Why lumens don't mean anything to plants. was going to post an explaination but this guy does a better job and I don't have to type \tQuote: Light Measurement for Horticultural Applications permalink Why is the lumen useless for horticultural applications? The answer to that lies in the definition of the lumen. What does it measure? The lumen is a SI unit that measures light according to how effective it is at stimulating human visual response. Let's look at some definitions: The candela (cd) is the luminous intensity, in a given direction, of a source that emits monochromatic radiation of frequency 540x10^12 hertz and that has a radiant intensity in that direction of 1/683 watt per steradian. A steradian is a unit solid angle (the 3d equivalent of a radian). There are 4pi steradians in a sphere. 540*10^12 hz is roughly 555nm, the peak of human visual sensitivity. This spectrum limitation is what makes photometric units useless for horticultural applications. 1 cd = 1/683w/sr 1 lm = 1 cd·sr = 1 cd·m2·m–2 1 lm = 1 cd*sr = 1/683w/sr * sr = 1/683w So lumens really are a measure of radiant flux (watts), but with the wavelength restriction. In the real world, very few light sources are monochromatic at 555nm, so to assign lumen ratings to lamps, it is necessary to weight the light emitted at other wavelengths. This is calculated by the Luminosity function. Now if you look at the curve of the luminosity function (CIE 1931) and compare it to the curve representing the absorption spectrum of chlorophyll a and b, you will see that the former peaks where the latter valleys. That is to say: the lumen is weighted to measure the light that is least useful for plant growth. Low Pressure Sodium lamps are virtually useless for growing plants. LPS emits nearly monochromatic light at 590nm, and has a very high luminous efficiency (in fact, it highest luminous efficiency of any gas discharge source currently available). If lumens (or luminous efficiency) were really any indication of usefulness for plant growth, we'd all be using LPS. The other thing to notice about the comparison is that the photopic efficiency of light approaches zero right about where it's effectiveness for photosynthesis peaks at both the blue and red end of the luminous efficiency curve. So rather than being a good indicator of how useful light is for plant growth, the lumen is just the opposite. Stated another way: for any two lamps having similar radiant efficiency but different luminous efficiency, the one having lower luminous efficiency is almost guaranteed to be more efficient for plant growth. What we are looking for is high radiant output, with a low lumen rating. In addition to the chlorophylls, there are a number of other photoreceptive pigments (e.g., carotenoids, xanthophylls; collectively called "accessory pigments") in plants that contribute to photosynthesis. These do absorb light between the chlorophyll absorption peaks. The amount and ratio of accessory pigments varies from species to species, and even among individuals of the same species. Because of these accessory pigments, the action spectrum of photosynthesis differs from the chlorophyll absorption spectrum. Lighting recommendations are often given in the nominal wattage consumed by the luminaire (rated watts) per square foot. Watts consumed by a luminaire don't give any indication about radiant or luminous efficiency, but at least don't have photopic bias. The argument that rated watts somehow favor incandescent lamps is a non-sequitur: incandescent lamp output is measured in lumens, too. It is certainly possible to provide the magic 7000 lm/sf using incandescent lamps, but this is not likely to have very good results, because so much of the incandescent output is heat. For lamps producing similar spectral power distribution, it is perfectly valid to recommend light levels in terms of the rated wattage of the lamp; for example, "50w-60w of Hortilux per square foot" is an entirely reasonable lighting recommendation that conveys useful information. It does assume that the light is going to be distributed fairly evenly throughout the growing area to produce somewhat uniform irradiance. Irradiance (energy received by the plants, measured by a radiometer in watts per unit area, e.g., w/f^2) is more useful information, but also says nothing about the spectrum. Lighting recommendations are also often given in lumens/unit area (e.g., 7000 lm/f^2). This has the same limitation as input watts/unit area: it divides the initial luminous flux (photopic lumens emitted by the lamp) over the area being lit, which assumes that the light is distributed fairly evenly throughout the lit space (for even illuminance). Actual illuminance measurements (in lux=lm/m^2 or foot-candles = lm/ft^2) would be more useful, but still suffer from spectral bias and are not informative about photosynthetically useful radiation. A caveat is that the lumen (or lux) can be useful to compare light levels when the source SPD is the same (or very similar). For example, outdoor or greenhouse light levels can be meaningfully discussed in terms of Lux because the SPD of sunlight is fairly constant (at least for unfiltered greenhouse glazing). Similarly, light levels can be compared in lumens, lux, or foot-candles between two grows using the same lamp model. Photometric units are not useful for comparing sources that emit different spectra. So while it is meaningful to recommend "7000lm/f^2 of Hortilux," it is not useful to recommend 7000 lm/f^2). Photosynthetically active radiation (PAR), measuring radiance in watts, is much better than photopic units (lumens or lux) or radiometric units (watts or watts/m^2). PAR is a mesaure of lamp output in the spectrum that is actually useful to the plants. However, this data is only available for a small number of lamp models. An additional limitation is that the PAR watt treats all wavelengths between 400 and 700nm equally, but all wavelengths within that range are not equially effective for plant growth. Despite these limitations, PAR efficacy (PAR watts emitted/watts consumed) is a decent indication of efficiency for horticultural lamps. PAR irradiance can be measured using a spectroradiometer (not cheap!). Irradiance at the plants can be measured directly, as photosynthetic photon flux density (PPFD). This is measured in Einsteins (E), which is not a SI unit. One Einstein is one mole of photons, regardless of wavelength. Irradiance levels on the surface of the earth occur on the scale of micro-Einsteins. PPFD is observed directly using quantum meters (starting around US$100) that measure microEinsteins per square meter per second, for photons having wavelengths between 400 and 700 nm. This has the same limitation as the PAR watt, by treating all wavelengths in the range as equally useful for plant growth. We can get a rough idea of the how efficient a lamp will be be for plant growth by looking at it's Spectral Power Distribution (SPD) and comparing it to the action spectrum of photosynthesis. The more useful lamps produce most of their light around the chlorophyll absorption peaks, and produce less light away from those peaks. However, without knowing either the radiant output (in watts) or PAR output (also in watts), we can't calculate anything objective, like it's overall PAR efficiency. knna has done some work to analyze the action spectrum of photosynthesis for Cannabis, and provided a helpful tool that analyzes an SPD and calculates photosynthetically useful radiation (PUR) produced by that lamp. Getting back to the original question, it is well documented that (within limits) higher ratios of blue/red reduce internode length, while lower b/r promotes flower production but also tends to increase internode length. Red light is generally more efficient than blue light, both for the lamp to produce, and for the plant to absorb. Tazawa concluded that"high-CRI" MH is the best all-around light source for plant growth. I have read several research reports that concluded that HPS consistently outproduces MH as the sole light source for several crops in controlled studies. I did read one study that compared MH vs HPS vs MH+HPS as supplemental light in a greenhouse in Quebec during winter, and concluded that there was little difference in yield under any of the tested configurations. The main limitation in all of those studies is that they did not compare specific lamp models, so the conclusions can't be assumed to be generally applicable, in particular to "spectrally enhanced" "horticultural" lamps. I still use MH in veg to control internode length. For lanky cultivars, I use MH during the first 2-3 weeks of flowering to help control stretch. penguin \t\t\t \t\t \t \t [/COLOR]