Halogen on the left, Ott on the right.
CD reflection spectrum
Moderators: rjlittlefield, ChrisR, Chris S., Pau
CD reflection spectrum
It was brought up in another thread about getting a light spectrum with the reflection off of a CD. I know that there is nothing scientific about this, but I did make images of the spectrum of light coming off of a Halogen lamp and from a compact fluorescent (Ott light). I processed both as RAW images and in ProPhoto RGB to give a larger gamut (not sure if it is useful or not, but red tended to overload in sRGB pretty badly). I put the WB to 5000K, 0 tint (not sure where to put it). I always notice a slight green tint to my images with an Ott light - looks pretty strong in the green on the spectrum (a phosphor in that region?)
Halogen on the left, Ott on the right.
Halogen on the left, Ott on the right.
-
rjlittlefield
- Site Admin
- Posts: 23626
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Mark,
This may be a silly question, but did you include anything like a slit when you were making these spectra? The reason I ask is that the fluorescent spectrum seems way too continuous even for an Ott. What I'd expect to see is more like at http://laser.physics.sunysb.edu/~wise/w ... s/jasleen/ , half-way down the page.
--Rik
This may be a silly question, but did you include anything like a slit when you were making these spectra? The reason I ask is that the fluorescent spectrum seems way too continuous even for an Ott. What I'd expect to see is more like at http://laser.physics.sunysb.edu/~wise/w ... s/jasleen/ , half-way down the page.
--Rik
If you are going to not use a slit you have to view the source from way across the room. Then with the source very small angularly speaking
you will see in the case of the compact fluorescent bulb, multiple images of the radiant tube in the different colors of the peaky phoshors.
In the case of something like an Ott light you will still see mercury peaks in blue green and yellow
The ott light will have a sort of contiuous spectrum. CFL usually don't
They are a bunch of lines.
you will see in the case of the compact fluorescent bulb, multiple images of the radiant tube in the different colors of the peaky phoshors.
In the case of something like an Ott light you will still see mercury peaks in blue green and yellow
The ott light will have a sort of contiuous spectrum. CFL usually don't
They are a bunch of lines.
Second try.
I don't have a proper slit to use, but I cut one out of cardboard. Had a hard time with the LED. I found that If I closed the aperture as far as it wuold go it helped make the lines visible (f45 in my case). I first had the slit vertical and that wasn't useful
Then I figured out the the slit should be horizontal. 
The Ott showed nice lines. The LED was more difficult to set up. The basic structure seems similar to the Ott, but the indigo line is way strong on the LED - I have noticed a lsight purple cast to my pics with the LEDs.
Ott on the left, LED on the right.
I don't have a proper slit to use, but I cut one out of cardboard. Had a hard time with the LED. I found that If I closed the aperture as far as it wuold go it helped make the lines visible (f45 in my case). I first had the slit vertical and that wasn't useful
Then I figured out the the slit should be horizontal. The Ott showed nice lines. The LED was more difficult to set up. The basic structure seems similar to the Ott, but the indigo line is way strong on the LED - I have noticed a lsight purple cast to my pics with the LEDs.
Ott on the left, LED on the right.
Those are exactly what would be expected.
Pretty good on try two. You can get the lines sharper on the fluorescent light
by getting a bit narrower slit. If you look at the lamp from across the room you can see an in focus image/reflection in each of the spectral line colors.
The LED spectrum looks just like you would expect the graph linked to in the other thread to look. The one at ocean optics.
Pretty good on try two. You can get the lines sharper on the fluorescent light
by getting a bit narrower slit. If you look at the lamp from across the room you can see an in focus image/reflection in each of the spectral line colors.
The LED spectrum looks just like you would expect the graph linked to in the other thread to look. The one at ocean optics.
-
rjlittlefield
- Site Admin
- Posts: 23626
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Bingo -- you have made a functioning spectrograph!
About the f/45, I suspect your lens wasn't focused on the slit and the small aperture was just giving you enough DOF to work. Can't think of any other reason.
After you've played with these things for a while, you'll figure out how to get quick indications without much setup at all.
See the image below for a quick indication of a Luxeon V Portable LED. Overall view on the left, zoomed on the spectrum in the middle, exposure whacked way down on the right. Left side f/2.8 @ 1/40, middle f/4.8 @ 1/80, right side f/8 @ 1/1000. All shots with a handheld Canon A710. I had to manually focus the right two shots because of course the camera kept wanting to focus on the CD instead of the reflection of the LED.
The big fan-shaped spectra on left and right are due to an overhead diffused fluorescent. You don't see lines in those because there was no slit and the light was big.
--Rik
Edit: to correct brand name of LED
About the f/45, I suspect your lens wasn't focused on the slit and the small aperture was just giving you enough DOF to work. Can't think of any other reason.
After you've played with these things for a while, you'll figure out how to get quick indications without much setup at all.
See the image below for a quick indication of a Luxeon V Portable LED. Overall view on the left, zoomed on the spectrum in the middle, exposure whacked way down on the right. Left side f/2.8 @ 1/40, middle f/4.8 @ 1/80, right side f/8 @ 1/1000. All shots with a handheld Canon A710. I had to manually focus the right two shots because of course the camera kept wanting to focus on the CD instead of the reflection of the LED.
The big fan-shaped spectra on left and right are due to an overhead diffused fluorescent. You don't see lines in those because there was no slit and the light was big.
--Rik
Edit: to correct brand name of LED
Last edited by rjlittlefield on Mon Dec 15, 2008 11:50 pm, edited 1 time in total.
-
rjlittlefield
- Site Admin
- Posts: 23626
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Caution: deranged tinker at work!
