Working Group

#649 Correction to Units Database - Light Base Values

Keith Bishoρ Wed 31 Oct 2018


All of this began because we have a user of one of our systems that wanted us to be able to convert a value between footcandle and lux and didn't understand why this type of operation wasn't supported. This lead to some research into lighting measurements and the recommendations at the bottom that should help correct some of the entries in our unit database that deal with lighting.


The base SI unit for the addition of lighting is the Candela (cd). The cd is the magnitude of luminous intensity (we are calling it luminescence which is contrary to the International Bureau of weights and measures but I don't think the exact name is important, but we should probably change it. As a note, Fantom has it listed correctly). The cd is roughly the amount of "light" traveling in a certain direction.

The next concept is luminous flux which is the amount of "light" emitted in the available direction. (This available direction is measured in steradians (sr) where a sphere would have a steradian value of 4π sr or roughly 12.57 sr). Luminous flux is measured by the lumen (lm) which resolves to base units of cd·sr. (The steradian is not technically a base unit but can not be reduced to a base unit because it's math resolves to m²/m² which removes all units)

Now that we have luminous flux defined, we can move on to illuminance (or luminous emittance). To build on the above concepts, this is the amount of lumens passing through and area. The SI unit for luminous flux is the lux (lx) with base units of cd·sr/m². The SAE unit for luminous flux is the footcandle (which we have listed as a measure of luminescence). The footcandle (fc or ftcd as we have listed) is equal to 10.76 lx

The measure of luminous intensity passing through an area is called luminance. The SI unit for luminance is the candela per square meter (cd/m²) which we have listed in our unit database under illumanence.


Based on the above discussion, it is recommended that we change the cd based units in the database to:

-- illuminance (m-2*cd1*sr)
lux, lx; m-2*cd1*sr
footcandle, ftcd; m-2*cd1*sr; 0.092937

-- luminance (m-2*cd1)
candelas_per_square_meter, cd/m²; m-2*cd1

-- luminous flux (cd1*sr)
lumen, lm; cd1*sr

-- luminous intensity (cd1)
candela, cd; cd1

Please let me know what you think.

(I understand that our unit database was effectively adopted from oBIX and have passed this discussion along to the chairman listed on their website. However, I do not believe that our correcting these entries relies on oBIX changing theirs)

Jay Herron Thu 1 Nov 2018

Great write-up Keith! I think this is an excellent addition.

I think it's worth mentioning that there are already some of these units in the current database, seen below:

-- illuminance (m-2*cd1)
lux, lx; m-2*cd1
candelas_per_square_meter, cd/m²; m-2*cd1

-- luminescence (cd1)
candela, cd; cd1
footcandle, ftcd; cd1

-- luminous flux (cd1)
lumen, lm; cd1

So this proposal is to change luminescence to luminous intensity, and change these units to use the sr unit according to the formal unit definitions, thereby correcting the definition of footcandle and moving candelas_per_square_meter from illuminance to the luminance quantity.

My only suggestion would be, since we're on the subject of changing these, to also include the following for completeness:

-- luminance (m-2*cd1)
nit, nt; m-2*cd1; 1.0
stilb, sb; m-2*cd1; 0.00010

-- illuminance (m-2*cd1*sr)
phot, ph; m-2*cd1*sr; 10000.0

Jeremy Yon Fri 2 Nov 2018

Hi all, I'd like to engage with this conversation as a representative from the lighting community ANSI C137. We're in the process of formalizing a group to help contribute to this excellent work specifically for lighting elements.

There can be a lot of confusion around how these terms and equations are technically used (versus in common vernacular), but before we jump in in to clarifying/simplifying the technical details - would someone be willing to add to this thread the context for how these data types are being utilized? Is this primarily limited to daylight sensing or are there other use case needs?

Thanks, and we're looking forward to contributing.

Matthew Giannini Thu 8 Nov 2018

Keith - thanks for an excellent write up, and Jay thanks for your clarifications. I would like to digest it a little, but sounds like these changes make sense.

@Jeremy - from your perspective, do you have any issues with the proposal that Keith put forward? The Units documentation may help shed light ;-) on how these units are used. Mostly I am concerned right now with ensuring the units are correctly defined and classified.

