Ok, so that is further off than mine… but please remember these are not a scientific instrument. We have test equipment at work that verifies the frequency and deviation on DWDM tuned and tunable optics… the device costs thousands of dollars and have a very narrow spectral range and better be accurate and they are but this is only a general tool for a hobby use.
The problem is that most lights we have access to have never been tested in a scientific manner for CCT and spectrum, so we do not have a known reference source to compare to. Tint, CCT, CRI and DUV vary based on batches and bins of emitters and can even vary within a given bin of any emitter. Those of us who have built dozens to hundreds of lights over the years have a good idea about CCT, CRI and DUV when we pick up a light, this will confirm our ideas. If this site and others were not around that hosted pics… And lets say that someone new to lights decided to get a light and did not understand CCT and CRI and just ordered random lights. They will eventually pick a favorite without knowing what any of the numbers are, how would you know what to reorder? This is where you have to go back to the specs of the light and find out what emitter, what CCT and what CRI the emitter is. If you know the emitter and have a good idea of the CCT you can and should go the the manufacturers datasheet and verify and study what you believe. For the newer light person, this tool can give you guidance for the veteran light person, it does things like confirm why 1 light out of 10 that looks amazing edges out the other 9. One could also confirm that proper LED’s were shipped in a bulk order, to some degree.
I find that it can be very close, but it takes work. Much like reading very low amperages, you must use a tool you trust and use it consistently. For me, this is what works…
1) For quick reads, dark room as much of the hot spot as possible on the sensor from about 8 inches as long as the LUX read is around 20 to 25 thousand.
2) For a more accurate read I have built a shoebox lined with white foam. The sensor sits in one end and the light is introduced thru a hole on the other end. Still read from 20 to 25 thousand lux.
One can still read at much lower lux, but that is also a much lower drive current and remember this is known to affect things like tint of the emitter. But, this is useful as you once again can find out why that one light with an underdriven SST-20 looks so much better than the others that have the same emitter, hint… look at the color coordinates.
I really like the tool, within it’s limits. I feel lucky to have such a cheap device available to me to help draw inferences about something as subjective as light. Would I like to have something more accurate? Sure, but it has never been something that I really felt a strong desire to finance, own and learn to operate.
I am typing off the cuff hear, likely I have said something wrong… if so correct away.