My understanding for LEDs is that they are more accurately measured using (in terms of Apogee sensors) the Sunlight calibrated sensor, but most of the hobbyists that own a PAR meter will have an "Electric" calibrated sensor, which is designed for halides and T5 type lights.
The only reason I can suspect this is exactly what Hoppy mentioned--- the more natural spectral curve of an LED is much closer to the curve of sunlight than a halide or T5 puts out.
Anyone who can explain this better, or correct me, please do.
So do you or anyone else happen to have a 3000-20,000$ radio spectrometer?
Well, I don't own it. It belongs to UNC Wilmington. We're studying a pile of common LED brands here soon--- on April 30th.
We will be looking at the actual graph of the individual LEDs, and then the resultant spectral graph of various planted tank and reef combinations.
Here's a list of what we will study (along with a myriad of common/likely combinations of these)
Cree XPG in cool, neutral, and warm white.
Cree XML in cool white
Cree XRE in royal blue and cool white
Cree XPE in blue, royal blue, and red.
Bridgelux 402 cool white 5600K.
Satistronics 10w pure white and 455nm royal blue
Satistronics 20w 15,000K cold white
Satistronics 3w pure white and red
Rebel neutral white, royal blue, cyan.
This basically covers the gamut of what we're likely to use--- LEDs from the "big three" name brands and then the chinese import Satistronics.
We will also take PAR data from raw LEDs with no secondary optics.