Thanks for the feedback and follow-up notes from Fluence.
This fixture was brand new, out of the box. Aging effects were unlikely.
The device utilized to capture the spectra is the following:black-comet-sr
And is calibrated prior to each use using the following with calibration data: SL1 CAL
The cosine receptor is CR2 from the same vendor.
The instrumentation is scientific grade but the facility is not. Meaning, measurements are not from an integration sphere but rather open air. This means purposeful color mixing is not occurring where placement could have an effect on the captured spectrum and, from a PPF perspective, we certainly do not capture all of the energy. It is, however, closer to what we might experience in use.
I would expect some variation simply from instrumentation but would still expect that, for spectra in particular, error to be less than the apprx. 2-5% difference relative to the notes from Fluence. At some point, I’ll need to go and evaluate spectral sensitivity relative to sensor placement to see if there is anything significant. For fixtures with the sparse far red leds, for instance, I can move the sensor around and see obvious differences within the far red region show up. Centerspot PPF would certainly be sensitive to sensor placement and is fixture design specific. There could always be some undetected maths error in there somewhere as well. For instance, sampled spectra bins do not line up with the nicely rounded red/blue/green bins although the contributed power in each sampled bin is very small. Or, because Blue is higher energy, an error in the energy calculation could be causing a bias across the bands.
This does point out that any spectral measurements should occur at the higher hanging heights to allow for better color mixing to occur. But, at the same time, their response is in essence marketing specifications derived from their engineering sample.
Assuming everything is accurate, would the differences in the measured ratios matter in use? Well, I don’t really know
Thanks. Good info.