Jump to content

Debayering a DSLR's Bayer matrix.


RAC

Recommended Posts

I thought i would have a go at the impossible. I've got two Canon 1000d's so i took it upon myself to either break it or make it. I looked all over the net and found nothing at all about doing this , no pics of anything! This could totaly change DSLR imaging if it works. Using an HA filter or any filter would now mean using every pixel.

Take one sensor.

8130310599_88111684ae_b.jpg

Canon 1000d CMOS sensor by meg rac, on Flickr

Remove the glass. It pops off slowly with a small screwdriver. All i'm going to do is pollish the front of it to get rid of that green layer as this is the bayer matrix. I used a dremel with a pad thing sharpend to a soft point.

Meguiar's ScratchX was the past of choice.

8130337428_1f42b26990_b.jpg

Canon 1000d CMOS sensor without front glass by meg rac, on Flickr

First test. You can see how more sensitive it is !!!!!!!!!!!!!!!!!!!!!!!!!!!!

8130309119_5492da7427_b.jpg

First test of seeing in the dark. by meg rac, on Flickr

Flat test. We wont be using the whole sensor as its very hard to get right out to the edges as there are lots of little wires so cropping will be a must. I'm going to have a better go at cleaning the sensor up.

8130308595_d2ee562863_b.jpg

Flat test by meg rac, on Flickr

So far for about an hours effort i think its a total success.

8130309713_e9f4502d43_b.jpg

Mono Canon 1000D! by meg rac, on Flickr

By the way i don't recomend you doing this if you only have one camera or you arn't good with small things!

  • Like 13
Link to comment
Share on other sites

Hmm... I think my 450D just gave me a nervous look! I have long wondered how hard would it be to remove the CFA, but never took it further... so if I understand correctly, you just need to try and remove the CFA layer without damaging or scratching the underlying surface? And in this case you used some kind of mild abrasive?

Actually, I might give this a go on my 300D IR mod as a test before I break the 450D :D

After that, just need to figure out how to configure software not to de-bayer a non-bayer...

Link to comment
Share on other sites

WOW! Now that is extremely interesting :) I have a couple of spare imaging assemblies from cameras I've already succeded in destroying with my extreme modding so this will be definitely a possibility for me. Now there's a surprise :D I've seen a couple of tries at debayering that didn't succeed.

Yes, I shall definitely have a go at this :) I love extreme mods :D I'll see if I can find that paste you quoted. Did you use it dry or with water or other lubricant?

I'm currently using a colour 1100D for NB and losing a lot of resolution which is a little bit crazy :D Actually, two sided by side - with Ha and OIII clip filters for widefield DSO imaging. I'm also in the process of making a filter wheel for my scope camera (another 1100D) and have NB filters on order.

It's good to see that the Bayer mask is on top of the microlenses :) And it looks like the mask is softer than the microlenses.

Edited by Gina
Link to comment
Share on other sites

After that, just need to figure out how to configure software not to de-bayer a non-bayer...
Ah yes - that's a point :eek: Hadn't thought of that! That is done in DSS, isn't it? I wonder if it needs a change by the software author or whether there's a way round it. Please let me know what you can find out :) Edited by Gina
Link to comment
Share on other sites

I don't know about DSS, but I think in many raw conversion processes there's an "ignore bayer" or similar option. I think PixInsight's raw conversion uses dcraw as a base, and that has an option to just convert without debayering. But I'm not sure where to find the options to do that in PI or other software.

Link to comment
Share on other sites

I don't know about DSS, but I think in many raw conversion processes there's an "ignore bayer" or similar option. I think PixInsight's raw conversion uses dcraw as a base, and that has an option to just convert without debayering. But I'm not sure where to find the options to do that in PI or other software.

Ah yes - maybe there is in DSS. I'll see if I can find anything.
Link to comment
Share on other sites

Been looking at the DSS manual - this page :- http://deepskystacke...al.htm#rawdecod

No mention of not debayering but there are different ways of using the pixels. I've been using the "Super-pixel" mode with NB filters and I imagine this would be alright with the Bayer mask removed - giving a vast increase in sensitivity at the expense of resolution. The "Bayer drizzle" mode looks interesting though I don't quite understand what they're saying. This may give the required result with full resolution - not sure. Something to try though. Think I'll post in the Software forum - someone may know the answer.