I couldn't resist adding a slit to the LED. Same setup as before, but with a couple of pieces of black tape stuck on the LED itself. Not exactly high-tech, but good enough to bring out a strong red emission line. That's probably the one that also shows in the high-res spectrum HERE -- the page labeled "SUREFIRE U2 DIGITAL ULTRA" and graph labeled "Spectrographic analysis of the LED in this flashlight (lowest setting)."
Fun stuff...but I need to go back to work now...
--Rik
Edit: to replace obsolete link.
I couldn't resist adding a slit to the LED. Same setup as before, but with a couple of pieces of black tape stuck on the LED itself. Not exactly high-tech, but good enough to bring out a strong red emission line. That's probably the one that also shows in the high-res spectrum HERE -- the page labeled "SUREFIRE U2 DIGITAL ULTRA" and graph labeled "Spectrographic analysis of the LED in this flashlight (lowest setting)."
Fun stuff...but I need to go back to work now...
--Rik
Edit: to replace obsolete link.
Last edited by rjlittlefield on Sun Jun 13, 2010 10:07 pm, edited 1 time in total.
I can vaguely see that LED red line. LED seems to have a more continuous spectrum in the reds and greens than the Ott. I just don't get the lines from it. I modified my slit and made it out of paper tape. I also show the difference in open aperture vs closed aperture. The lines get cleaner the more I stop it down. LED se
All four side by side
Ott, warm fluorescent, LED, Halogen
All four side by side
Ott, warm fluorescent, LED, Halogen
Rik
I suspect that red line is actually the image of the slit rather than a spectral line. We would not expect an LED to have a any line structure other than a peak in the excitation of a white one. But even that is a band not a line. What you see from a compact fluorescent , those are lines.
Try moving your head with respect to the red line to see if parallax will shift it to another color. With many configurations of spectroscope you can see the slit. Some actually use the slit as a wavelength index. You turn a micrometer and put the line of interest on the slit and then read the wavelength. (Beck hand specs)
The link you included to the guy with LED spectra is great. I used to look at his stuff a very long time ago before he got his Ocean Optics Spectrometer
edit: I stand corrected . The LED guy's website DOES show some whites with a little sharp peak in the RED. Must be Warm Whites.
His spectra demonstrate a couple of interesting things.
First of all in the many white LEDs he measures the blue peak which is the
the excitation light for the white phosphor drifts all over the place with respect to wavelength. Anywhere from around 420 nanometers to 480 which is what you get with a Silicon Carbide blue LED as exciter.
Second the peak can be also of varying breadth (Called FWHM =Full Width Half Max if memory serves) depending on the manufacture of the LED and its operating conditions. The location and breadth of the peak affects the shape of the trough before the phosphor kicks in.
Third , every one of the different white LEDs had a distinctive and unique spectrum. The peak and the width of the phosphor spectrum demonstrates variation as well though not quite as much as the excitation.
This really means they are not something you would want to use for photography. Unless you were going to standardize on them and only use one set that you got zeroed in , if zeroing in is even possible. They DO have a long life.
In contrast to this your friendly quartz halogen bulbs of the same type and the run at a constant voltage yields black body curves that are repeatable enough for convenient use for photography. (probably plus or minus about 20 degrees K). That is also why they are used for absorbtion spectroscopy sources in the visible range. (Selected and with regulated supplies) A regulated DC supply for your qh photo lamps might not be a bad thing. Maybe a decent volt and/or ammeter too.
The only other suitable source for spectroscopy in the visible range is the Xenon short arc lamp or flash lamp. OK Carbon Arcs too.
I suspect that red line is actually the image of the slit rather than a spectral line. We would not expect an LED to have a any line structure other than a peak in the excitation of a white one. But even that is a band not a line. What you see from a compact fluorescent , those are lines.
Try moving your head with respect to the red line to see if parallax will shift it to another color. With many configurations of spectroscope you can see the slit. Some actually use the slit as a wavelength index. You turn a micrometer and put the line of interest on the slit and then read the wavelength. (Beck hand specs)
The link you included to the guy with LED spectra is great. I used to look at his stuff a very long time ago before he got his Ocean Optics Spectrometer
edit: I stand corrected . The LED guy's website DOES show some whites with a little sharp peak in the RED. Must be Warm Whites.
His spectra demonstrate a couple of interesting things.
First of all in the many white LEDs he measures the blue peak which is the
the excitation light for the white phosphor drifts all over the place with respect to wavelength. Anywhere from around 420 nanometers to 480 which is what you get with a Silicon Carbide blue LED as exciter.
Second the peak can be also of varying breadth (Called FWHM =Full Width Half Max if memory serves) depending on the manufacture of the LED and its operating conditions. The location and breadth of the peak affects the shape of the trough before the phosphor kicks in.
Third , every one of the different white LEDs had a distinctive and unique spectrum. The peak and the width of the phosphor spectrum demonstrates variation as well though not quite as much as the excitation.
This really means they are not something you would want to use for photography. Unless you were going to standardize on them and only use one set that you got zeroed in , if zeroing in is even possible. They DO have a long life.
In contrast to this your friendly quartz halogen bulbs of the same type and the run at a constant voltage yields black body curves that are repeatable enough for convenient use for photography. (probably plus or minus about 20 degrees K). That is also why they are used for absorbtion spectroscopy sources in the visible range. (Selected and with regulated supplies) A regulated DC supply for your qh photo lamps might not be a bad thing. Maybe a decent volt and/or ammeter too.
The only other suitable source for spectroscopy in the visible range is the Xenon short arc lamp or flash lamp. OK Carbon Arcs too.