A simple DSLR spectrograph

For those interested in imaging and image processing with DSLR, CCD, webcam or film.
Post Reply
Message
Author
User avatar
rwilkinson
Member
Posts: 1286
Joined: Mon Oct 12, 2009 6:47 pm
Location: Bolton
Contact:

A simple DSLR spectrograph

#1 Post by rwilkinson » Sat Jun 18, 2011 3:01 pm

Following my experiment with a DIY spectroscope described elsewhere: http://www.boltonastro.org.uk/Forums/vi ... ?f=5&t=168
I've built a spectrograph version, designed to image the spectra easily with a DSLR camera:
spectrograph.jpg
A cardboard-box spectrograph with DSLR fitted.
spectrograph.jpg (181.79 KiB) Viewed 4706 times
And here are a couple of spectra which I captured with it:
daylight.jpg
Continuous daylight spectrum showing some absorbtion lines
daylight.jpg (7.01 KiB) Viewed 4706 times
lamp.jpg
Line spectrum from fluorescent lamp
lamp.jpg (9.29 KiB) Viewed 4706 times

User avatar
rwilkinson
Member
Posts: 1286
Joined: Mon Oct 12, 2009 6:47 pm
Location: Bolton
Contact:

Element identification from the spectrograph

#2 Post by rwilkinson » Sat Jun 18, 2011 7:28 pm

I've now worked out how to use the IRIS and VisualSpec programs to convert my rainbow-stripe pictures into a diagram of intensity v. wavelength:
daylight1.gif
Diagram from VisualSpec plotting intensity v wavelength (in Angstroms), indicating the absorbtion lines for particular elements
daylight1.gif (5.58 KiB) Viewed 4703 times
The dips in the graph are absorbtion lines which relate to particular atoms and molecules in the Sun (and the Earth's atmosphere) - so now we can see what our Sun's made of!

User avatar
rwilkinson
Member
Posts: 1286
Joined: Mon Oct 12, 2009 6:47 pm
Location: Bolton
Contact:

Re: A simple DSLR spectrograph

#3 Post by rwilkinson » Mon Jun 20, 2011 6:51 am

The VisualSpec profile from my colour camera (Pentax K110D DSLR) is clearly disrupted by the pass-bands of the Red, Green & Blue filters in front of its CCD sensor, so I tried my 1004x-JG monochrome camera instead:

This is clearly much flatter across the visible region, and its much lower-resolution CCD is still good enough to show the spectral lines.

At first it seems strange to use a black & white camera on colour spectra - but remember that the diffraction process spreads the colour information into a linear dimension, so the intensity vs distance is all that's required.
In fact the VisualSpec program works on monochrome spectra anyway.
Attachments
mono1.gif
VisualSpec's profile of daylight spectrum captured with my 1004X-JG monochrome camera
mono1.gif (4.68 KiB) Viewed 4510 times

User avatar
rwilkinson
Member
Posts: 1286
Joined: Mon Oct 12, 2009 6:47 pm
Location: Bolton
Contact:

Re: A simple DSLR spectrograph

#4 Post by rwilkinson » Wed Jan 18, 2012 2:48 pm

Here's the previous daylight spectrum corrected for the response of my camera:
daylight.png
Daylight (Solar) spectrum after correction for camera response
daylight.png (8.04 KiB) Viewed 4478 times
and the re-synthesised colour spectrum:
daylight-col.jpg
Colour spectrum of the Sun synthesized in VSpec
daylight-col.jpg (16.24 KiB) Viewed 4478 times

Post Reply