Ever wondered what is so special about the "astronomy version" of the Canon 20Da? Check out:

http://www.astropix.com/HTML/M_DAP/M011/M011.HTM

The most interesting part is the first paragraph which I've pasted below in red

The Canon 20Da is a 20D that has been modified for astronomical photography. The three primary features being a low-pass filter in front of the CMOS sensor that lets more of the hydrogen-alpha wavelength through for red emission nebula, a function that allows focusing on the LCD on the back of the camera in real time, and reduced noise. The increased hydrogen-alpha light passing through the filter in the 20Da is supposed to be 2.5x greater than the 20D.

What does that mean for UW photography???

1) LOW-PASS FILTER modification
If you read more of that webpage you'll learn that digital cameras have a filter in front of the sensor to remove infrared light. Apparently it also removes some of the red light. The hydrogen-alpha light that is mentioned has a wavelength of 656nm, nicely deep red. So the 20Da has 2.5 times higher sensitivity in this range. BINGO this is the same spectral range where UW photography is struggling most because of red absorption by water. I've been digging up experimental data for light transmission in water and things rapidly get very bad between approximately 650-660nm and then stay about the same up to 685nm after which all goes to hell (but you're into infrared at 700 so that's ok). So the 20Da filter should help us most right where red absorption is worst. GREAT for MAGIC FILTERS!!!.

2) ALLOW FOCUSSING ON THE LCD modification
We just had a post with everyone raving about a new gadget that projects the eyepiece image onto a separate LCD panel (and pestering poor Ike to come up pronto with a new back plate to support it). If I don't get it wrong the 20Da apparently has that already taken care of while using the same LCD that is already part of the camera!!! So no reason to boost Ike's workload (or wallet ).

3) REDUCED NOISE
I expect there won't be too many people complaining about reduced noise, assuming it doesn't compromise other properties.


So I am wondering if anyone has had a look at the 20Da for scuba. I don't think your average photo shop will have a few of them in stock but perhaps someone on this list has an astronomy buddy or can get their hands on one. It would be great to know if this could really be as good as the little paragraph on the top suggests. It is a relatively expensive upgrade (US$2199 at B&H) but it sure looks nice.

To get back to the topic title, if the astronomy people could get Canon to do this, couldn't we push for a 20Ds with the same features but with a sensor filter that attenuates the blue and green so we have a magic filter build right into the camera. Just dreaming

Bart

I just followed the link on the original webpage to the HUTECH company

http://www.sciencecenter.net/hutech/canon/index.htm

The good news is this company can do the sensor-filter modification on either new or old Canon 300D, 350D, 20D 20Da, and 5D. They can either replace the filter by a transparent filter (apparently without any filter autofocus no longer works), or a better daylight filter that let's the camera behave as normal but without the blocking of deep red light. The modified Canon 350D would cost you US$1350, for the 20D it is US$1995, and the modified 5D is US$3795. The modification iself is about US$600. The 20Da which also has the other modifications may still be the most interesting.

The filter mentioned above sits right in front of the CCD (there's a nice picture on the hutech web page), but they can also install a user-replaceable filter between lens and body. If we can get a magick filter to go into that spot with a transparent filter in front of the sensor we could have the best of both worlds, and still be able to use the camera normally for topside shooting by inserting an IR filter.

Bart

PS They also do it for the Fuji finepix, but as far as I can see not for any of the Nikon cameras

Using the magic filters this way would be cool however they note that

"Front-mounted filters do not affect the camera's autofocus operation but, when installed, preclude the use of Canon EF-S lenses or some lenses designed for APS-C sized sensors."


Which might be a limitation for some.


Allen

Yes there are limitations but the inability to use EF-S lenses isn't a serious one in my opinion.

A more obvious disadvantage is that you have to shell out 600 to 1000 dollars extra for the Hutech modification or the full 20Da setup, respectively.
If you go for the fully transparent "filter" then you have the topside problem that you MUST use an IR filter because IR radiation gets picked up by both the blue and green pixels.
Another disadvantage is that you apparently can't do this with Nikon cameras (but perhaps the Nikon designers put in a better IR filter in the first place that doesn't degrade the deep red)

The "no EF-S" issue can be easily circumvented by not using the "front-mounted filter", but just using a normal filter screwed on the end of the lens. A better way is not to buy EF-S lenses for this purpose at all. Since the basic concept of magick filters is to re-establish the color balance by throwing away blue and green light the goal should be to get as much light as possible into the system. The more light you have, the more you can throw out, and the deeper you can, in principle, push the useful depth of magick filters. Since Canon EF-S lenses are not very fast they are not ideal. The 28mm f1.8, 35mm f2.0 and 50mm f1.4 (or the equivalent Nikon choices) are more interesting options, or the f1.4 L series primes if you can affort them.

A second requirement to push the magick deeper is to get filters that let through less blue and green light. The obvious current solution is to stack up two magick filters. Has anyone tried this?? One problem is that deviations for proper absorption compensation may not be a problem when you use a single magick filter, in other words the WB can correct a lot of imperfections, but when you stack two filters together the deviations get magnified so it would be interesting to find out if this is still acceptable. Other filters for "deep magick" (TM ) may need to be considered to get the best results, including specially designed thin-layer filters instead of absorption filters.

Magick filters were never meant to be the "total solution" to solve all UW shooting needs but from the stories we've heard and the pictures we've seen they sure have a lot to offer in the 5-15m depth range in clear tropical waters. That's were I hope to do a lot of my photography once I get my Canon G2 upgraded to a dSLR AND find the opportunity to practice instead of boring people on this list with theories.

Bart