HPS with out a doubt.... Chlorophyl absorption rate is the highest during the red wave length... There are 2 types of chlorophyl, a and b, a is the primary chlorophyl and absorbs the red spectrum (measured in nm) This is just off the top of my head, I would have to get out my bot book from a couple semesters ago.
Yep, A and B, each of them have 2 peaks. A peaks at 455 and 625nm, B peaks at 475 and 665nm. Testing has shown that the higher wavelengths (red spectrum) are slightly more efficient than the blue, but that isn't to say you should give just the higher wavelengths. The most efficient production of sugars will be from combining all photosynthetic peaks, possibly going slightly higher on the red, but not neglecting the blues. From looking at seasonal spectral changes in the sun, in autumn the sunlight will shift more toward higher wavelengths, so flowering is typically accompanied by higher 625-665nm levels than earlier in the season when going through veg. This doesn't mean that it cuts out the 455-475nm peaks, simply that the intensity of light is stronger in the higher wavelengths.
I got a request to elaborate on PAR a little. There is a lot of very scientific info on PAR and photosynthesis online, so my best advice is to do some research. The subject of light spectrum is a broad and complicated field. I am by no means an expert, I simply have spent a lot of time over the last few years doing research on the subject.
To give an overview. When us humans think about light we think about how bright something is to us. How much it illuminates, and how far that illumination penetrates. This is the idea behind luminosity, which is the basis for many of the ways that humans quantify the intensity of a light source (luminous energy, luminous flux, luminous intensity, luminance, illuminance, luminous emittance, luminous efficacy, etc...). Basically lumens are a measure of light in the spectrum peaking at roughly 550nm, and sharply dropping off. This is what we see and makes a lot of sense when you are talking about how bright a light source appears to us.
Plants, on the other hand, see a much wider range of light colors than we do (excluding greens). When we talk about a plant 'seeing' what we really mean is that it is able to perform photosynthesis. As I mentioned before, photosynthesis (the production of sugars) peaks at two points for each type of chlorophyll, of which there are two. PAR (photosynthetic active radiation) is a measurement of the intensity of a light source with peaks around 450nm and 665nm. The levels will taper down from a high blue to a low green, then back to a high red; following the spectrum of light that plants are able to use for photosynthesis. Instead of measuring how bright we see a light source, it is measuring how bright the sources of photosynthetic light are.
The end result of all this is that most lights are advertised based on their luminosity, which ends up being what hobbyist gardeners use to evaluate their light source. Professional gardeners understand this difference, and will ignore measures of luminosity and instead focus on PAR, since that is actually rating what they care about.
Hope that helps!