficiency which cannot be obtained.
However, using photographer addon setting EV to descriptive office lighting (which is what I work with normally), exposure ends up at 0 - so for a night shot at 0 exposure, you could light it as you would expect an office to be lit and have a well exposed scene. Since I usually do “render a nice image” rather than a “lighting study”, that is good enough for me though. But it really should have to be that cryptic.
Thanks for the article but I already read it before.
I already tried the formula “1 Lumen = 1/683 Watt = 0,001464 Watt”.
I made some tests and i think it doesn’t work.
An example for a lamp at 350 Lumen : 350 lm * 0,001464 = 0,51 Watt ?
I entered that value (0,51) into a blender spot lamp I think the result was really too dark in comparison with a real lamp at 350 lumen.
Or maybe I’m doing something wrong ?
As CarlG mentioned there is also the problem of efficacy/efficiency. But I guess we need to know if we are speaking about electric wattage or light emission wattage (light power).
I guess if we speak about electric wattage we need to know about the efficacy/efficiency which is used into blender.
If we speak about light emission wattage we need to find a formula to convert wattage into lumen.
Although watt and lumen are used differently in the industry, at their core both of those units can be seen to describe energy-per-second. Therefore it is possible to convert between them.
Look at these for reference: Link 1 Link 2
Ok, I get it, I think. The conversion is between lumen and the wattage Cycles expects, and has nothing to do with power wattage?
Instead of Lum*0.0015, another one easy to remember would be Lum/666
Turning on a single phillips hue at 800 lum, I get similar results in cycles when matching camera settings (ISO800, 1/10", f/4, no extra glass) using photography addon (Candle) when doing that conversion factor. In exposure terms that would be equal to 3.68.
I have to try this out more, but it appears to be good.
BTW IMO the only way to have realistic lighting is with High EV HDR maps only.
When you try to mix HDR with blender lights there is no way to make them scientifically consistent each other.
It becomes more a eye pleasing only evaluation.
Also I would not use photographer addon because as far as I remember it doesnt bake settings into the EXR file, I would set desired exposure by manually set lights values as explained in the post I’ve linked before.
That’s all about your workflow tho.
What do you mean “bake settings”? It’s only a tool for setting exposure in a photographer friendly way; ISO/Shutter/Aperture. To me the exposure setting is cryptic. Exposure 0, what does that even mean? You can also disable it after setting everything up, exposure will still be set to whatever value was determined.
Scientifically? No. Visually, yes. Preset the camera exposure to known values which will produce expected results, then replace sun&sky with hdri and adjust its strength until it produces similar results. Now you can use those interior lighting assets without tweaking all of them.
For lighting interior scenes, I much prefer sun & sky (only sky is preferred due to noise) over hdri.
You can’t derive a luxmap in cycles. Try Dialux or some high end architect software for that. You can only do reflected light using filmic false color. The problem using a white diffuse as a material override is that all emission surfaces will be overridden as well, and obviously all energy from specular bounces are lost.
Use filmic false color to help you set exposure, as that is what photography deals with. But luxmaps are out of reach, so you can’t do architectural lighting studies.
With filmic false color, RGB 0.18 will show up as grey. Slightly above or below will show up as slightly green. Test with using emission shader at that color with 0 exposure.
Blender just needs to set / agreed to a default for the representation of light intensity in the value’s of Lux and Lumens.
Like Dialux or Relux has done as well. It’s not a real representation of light intensity; but a suitable one to do the job right.
A good formula is needed to take different factors in to matter. A conversion from Watt’s/Square meter is easy. But to find a good relationship between the factors Lumens > Lux > Candela; all need to precisely interact within the Filmic color management to get a close to an accurate representation of the real world values.
Right now the representation is not accurate at all and confusing. Like; when using an IES file. It contains all of these data (Lumens > Lux > Candela), but still; after importing the IES file; none of the point light settings are adjusted correctly.
In fact, when you try to get as close to the representation of brightness of a program like Dialux; you’ll need to manually set the power to 1 Watt, Size (also defined by IES file); would set it to just above 0 or maybe try the size of the actual armature of the IES file, etc.
Is 1 watt the case for Blender? It just doesn’t seem like it to me. Also, in Blender size 0 appears to give the correct and expected “throw pattern” from the armature, but not the correct shadow softness.
I just visited the site today of a project I’ve done, and don’t you just hate when you’ve not been given all the required info - such as “the room is facing an interior courtyard with tall buildings, so no sun will ever shine there and also daylight will be limited” - I was very worried about this room.
You cannot use Lumens for intensity. It doesn’t work. It’s a measurement of perceptual energy via luminous flux, while Joules are a measurement of physical energy, or radiant flux.
Setting the values of an IES lamp (IES file by the lamp manufacturer) to 1 Watt and the size/radius to the size of the corresponding armature (in Blender) made a close to almost the same result as when rendering a scene in Dialux; a program for light calculations. Leaving it at default settings looks just ridiculous (an inaccurate representation).
Not saying Dialux has the right way to simulate/represent the light tough. But they have a default. Without a default for the representation of light (simulated), you can’t really assign a decent value in numbers. Any number could be assigned to any light intensity.
Sorry. Maybe I’m not using exactly the right words to describe what I mean.
But to keep it simple (light calculation wise, with all the scientific factors and formula’s we can use) I’m sticking to using Lumens, candela and lux to represent light value’s; like the light calculation programs do. Right now in Blender, I believe it’s Watt/Square meter.
The reason is that luminous flux is a perceptual phenomena. That is, if you output precisely the same Joules of energy of a reddish, greenish, and blueish light, they will appear of different luminance to our perceptual systems; the greenish will seem the brightest, the reddish next, with the blueish light being darkest.
As a result, you can’t rely on lumens to communicate light intensity when discussing coloured light. It simply does not work to communicate unambiguous energy levels.
What matters in a light transport model is energy, and the amount of energy output from a given source. That energy is radiant flux.
Hence the difference. If one does not understand this crucial and fundamental difference between radiant flux and luminous flux, no amount of discussion of numbers will matter. There is no “simple” here, just dead wrong.
You are making it quite clear to me why I should not talk in terms of lumens for light intensity.
Now I do use the Program Dialux a lot, here armatures are always defined in terms of Lumens, lux and candela. Shouldn’t they be using radiant flux then?
Do know that I am only talking about the representation of light intensity simulated in blender. Isn’t this just perceptual in a sort of way?
Imagine we had an idealized HDR display that doesn’t exist yet. When we emit peak intensity across all three colours of the base lights in a pixel of that display, it will emit 10000 nits of brightness. As you know, a cd/m^2 aka nit is a luminous flux based measurement, meaning it is perceptual and based around our sensitivity to certain specific wavelengths of light.
So what would happen if we turned on only the green light in a pixel? How many nits of brightness do we have now? What about red? Blue? What if we mixed two of the lights?
The point is, if this simple example were our modeling environment, we would be emitting radically different quantities of radiant energy to “match” the same luminance level. That could be quite tricky in a modeling environment where we have and need granular control over the chromaticity of the light reflected or emitted.
What we’d end up with is a strange UI control where while our lumens would be identical, the actual radiant energy of the light would be all over the map depending on the chromaticity chosen.
On the other hand, if we use the radiant energy measurement, we are always assured that the energy is constant, and only the perceptual result shifts, which is what we’d expect; blues should appear darker than reds which should appear darker still than yellows and greens at equal energy emission.
