How do you define 'bluest'? It is based on perception? Is it based on all wavelengths from 400nm to 460nm (includes actinic), or just blue (450-460nm)? Do you want the bluest bulb regardless of what other spectrums are present... as in, it might have the most blue but not look like it because it also has lots of daylight?
This is the point where perception meets reality. Which wattage you are interested in will also change the responses.
For example: did you know most 20,000K's have less actinic in them than most 10,000Kish bulbs? A 20,000K isnt so much a daylight bulb with more blue as it is a daylight bulb with all the spectrums other than blue taken out. There are many 'daylight' looking bulbs that often have more actnic, and more blue, than a 20,000Kish bulb. Rather than having that 'spike' on the spectral graph that is very narrow in the 440-460nm range, they have less dramatic, yet broader coverage in spectrums from 400-460nm, meaning when you 'measure' the total amount of 'blue' light, they really have more than the one's that 'look' bluer to the eye. 'Blue' looking bulbs aren't so much about having MORE blue than a daylight bulb as they are about just having less red, orange, yellow, green... etc.
Some of your BLUEST bulbs, depending on how you define it:
pheonix 14,000K 250wattDE: .63watts/m2/nm of about 452nm wavelength light. It is really one of the bluest bulbs out there because it has a huge blue spike, yet little of anything else. It really lacks actinic though too, so corals tend to look washed out and chalky under this monochromatic blue light. Its a prime example of a light that makes the tank look cool and blue, yet makes the corals look pale.
Aquaconnect 14,000K 400wattHQI: about a 452nm narrow band spike that reaches about 1.52 watts/m2/nm on Sanjay's charts. The actinic spike at 410nm is only about 0.47 watts/m2/nm. Like many 'blue' bulbs in the 400 watt range though, these bulbs seem more 'daylight' because even though their proportion of other daylight spectrums is so small, our eyes register green and red much better than blue so this bulb actually looks like it has more daylight in proportion to blue than the 250 watt version. Really its just our eyes. A similar comparison would be the pheonix 250... on an e-ballast its pretty much monochromatic blue, but raising up the current to a HQI ballast makes the bulb look more 'white'. Its because while the boost in power raises the spectral curve evenly about .01 watts, this is only a boost in blue of about 3% but results in about a 15% boost in many daylight colors... and since our eyes once again dont favor blue very well, it looks like the relative amount of daylight has increased. Aside from the low actinic output and a good amount of daylight, this bulb is THE BLUEST halide on the market now since it has that one huge spike in the 452nm range. In person, its not the bluest, but it really is. Our eyes also see different colors at different proportions based on intensity (so the lumen/photometric scale is somewhat flawed in this respect). For instance, we see blue better in dim light (since a good deal of our blue color perception is with our rods, normally reserved for dim/night vision) and as it gets more intense, we see less of it. For this reason, less intense lights of the exact same spectrum will look bluer.
G-man 14,500K, BLV/Ushio 14,000K (all same bulb): 0.6watts/m2/nm of 452nm blue... so for a daylight bulb, it has almost the same blue spike as a pheonix at 250 watts. The thing is, this spike isnt alone. This bulb also has a .2-.25 watt/m2/nm spike at 405nm, a .25-.32 w/m2/nm spike at 415nm, a .4 watt spike at 425nm, and a .31 w/m2/nm spike at about 425nm, meaning that even though this bulb looks like a daylight bulb, and it doesnt have the 'blue spike' like at 450-460nm like many 20,000Kish bulbs its volume of blue and actinic trumps that of any other halide on the market. Many 20,000K's have a blue spike and then little else, little actinic even. You just dont see all that blue when the bulb is over a tank because it also has gobs of red, orange, and a green to round it out. Overall, an awesome bulb for output and spectrum though. Makes coral pigments very rich.
Ushio 20,000K 400watt HQI: has a 1.18 w/m2/nm spike at about 458nm but only a .2w/m2/nm peak at about 415nm.
XM 20,000K 400watt M58: has a 455nm blue spike of almost 1.5 w/m2/nm, but only about a 412nm actinic spike of .28 watts/m2/nm. What makes this bulb 'look' bluer to the human eye though is that unlike the Aquaconnect 14,000K it lacks most daylight all together... less than a third of the daylight spectrums compared to the AC 14kK. Its pretty much just blue though, and still keep in mind that if you compare all actinic & blue spectrums (usually done by taking the volume under the line on the spectral graph, or integration of the line in these spectrums), the Ushio/BLV 14,000K has more blue/actinic in its output. That one huge blue spike is all the XM has, where the Ushio has a few nice actinic spikes, extra blue spikes, etc.
