Giant Magnesium Lines in Lovely Green Colour

[thesmiths]

There were some comments on our previous spectroscopy images that asked, what happened to all the lovely colours? Well, the monochrome DMK51 does have higher resolution and is actually easier to use than a colour camera. But we went back to using the Canon 40D as an experiment. The Canon does have a much bigger sensor (3888 pixels wide vs 1600 for the ICX274 chip in the DMK51). The pixels are quite a bit larger though (5.7 vs 4.4 microns) and then there's the Bayer matrix, which in the green lowers the resolution by another factor of two. But real colour is a great thing to capture. We have actually just started putting eyepieces into our spectrometers to see the spectra in real colour with full "human eye" resolution. The view is pretty stunning. Maybe we all get too fixated on imaging sometimes.



So here is out latest setup with our Canon 40D installed. The differences from our last setup (where we detected the solar Doppler rotation) is the choice of lenses and the new foam board and cardboard box we built around it. The collimating lens is now a 100mm f2.8 Fujinon and the imaging lens is a 150mm f4 Pentax (with a 2x APS tele-converter, making it a 300mm f8). Along with the 5 micron slit and the 100mm f4 "solar telescope", these lenses seem a good match for the intended observations. As before, we use a 1800 l/mm grating and the angle between the two lenses is 14.5 degrees.



This is a view of the grating with the collimating lens on the left and the imaging lens on the right. You can actually see the reflection of the imaging lens in the grating. The angle is set for approximately 77 degrees, which is second order in the green (-2 actually) and you can see what a very small cross section of the grating that is visible. The angle is chosen to detect the "magnesium triplet" lines near the wavelengths of 517nm and 518nm that are one of the major features of the solar spectrum.



Just to see how far we've progressed in a few months, here is a photo of one of our earlier setups that also used the Canon 40D for imaging. In this case, the imaging lens was a 200mm Canon f2.8L and the collimator was a 1 inch f5 achromatic lens (the one closest to the camera). The slit was 10 microns and the angles were set for first order (the grating is rotated about 8 degrees from normal to the camera). The image below compares the results of the old and new setup (click on the image to view in full resolution):



At the top is one of our earlier magnesium images taken with the older setup and at the bottom is a new image taken at "super high resolution". The whole width of the image is now just 4.2nm and the "magnesium triplet" fills almost half the image. The centre of the lower image is at 517nm. We've put arrows to show which lines in the new image correspond to the lines in the old image. The new one is so "magnified" it's hard to see the forest from the trees, as they say. But such a pretty green, isn't it?

[swisswalter]

Hi Douglas

that looks really professional

[marktownley]

Lots of fun and way cool!

[Montana]

Wow!!! I love the colour! absolutely amazing lines : : :

Alexandra

[solarchat]

great work thesmiths... I am just constantly fascinated at all the data I see in my spectrograph. Still trying to find the best way to get it out to the public but it is really awesome.

[colinbm]

Fantastic spectroscopy guys
Col

[Merlin66]

Well done!
(IMHO I would have gone for the DMK and used the artificial "colouring" available after calibration...)
Very nice result nevertheless.

[thesmiths]

To Merlin: yes, I think you're right, the monochrome DMK51 does give sharper results (click on image to see full 1600 pixel wide resolution):



I looked at the colour balance of the Canon 40D magnesium image and amazingly it was 100% green, with basically no contribution from red and blue across the whole spectral range (which is of course extremely narrow here). I then just set the blue and red to zero on the monochrome image.

The smaller pixels of the DMK51 mean the image is 5.7/4.4 or about 30% more "enlarged" than on the Canon 40D. Some of the very faint lines are likely due to atmospheric absorption and the intensity of these may change from day to day and the time of day (e.g. if they are from water vapour).

I recall in our very first attempt at spectroscopy in September using a Paton Hawksley spectroscope, the "magnesium triplet" was resolved as a narrowly spaced doublet. Now it takes up almost the whole width of a DMK51. That's progress!

[Merlin66]

Nice image!
If you check the BASS2000 solar spectrum you'll see the majority of the "other" lines are Fe...
(Soft focus on the right hand side?)

[thesmiths]

To Merlin: yes, the focus does seem to vary slightly even over such a narrow spectral range. So I guess it's related to geometry. Any ideas what it could be? On another issue, do you know what causes the "smile" (as C. Buil calls it)? And how to remedy it at the instrument level rather than in post-processing?

[Merlin66]

The focus is probably due to field curvature and some achromatic aberration (lens aberrations) you could try tilting the camera slightly relative to the lens.
In the Lhires and the Spectra-L200 the "sweet spot" is only around 10mm - you need to refocus beyond that.
"Smile" is a natural distortion caused by the combination of optics and dispersion. It can't be removed within the system (it's even worse when you play with prisms!!) IRIS can fix it for you.