Monoing a DSLR

[gregbradley]

Monochrome modified DSLRs are becoming more common. They seem to be quite good performers and open up narrowband imaging to a wider number of people.

Here is an article about how to modify a DSLR to monochrome.

I read on another CN thread that Nikon D90 and D5100s were easier to modify this way for some reason.

This site refers to Canon 1100D mostly but they offer larger DSLRs already modified as well.

http://www.jtwastronomy.com/tutorials/debayer.html

Greg.

[Geoff45 (Geoff)]

Looks interesting, but I wouldn't be game to try it.
Geoff

[gregbradley]

I'd love to see someone take the plunge and buy one and post their results. Its got to be the cheapest 12mp APSc mono sensor camera around.

Greg.

[rustigsmed (Russell)]

Originally Posted by gregbradley

I'd love to see someone take the plunge and buy one and post their results. Its got to be the cheapest 12mp APSc mono sensor camera around.

Greg.

yep you can get a new 1100d for $359 grey import off kogan. now that is a cheap mono camera.

[rally]

Greg,

Look here
Its a long read but you need to read it to get the full picture !

http://stargazerslounge.com/topic/16..._%20debayering

You'll find a lot of info on techniques and differing ways for differing DSLRs
You'll also find an IIS member who is active in this thread.

Rally

[Peter Ward]

Can't see the point.

The probability of causing the CMOS to fail I's suggest is *very* high, and by removing the micro-lens structure you are probably dropping the fill factor by 20-30% (I'd guess that to be similar the ND factor of the Bayer matrix).

Simply removing the IR blocking filter and CMOS cover slips I'd suggest is a way better mod.

[gregbradley]

Narrowband for a DSLR is the main gain.

Greg.

[rcheshire (Rowland)]

Alistair has done the same - ATM forum.

[rcheshire (Rowland)]

CMOSs are much more rugged than they appear. Agree though, the purpose of a microlens is to improve pixel performance. What cost a mono APS-C CCD, by comparison to a CMOS. I think, the main reason is a cheap alternative. Cooling, particularly the 1000D, 450D is dead easy, as well.

[gregbradley]

Originally Posted by rcheshire

CMOSs are much more rugged than they appear. Agree though, the purpose of a microlens is to improve pixel performance. What cost a mono APS-C CCD, by comparison to a CMOS. I think, the main reason is a cheap alternative. Cooling, particularly the 1000D, 450D is dead easy, as well.

Exactly. There is still a considerable gain in improved sensitivity despite losing the microlens if those who have done this are right.

Greg.

[alistairsam]

hi,

I'm active in that SGL thread and have debayered a 350d, working on a 600d. other one's we're working on are 450d, 600d, 1100d.

here are some results from a guy in portugal who used my wooden tool method to remove the cfa. all with a 350d.
http://www.flickr.com/photos/37419943@N08/9731109519/
http://www.flickr.com/photos/37419943@N08/9402541477/
http://www.flickr.com/photos/37419943@N08/9479299636/

this one is from a mono 450d
https://dl.dropboxusercontent.com/u/...ack_L_full.jpg

some results from a mono 600d
http://www.jtwastronomy.com/products/monochrome.html

to summarize, the biggest issue is removing the cover glass.
what we've found is that 1000d is the easiest to remove the glass. comes off in one piece. 350d is the next. 450d is the third hardest.
1100d is proving the toughest.

some have had success by using a hot air gun on the glass to break the bond but there have been a few failures due to heat. it worked when she used a cold finger underneath the sensor while heating the glass.
requires hot air at 320C for about 15 seconds and you see fringe patterns under the glass where the bond starts to break.

cfa on the 350d is the easiest to remove, 1000d as well, but its quite noisy although its easy to cool.

I've got my 600d ready to be debayered. will post results.

I guess as Greg mentioned, possibility of narrowband at a fraction of the price of a mono is the biggest attraction.
Loss of microlenses is more than made up by gain in 3 pixels. so that's not noticeable at all.

with a QE of around 30 to 40%, they're far less sensitive than a mono, but again, it comes down to cost.
even if you ruin one or two sensors, you'd have only lost $200 odd. so you'd get a large mono chip for a very low outlay if you succeed.

and if you are capable of adding cooling, then all it requires is longer subs to get decent results.
it'll never compare to a mono ccd, but it does get close as in this comparison I did with a mono 350d and a qhy9
https://www.dropbox.com/s/rvij6c77e9...vsmono350d.jpg

will post results in the atm thread.

Alistair