One is a traditional incandescent 60-watt bulb that puts out 840 lumens.
The other is a compact fluorescent 23-watt bulb that supposedly puts out 1600 lumens. It claims to be the equivalent of a 100-watt bulb.
One would assume that the second bulb would be considerably brighter than the first. So why is it considerably dimmer? And what is a “lumen,” if it’s not a measurement of light output?
And it’s likely that you are mistaken when you say the CFL appears dimmer than the incandescent. For a couple of reasons:
the color of the light is probably different – many CFL’s (especially cheap ones) are ‘bluer’ than the orange-yellow of incandescents. And the sensitivity of eyes varies depending on the color; many people see blue light as less bright than yellow light.
Incandescents are almost a point source; all the light comes from a filament that is about 1/2 inch long. In CFL’s, the light comes from all along the tube. For the typical twisted spiral CFL, that’s a tube about a foot long. So the source of the light is much more spread out, and may not seem as bright as trying to look at the glaring filament of an incandescent bulb.
Follow-up question, then: Considering this difference only (and ignoring the color difference, or suppose we get a CFL color-equivalent to a real light bulb): Does this light spread out along the length of the tube make a difference in the perceived brightness on a piece of paper 3 feet away?
That is: If I am using this light to illuminate a page of a book I am reading, will the page be as well illuminated by the foot-long twisty CFL as it would be by the incandescent filament?
What I’m saying is this: I have a lamp that takes a 60-watt bulb. The bulb burned out, and I replaced it with a compact fluorescent. The room now looks darker. I understand the difference in color, and it’s not that. According to the lumens listed on the packaging, the new bulb should be almost twice as bright as the old one. It’s definitely not.
I’ve had a similar experience with them. I am in a room right now that has a very high ceiling, almost 30’ at the center with dark wooden support beams coming down at an angle from the center to the walls. Picture a pyramid shape made of dark wooden beams. I have a little 13W spiral-shaped CFL in a table lamp that is supposed to be equal to a 60W incandescent. If I turn off the overhead light and rely just on that lamp I can only see about 1/2 of the beams before they begin to blur into shadows, and the very top of the ceiling is almost completely black. In the same lamp, with the same shade, in the same position, when I use an actual 60W incandescent reading light I can see the detail in the wood grain in the beams at the very top of the ceiling. It is like night and day, literally. So if nothing else they don’t throw the same lumen as far as incandescent lights do, at least in my own limited experience.
A lumen is a measure of visible light. In paticular, it is a measure of Luminous Flux – the measure of visible light output in all direction from a source. This is distinct from the Radiant Flux, which is a measure of all wavelength output from a source in all directions. An ordinary incandescent bulb puts most of its light into the infrared, where it’s not visible, and may have a very high output in watts, but it doesn’t do your vision much good. Sources with higher output in lumens ought to be giving you visibly brighter situations. In your case, the 1600 lumen source ought to be twice as bright at an 80 lumen source.
The value is weighted by the response of the eye, so a fluorescent source – unless it’s a very modern one that mimics an incandescent one in wavelength output – will not appear to be the same as an incandescent source. Traditional fluorescents had much more blue output and its light was concentrated mainly in a few strong lines, so its light appeared very different from that of a tungsaten filament lamp. Your statement (if I read it right) that the 1600 lumen bulb was not as bright as an 800 liumen bulb suggests that something is wrong.
The measurement and science of radiometry and photometry is a welter of subtly different definitions and confusingly named units – nits, lux, phots, lumens, candels, candelas, and so on. These might help. Or they might just confuse you further:
As CalMeacham says, the Lumen measure is supposed to take into account all the aspects of the eye’s response to create a normalised figure of merit that can actually be used for exactly the purpose of comparing the apparent brightness in a manner useful to humans.
About the only differences that are not taken into account are the effects of line spectra when illuminating coloured surfaces. The luminance is measured by illuminating a flat white surface (and they specify surfaces that have utterly flat spectral reflectance when they measure.) However CFL blubs are line emitters - typically tri-phosphor for the modern nicer ones. And you get a number of peaks in the output, not a continuum as you get with an incandescent bulb. The line emission is balanced so you perceive the light as having the same colour as a black body radiator of some given temperature (hence the marking of a colour temperature.) Critically the CFL does not have the same emission spectrum as a black body.
But when you illuminate a room, most of use don’t have flat white everything. The usual complaint is that the colours don’t look right. Which they won’t. Some objects shift colour worse than others, but it is inevitable that some colours shift badly. This can also affect your perception of contrast in odd ways. Overall you might find that things don’t seem to look as nice, or that you actually have more difficulty seeing things as well defined as you might hope. Maybe. A lot of this is why CFLs have attracted something of a bad reputation. Outright lies when it came to specifying equivalent light output (ie 18 watts CFL = 60 watts normal) didn’t help. The Lumen rating should stamp this out. So, in the end, it does sound that there is soe other problem, just rotten colour rendition with a CFL isn’t going to make an apparent 2:1 difference. But it will make some.