Addendum to The Noble Art of Optical Communications article

(To see full size picture, click on the thumbnails in the text.)

Well over a month after I had finished the article about communication by means of IR light, I ran into two things that I think are worth mentioning.

The first thing was an advert on the back of the Dec-04 issue of Sky & Teleskope. It was from Celestron, a well known manufacturer of amateur telescopes, about a CCD camera with some software that seemed very interesting. I had not even heard of this one! Go to their home page http://www.celestron.com/mainf.htm and look for “Neximage”!

Let it be said that it is a CCD (not CMOS) colour video camera with a 3.6 x 2.8mm size sensor. It has 640 x 480 square pixels of 5.6µm, with < 1 lux sensitivity. They can be run in several “binned” modes, like 2x2 pixels, making it a 320x240 camera. It will take quite a lens to resolve 5.6µm (179 lines/mm!) anyway. The camera is entirely powered from the USB connection. Supposedly it is the Philips PC camera, so other PC cameras may do as well.

The price ($150) was a little more than I was willing to pay, but some calling around revealed that the large NYC photo store B&H, http://www.bhphotovideo.com/  had them in stock for $100 and the shipping cost was half of what the California sales tax (7.5%) would have been…. That makes it what we call a “no brainer”.

The camera was a very pleasant acquaintance, working at least as well as one can expect, but the big deal is the software! You can download a PDF manual from the Celestron site and read about it! While photographing celestial objects the atmosphere is the limiting factor. Even a small (6”) telescope can resolve 1 second of arc, but the atmosphere is often worse than this. One way around it is to take a lot of pictures, select the best ones, and stack them in Adobe Photoshop or similar. The program that came with the camera, RegiStax, has some very advanced routines! It selects the good pictures from a video file and puts the bad ones on hold. Then it registers them so they are accurately on top of each other, in spite of the image moving around due to the atmospheric refraction or poor tracking of the telescope. When stacked, the noise cancels out and real features are enhanced. Finally the image detail is enhanced, using wavelet technology. All steps, exposure time, file size, number of images, running time, film speed (fps), gain, gamma, colour balance… are controllable. In my opinion, the program is worth more than the camera, and the camera is good! This is one attractive deal! It is relatively easy to make an adapter for the camera so it can be used with other than telescopes. The camera housing has a ~6mm long 28.5x0.75mm thread and the CCD is recessed ~9mm behind the start of the thread.

I have made a C16 adapter, as there are many good and fast lenses for 16mm film and TV cameras. They give a large enough image area, as a matter in fact, already a lens for double-8mm will do. Especially the lenses by Kern, for Paillard cameras seem to be among the best. Their Switar series are superb, and the Yvar lenses are just a little less fast.

One more advantage, for the IR experimenter, is that this camera does not have much of a filter, cutting IR. So it is fairly sensitive to it. With a wide angle C16 or D-8 lens, 5-8mm or so, it would become a good general purpose camera-tool for aligning IR optics.

I set up the equipment I had used in the article and dragged in my telescope in the lab so I could observe the spot on the shopping centre (some 200m away), put there with a 15mW IR laser diode and a 25/1.3 lens from an 8mm projector. The aiming of the telescope with such a small “film” was very difficult. The scope has a focal length of 2010mm so it is like a 20 meter telephoto lens for regular film! I put on a field flattener corrector cell that also shortens the FL to some 1350mm and increases the aperture from f/10 to f/6.3.

A shot with a regular, zoomed up, digital camera, in daylight, shows the “target area”:

The aiming point is marked by the three white bars. Just below it we can barely see a handrail, under which are vertical bars are separated horizontally by about 10cm.

A shot with the NexImage camera and the 1350mm telescope reveals some detail. This, and the following NexImage pictures are not cropped, and thus to the same scale, so the separation of the bars can be used as a rough scale. From this angle it is 10cm centre to centre.

This picture is taken at night, the only illumination is mostly the stray lights from the parking lot below the rail. The image was very noisy, and warm air streaming out of the window, mixing with the colder night air outside, made the image quite unstable. It is evident what a great job the RegiStax program did on the ~210 picture stack! There is virtually no noise left and the registration of the images is near perfect. I may not even have focused the telescope very accurately. It was not easy with a noisy image, moving around!

Anyhow, the IR spot was clearly visible. Here are two versions of the same image.

The darker “doughnut” is a speck of dust on the protection glass covering the CCD, casting a shadow.

It can be seen that the spot is on the order of 3 ~ 5cm and that it is round. The grainy texture is real, it was not moving or changing within the spot. I cold just barely see it in the noisy video, but the stacking got rid of the noise and brought out the features.

It should be said that I used the same 15mW Laser Diode with a 60mA DC current.

The spot size, as we experienced it when detecting the signal with a small detector diode and a receiver, seemed to be some 15x10cm, but I had no means to focus it at such distance, as I could not see the spot. Now it was easy. The signal strength in this spot must be many times more than the one we measured, but also more difficult to find and more prone to atmospheric “wiggle”. It may be noticed that the exit aperture of the Paillard projector lens is 19mm so if the spot is still 50mm after 200m, the collimation is excellent, a very good lens! As a matter in fact, a faint ring is visible around the spot in the pictures above. I believe this to be the first diffraction peak, the Aieres disc seen around stars in good telescopes. An artifact of a diffraction limited optical system.

Slightly de-focusing the projector lens on the laser diode gives a larger spot. In one direction it gave this peculiar figure! This could very well be the “10 x 15cm spot” we noticed as we were waving the small detector diode back and forth, trying to get an idea of the size of the spot that we could not see.

The second issue.

At the same time my son and I were roaming through some boxes with Old Noble Junk. We came across a night vision device of Russian origin. I used to import these and all went well except for with the Customs officers at LAX. They had their own agenda about what nations U.S. should trade with, they made life very difficult and the venture impossible.

It is my pleasure, and duty I think, to give them all the credit they deserve. The fact that Russia had “Most Favored Nation Trading Status” at the time did not bother them. In writing a Letter of Complaint to the Department of Commerce, under which the Customs are organized, I got the predictable response, praising the skills and professionalism of their officers. Obviously, they did not know what they were talking about. I was dealing with an excellent company ISTA in St. Petersburg, Russia. They are staffed with very nice, professional and patient people. The products they found for me were very good as well, with great potential for import to here.

Anyhow, this device, the least expensive of all kinds of NV devices available, is also quite IR sensitive. It is an electrostatic image intensifier tube, a technology used already in the WWII. I refrain from trying to determine what “Generation” it may belong to, as nomenclature in this field is a deep and messy swamp inhabited by creatures who knows it all. Anyhow, I looked for the spot and saw it easily. A device like this could also be useful in aligning IR optics but almost too sensitive for nearby work.

I took a picture through it, on free hand so it is not very sharp, but you can see the small spot above the hand rail, just to the left of a light pole, in the target area marked in an earlier picture. It was easy to focus the laser beam well, using the view in this scope.

The  setup, in my crammed lab, can be seen below:

The Celestron 8-inch mirror telescope in the middle. To the right, on a camera tripod, the Tx/Rx assembly used in the article. To the left of the telescope tripod is a laptop, displaying the picture from the NexImage camera mounted on the telescope.

Carl G. Lodström, Ventura, California, in November 2004.