Link to comment
Share on other sites

All i used was a dremel with one of these tips(the rounded top one). I cut the tip a bit sharper. And this stuff straight on the end. Without the paste nothing happend. As so asl the clouds go away i'll get a photo. I might put this sesnor in my cooled camera or just run it uncooled for a quick test and compare imaging times verses my cooled colour one. Do a run of a few 5min subs at the same object for each camera maybe.

I haven't thought too much about the processing. I'll see what happens. It may not matter what the software does?

  • Like 1
Link to comment
Share on other sites

I also want to have a go with a test sensor(even an old webcam senor) with some form of paint stripper to see if it'll work without damaging the sensor. This would make it more easy and should give a cleaner finsh.

Anyway its a good thing i didn't listen to the internet and not give it a shot.

  • Like 1
Link to comment
Share on other sites

It's good to see that the Bayer mask is on top of the microlenses :) And it looks like the mask is softer than the microlenses.

Looking around the net it looks like the microlenses are on top of the bayer filter?. Haha well not anymore if they were!

Link to comment
Share on other sites

I also want to have a go with a test sensor(even an old webcam senor) with some form of paint stripper to see if it'll work without damaging the sensor. This would make it more easy and should give a cleaner finsh.

Anyway its a good thing i didn't listen to the internet and not give it a shot.

The micro lenses will be a moulded plastic so mind how you go with the paint stripper.

I'd also be worried about making the lenses 'frosted' with the dremmel - even using a soft tip?

Link to comment
Share on other sites

Looking around the net it looks like the microlenses are on top of the bayer filter?. Haha well not anymore if they were!

Well, whatever, your result clearly shows the increase in sensitivity is not reduced too much by your mod if you have removed the microlenses. The proof of the pudding is in the eating. Maybe the later sensors have the CFA on the top - manufacturers do make changes to their products :D And they often don't tell us - their customers.
Link to comment
Share on other sites

Removing the CFA layer should give a significant increase in resolution and sensitivity.

Yes, certainly would seem so :)
This company can supply cameras suitably modified for B/W, IR, UV and Ha work. Looks interesting :) I don't think I would have the bottle to try it on my own bodies lol
Not cheap but still cheaper than a mono astro CCD of similar resolution. Edited by Gina
Link to comment
Share on other sites

I have a Dremel - well actually the Draper equivalent, with the assorted tools, and I'll see if our local cheapo shop has the buffing paste you mention. If not I'll get it from Amazon :)

How difficult is it to know/see how far to grind down? It would be rather distressing to go too far! :eek:

Link to comment
Share on other sites

I guess if the microlenses are plastic and the CFA is also plastic but the sensor photodiodes and associated circuitry is silicon then the plastic layers being softer than the silicon can be eroded easily with little affect on the much harder silicon.

Whether the plastic could be disolved or softened with paint stripper I don't know but I'm pretty sure paint stripper wouldn't affect the silicon circuitry.

I think trials could be conducted with webcam sensors as these will probably be much the same as DSLR sensors but on a smaller scale. They certainly have Bayer masks but may not have microlenses.

Link to comment
Share on other sites

Out of interest, what is the difference between this mod and the normal mods where the filters are removed.

The filters are separate from the sensor - held in a plastic frame with glue and micro sticky tape. The Bayer mask is part of the sensor chip and it's removal involves removing the glass cover plate and then scraping away at the actual sensor. Extreme modding at it's very most extreme! Edited by Gina
Link to comment
Share on other sites

Thanks Gina, so what are the advantages, or are we about to find out.

Its like taking its sun glasses off. If you were to put an ha filter infront of a standard dslr only one out of four pixels would be getting the light but with the bayer filter removed all four bixels will be getting the same amount of light.

I'll find out what its like to use on the first clear night i get. I don't have an ha filter yet but i think i should order a clip in one. But even just for luminance the sensitivity is massive over a standard sensor.

I have never seen a test like this for a dslr yet so it will be very interesting.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.