Anyone know spectrometers?

Baldy · March 22, 2007

I need to produce graphs of light intensity versus wavelength for various light sources like flashes, light bulbs and the sun.

Does anyone know how this is done? Would spectrometers like these do the trick? http://www.photon-control.com/spectroscopy.html

Are there testing labs I can get this done at?

Thanks!
Baldy

ian408 · March 22, 2007

Scientific Equipment Repair
(650) 938-3169

Might have your answer.

Art Scott · March 22, 2007

Ring DOCIT...he's PHD is in Physics {or is still working on the thesis for it:D), isn't it...he should have all the answers.

The only spectrometer I have used was used to burn iron samples to give us chemical composition for the foundry I was a lab tech at....sorry I couldn't be more help.

ziggy53 · March 22, 2007

Baldy wrote:

I need to produce graphs of light intensity versus wavelength for various light sources like flashes, light bulbs and the sun.

Does anyone know how this is done? Would spectrometers like these do the trick? http://www.photon-control.com/spectroscopy.html

Are there testing labs I can get this done at?

Thanks!
Baldy

If you don't need too much accuracy, the PhotoShop "Levels" is a type of visible light spectrometer.

You would (ideally) take the lens off of the camera, diffuse the camera opening, capture an image pointing the camera body in the direction of the source, (RAW gives the best flexibility) balancing for the white balance (before or after), and finally viewing the Levels in PhotoShop.

It would be difficult to calibrate, but if you use known spectra, Sodium and then Lithium flame tests might be suitable markers for calibration.

If you just need an inexpensive qualitative scope:
http://webmineral.com/chem/spectra/scopeinstruct.pdf

Process the following image as above to see a fluorescent spectra:

Baldy · March 23, 2007

I've been looking at buying one of these for $1,500 and then selling it on eBay when I get done with it:

http://www.photon-control.com/spectroscopy.html

We're still getting killed by flash photos of fair caucasians taken with digital cameras where someone's face goes nuclear. We absolutely have to understand this problem and far as I can tell, no one does. Newborns are the worst.

There is something very different about the light falling on their skin from the flash than from sunlight, and something different again about the way digital cameras record it.

So I have to be able to produce charts like this:

I'll try various filters on the flash head and the camera lens to see what works, but I need to chart how the filters are modifying the light. I also need to measure what happens to the light when it bounces off ceilings. And what happens when the flash is at different angles of incidence to the face.

I think I know what's happening: red and near-infrared light is penetrating the surface layer of skin, reflecting off the hemoglobin (like red-eye), and the camera is able to record it. Film cameras never did. If you have much melanin in your skin, it reflects the light as it's designed to do, so no reflection off the hemoglobin and no too red problem. The face is where it happens because it's so vascular.

Medical researchers have published studies showing that people who eat lots of colorful fruits and veggies build up a reflective layer of caretenoids in their skin, which acts like melanin to reflect light, protecting their skin. In other words, the substances in colorful fruits and veggies that protect plants from sunburn get co-opted by us for the same purpose.

Problem is, 90% of the population doesn't eat much fruit or veggies, so they go nuclear when the flash fires. (And without a reflective layer, they get skin cancer if exposed to sun). They make the reasonable assumption that something went wrong in printing, because they see other shots where they don't look that way (outdoor shots).

wxwax · March 23, 2007

That's extremely interesting, Baldy.

Is your goal to produce an explainer for the victims?

claudermilk · March 23, 2007

Sounds really interesting. Do you happen to have any links to these studies?

I guess reading between the lines, the message is "eat your veggies!"

ziggy53 · March 23, 2007

Baldy,

Most modern electronic flash have UV filters incorporated into the flash tube or as a surface coating. (Studio flash tubes can be either UV filtered or not.) The lack of UV would probably explain the difference between flash and genuine daylight.

Notice on the following link (a little more than halfway down the page) that Nikon used to produce an unfiltered flash, the SB-140, that did not attenuate the UV, except with an external filter. (Some Honeywell flashes were likewise unfiltered, and could cause odd colorations in white clothing, including bridal gowns.)

http://msp.rmit.edu.au/Article_03/02a.html

The Mecablitz CT 45 spectrum is more indicative of most modern compact electronic flash units.

ziggy53 · March 23, 2007

Baldy wrote:

I've been looking at buying one of these for $1,500 and then selling it on eBay when I get done with it:

http://www.photon-control.com/spectroscopy.html

...

If you are looking for wavelengths in the UV/Visible, then model SPM-002-D is probably your best choice, for $1800USD.

Baldy · March 23, 2007

wxwax wrote:

That's extremely interesting, Baldy.

Is your goal to produce an explainer for the victims?

The #1 goal is to come up with ways to avoid it before the pixels hit the memory card. For pros, it may not be too difficult because it probably involves getting the flash off the camera, bouncing it off the ceiling, and/or putting a filter over the flash or lens. I'll come up with both photographic examples and light spectrum charts to show the effects of each of those.

For the poor person who wants to take pics of their newborns with a consumer camera, we'll see... Maybe we can have an influence on camera manufacturers when we have compelling data to show.

Baldy · March 23, 2007

claudermilk wrote:

Sounds really interesting. Do you happen to have any links to these studies?

I guess reading between the lines, the message is "eat your veggies!"

Here's a reference to the NIR flashes put out that makes coloration too red (last couple of paragraphs).

http://www.cci-icc.gc.ca/publications/newsletters/news34/science_e.aspx

There are a lot of references in the medical field to reflectivity of skin in the red and near-infrared regions because they are interested in detecting disease, such as skin cancer. In their case, melanin and carotenoids interfere with their imaging because it stops the light before it can hit the hemoglobin and reflect back, just the opposite of our problem. Here's one of the carotenoid references:

http://jn.nutrition.org/cgi/content/full/132/3/399

This is a terrible thing to say, but in the interest of scientific observation I'll say it anyway.... I most often see skin go nuclear under flash on someone who doesn't look like they get most of their calories from colorful fruits and veggies:

Permission to flame for such a terrible comment without enough quantitative data to back it up is granted.

DavidTO · March 23, 2007

Or wearing lots of make-up, like Arnold!

Baldy · March 23, 2007

ziggy53 wrote:

Baldy,

Most modern electronic flash have UV filters incorporated into the flash tube or as a surface coating. (Studio flash tubes can be either UV filtered or not.) The lack of UV would probably explain the difference between flash and genuine daylight.

Notice on the following link (a little more than halfway down the page) that Nikon used to produce an unfiltered flash, the SB-140, that did not attenuate the UV, except with an external filter. (Some Honeywell flashes were likewise unfiltered, and could cause odd colorations in white clothing, including bridal gowns.)

http://msp.rmit.edu.au/Article_03/02a.html

The Mecablitz CT 45 spectrum is more indicative of most modern compact electronic flash units.

Nice link, Ziggy, thanks for that.

I'm not focused on UV, which makes wedding gowns look blue. It does it by causing the anti-stain coating on wedding dresses to fluoresce—UV light falling on the dress reflects back to the camera in the visible (blue) range, where the camera can record it.

My problem is on the other end of the spectrum in the reds and near-infrared. The issue is exactly as your link pointed out: flash has a lot of it. Here's a reference to daylight, which has much more blue (which kills red) and much less red and near-infrared:

http://www.lrc.rpi.edu/programs/NLPIP/lightingAnswers/fullSpectrum/comparisons.asp

Anyone know spectrometers?

Comments