Jeremy Yon Tue 13 Nov 2018

Got it, here is some info:

To answer the simple question - the conversion between footcandle and lux is just sq-ft to sq-m in the denominator, so 1 fc = 10.764 lux.

-Illuminance is the light incident at a point on a surface, in footcandle (fc) or lux. The equation for Illuminance (we signify this by an E) is E=I(cos-theta)/d^2, where I is the Luminous Intensity (amount of light in a specific direction from a point source of light), theta is the angle of incidence, and d is the distance away in feet. If you imagine a flashlight pointed at a wall directly (cos theta = 1), then this is talking about the amount of light at the center point increasing as you get closer and lower further away, NOT the size of the circle.

So unfortunately, no - the proposed definition for illuminance is not correct. What looks like an area factor in the units is actually a squared-distance, Illuminance is at a point not over an area.

The other terms are easier to explain from a different direction. -> Luminous Flux is basically all the light emitted by a light source in all directions (assumed to be a point in most cases). Think of a light emitting point at the center of a sphere - if you add up all the energy hitting the inside surface of the sphere, you'd get total lumens (We typically measure this exactly like that - inside a big calibrated white sphere - 2 or 3 meters in diameter) -> Luminous Intensity is that same light emitting point, but only looking at a solid angle - so a cone within the sphere (lumens/steradian). (We typically measure this in a huge room with an automated moving mirror and a fixed detector)

So we rate any EMITTER of light as Luminous Flux (lumens) to represent the total output and then we test the products to get the directional Luminous Intensity of light (in even increments, stored in a computer file or output table) so that we can simulate the expected Illuminanace (fc or lux) in a room. NOTE the above Illuminance equation is for every point at every angle of light bouncing off of every surface - so the real equations in practice get huge and so we use simulations.

Finally - Luminance, even though the units are similar, is from a COMPLETELY different point of view - that of an observer looking at a surface. In this instance, that IS an area function (ft2 or m2), but instead of assuming the light source is a point, it is measuring from a point and looking at the light source as an area. It's annoying, but the two are not equatable in real life - I could go on to more detail, but this is already getting to be a long post.

Please let me know if further explanation would be helpful or if some diagrams/pictures would help in your need.

Matthew Giannini Wed 14 Nov 2018

Keith has offered to champion a working group on the topic of lighting. I will convert this forum topic to a working group and make the Keith the champion. I encourage everyone on this thread to join the group.

@Keith - it might be good to make a post summarizing the goals/outputs for this particular working group. It sounds right now like the issue is primarily around fixing the unit database for lighting units.

Keith Bishoρ Wed 14 Nov 2018

Welcome to the "lighting" working group. (Please make sure you are listed under the members list towards the upper right of this page. If you are having an issue with this, please let me know.)

I have 2 existing goals for this this group:

  • Finalize a proposal to correct the "lighting" units as they currently exist in the unit database file.
  • Review the existing tag structure that exists for lighting and potentially propose an update as I know light architecture/technologies have advanced since this was published. This will be a more substantial work product as it will have to fit with other standards that haystack is currently working with to "merge".
  • and we can discuss any other topic regarding "lighting" that would be helpful for the community to adopt

As for schedule, I plan on giving this working group a week for more members to join and then we will get this started.

Looking forward to working with everyone on this.

Keith Bishoρ Tue 9 Apr

The lighting WG is proposing changing the unit database entries for those that deal with photometry from:

-- illuminance (m-2*cd1)
lux, lx; m-2*cd1
candelas_per_square_meter, cd/m²; m-2*cd1

-- luminescence (cd1)
candela, cd; cd1
footcandle, ftcd; cd1

-- luminous flux (cd1)
lumen, lm; cd1

To new entries:

-- illuminance (m-2*cd*sr)
lux, lx; m-2*cd1
footcandle, fc; m-2*cd1; 0.092937
phot, ph; m-2*cd1; 10000.0

-- luminance (m-2*cd)
candelas_per_square_meter, cd/m²; m-2*cd1
candels_per_square_foot, cd/ft²; m-2*cd1; 0.092937

-- luminous flux (cd*sr)
lumen, lm; cd1

-- luminous intensity (cd)
candela, cd; cd1

Please find our supporting documentation at:

(I have to apologize for not noticing that we've had further interest in this WG and I failed to get you added to our discussions. We'll add you to the next discussions)

Let us know if you have any issues with these changes.

Login or Signup to reply.