On 11 Nov, 15:07, pete wrote:
On Tue, 11 Nov 2008 05:53:59 -0800 (PST), wrote:

The packaging is another example of marketing 'though. It claims the
11W is equivalent to 60W, and even gives a comparison of a 60W
incandescent providing 620 Lm, and itself offering 640 Lm. The only
problem with that is that a standard 240 V GLS 60W (Frosted) *gives
700 lumen, and it is the so-called 'softlight elegance' that offers
only 620 lumen.

After the initial discussion here about the brightness of these bulbs
I did some real-life tests. The setup was a sheet of A0 white paper
(from a flip-chart) laid flat on a table underneath a ceiling hung
light fitting with a lampshade. I pointed my DSLR at the paper and
noted the exposure timing. All other parameters remaining constant.

With a "normal" 60W pearl bulb the exposure meter reported 1/90 sec.
With an 11W CFL, the meter said 1/60 sec.
With an 8W CFL the meter also read 1/60 sec.

While I'm only looking at _relative_ measurements, and I was concerned
only with light for reading by (hence seeing what was reflected off the
paper), it supported my view that CFLs were dimmer than tungsten bulbs.
The interesting result was that 8W bulbs threw just as much light in
a downwards direction as the 11W bulbs. I reckon this is because the
light is projected downwards off the end of the bulbs, not off the
body. Since 11W bulbs have longer bodies, but the same x-section, the
usable light output is the same. The sideways emitted light is largely
absorbed by the lamp shade (hence it's name!), even though it had a matte
white interior.

Now I'm quite happy to accept the manufacturers' claims that the absolute
light output, measured in laboratory conditions, is similar to tungstens.
However, I have convinced myself that in a normal domestic ceiling-hung
fitting, the shape of the bulb reduces the amount of usable light by far
too much.

(Sorry about the full quote, but I think it is all needed for the
context)

That is why lamp light output is measured by using an integrating
sphe

http://www.lightingsciences.com/stat...ing-sphere.pdf
http://physics.nist.gov/Divisions/Di...ela/photo.html

Lamp manufacturers will give the light intensity distribution of their
lamps - see p99 of:

http://www.gelighting.com/eu/resourc...luorescent.pdf

....and you'll note from the spectra on the same page that it is not a
black body spectrum. Your DSLR light meter's response will vary
according to the wavelength of light incident upon it. My camera will
automatically recognise daylight, tungsten, fluorescent, and candle
lighting and adjust exposure automatically: it is possible yours may
not have, and has got the exposure incorrect.

I'm not trying to defend CFLs here, but in order to criticise, the
measurements need to be correct, and I'm afraid you have not
controlled for all the significant variables.

I think it is Cadmium Selenide photodetectors that have a wavelength
response that is substantially similar to (but not identical to) the
human eye. A fair amount of work has been done to determine and codify
the differences so comparisons between different sources of light can
be made in a valid way. I suspect your quick check with the camera is
insufficiently rigorous. Sorry. You do bring up an intereasting point
about light distribution though. Now that you have mentioned it, I do
notice the same.

Regards,

Sid