Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Debayering a DSLR's Bayer matrix.


RAC

Recommended Posts

I gave up when my expenditure of DSLRs that weren't working from ebay reached the cost of a proper astro camera.  I only took it that far because I hate being defeated!  Proper astro cameras are a lot better than a debayered DSLR.

  • Like 2
Link to comment
Share on other sites

I got my second sensor, and messed that one up.  I had a slip of the X-acto knife as I was gently prying up the cover glass.  I didn't break any gold wires, but I did scratch the sensor surface.  Probably a bit too deeply.  Half the frame is black, the other half has wave after wave of wide, darker horizontal lines scrolling up through the picture.  But, I am getting better at removing the Bayer matrix.  The oak dowel sharpened up did work for roughly removing the layers.  I added a touch of polishing material to help get the surface cleaner.  The scratch was just too much, though...

In any case, I ordered one more sensor.  I may mess with the half-working one for now, just to see how it works out.  But, I'm not sure I'll modify the 3rd sensor, yet.  Maybe I'll give it one more go, but I may just skip it, install the stock sensor as-is, and save up a little more for a cooled CCD.  Fun project, though.

Edited by Noobulosity
Link to comment
Share on other sites

  • 3 weeks later...

Well, sensor #3 just showed up in the mail today.  I'm still not 100% positive on trying this again, but I completely understand the feeling of determination in succeeding at this.  I really want to try it again.

Link to comment
Share on other sites

Huzzah!  Replacement sensor has been debayered!  It's not perfect, by any means.  There's still a little debris around the edges and a few specks on the center portion, but it works.  A few images...

48294708741_d8d132ae5f_b.jpg

A close-up of one of the corners, you can see the depth of debayering isn't perfectly even, but I think flats will help smooth it out.  (Click for large version.)
48294806307_f242180353_b.jpgTest Flat - Closeup Corner by Eric Schubert

Test shot out the window.  I'm not 100% sure on how DCRAW is doing its thing.  A close look still shows a few pixels here and there in the field as magenta.  But, we'll see how astro images turn out.  That could just be sensor noise that will disappear with calibration frames.
48294708451_249788eba6_b.jpg

 

  • Like 4
Link to comment
Share on other sites

Tested the latest sensor last night. It's definitely noticeably more-sensitive to any light coming in. However, I noticed some odd bright streaking just left of center. I'm guessing it's because of how I removed the film from the sensor in that area. Maybe I can even it out. Or maybe it's because of micro scratches. No idea, yet.

In the open area next to it, I was able to get decent data on M8, Lagoon Nebula in narrowband H-alpha. The example here is heavily-cropped to get just the nebula in the lower right of the full frame.

13x 2.5' light frames, ISO 200 (Yep, 200!)

Included a few darks, a dozen flats taken against a solid red computer screen, and a couple dozen bias frames.

Only editing done here is minor stretching, cropping, and converting to full B/W.

48313277862_f5a73bbcfd_b.jpg

48313131756_8596f2df03_b.jpg

  • Like 2
Link to comment
Share on other sites

5 hours ago, Noobulosity said:

Tested the latest sensor last night. It's definitely noticeably more-sensitive to any light coming in. However, I noticed some odd bright streaking just left of center. I'm guessing it's because of how I removed the film from the sensor in that area. Maybe I can even it out. Or maybe it's because of micro scratches. No idea, yet.

...

Hi, may i give you a little tip after succesfull debayering 15+ nikon d600/d610 cameras? 

Never use heat on the glass, you may solder out the chip or overheat the board. The glass has to be removed mechanically. 

Never use [removed word] like wooden sticks or polisher paste or something like that.. you will get broken lines and never get rid of dust and leftovers. Also dont cover the golden wires with an epoxy, they will be teared as temperature rise.

Debayering is a CHEMICAL process, with right chemicals your sensor will be mirrolike. 

I add a pic of bayer filter curls up into tube to illustrate how it works. Also i can make debayered imx128 sensor (d600, d610, a7, a7-2) for a reasonable price. In nearest future i will try a7r and d800, i suppose the process will be the same.

And a debayered full frame is of course better camera than someting like asi1600mm, it is just 4 times bigger with bigger pixels and strong processing.

 

20190309_152647.jpg

20190418_181933.jpg

  • Like 1
Link to comment
Share on other sites

10 hours ago, YellowPacman said:

Hi, may i give you a little tip after succesfull debayering 15+ nikon d600/d610 cameras? 

Never use heat on the glass, you may solder out the chip or overheat the board. The glass has to be removed mechanically. 

Never use [removed word] like wooden sticks or polisher paste or something like that.. you will get broken lines and never get rid of dust and leftovers. Also dont cover the golden wires with an epoxy, they will be teared as temperature rise.

Debayering is a CHEMICAL process, with right chemicals your sensor will be mirrolike. 

I add a pic of bayer filter curls up into tube to illustrate how it works. Also i can make debayered imx128 sensor (d600, d610, a7, a7-2) for a reasonable price. In nearest future i will try a7r and d800, i suppose the process will be the same.

And a debayered full frame is of course better camera than someting like asi1600mm, it is just 4 times bigger with bigger pixels and strong processing.

 

20190309_152647.jpg

20190418_181933.jpg

Thanks for the tips!  I'd love to learn a bit more about the process you use to remove the CFA.  Would you be willing to share what steps you take and which chemical(s) are involved?

I can agree about the debris...  that stuff is hard to clean off the sensor after scraping it up.

Link to comment
Share on other sites

Got a chance to mess with the camera again tonight.  Looks like we'll have clouds for at least a few days, so I wanted to jump on the chance to test it while the night sky was clear.

I don't have time to process the pics tonight.  However, I was able to pin down the source of that weird illumination in my images.

I was imaging with the rear screen on.  It may be specific to live view, though I didn't test LV vs just having the info displayed.  But, I turned off the screen and VOILA!  The odd illumination was eliminated!  I'll post pics after I get to processing them tomorrow.

For now, I'm off to bed!  Clear skies!

  • Like 1
Link to comment
Share on other sites

Got a chance to quickly process my images over lunch today.  Here's a monochrome stack of the entire frame.  I'm sure I could pull out more detail with some time, but I didn't have a lot of time to work on it, yet.  Definitely seems like it's a lot more sensitive.  I'm hoping to compare this with a quick stack of about the same integration time from my unmodified Canon 7DII.

48323822237_bf95881059_b.jpg

  • Like 2
Link to comment
Share on other sites

15 hours ago, Noobulosity said:

Thanks for the tips!  I'd love to learn a bit more about the process you use to remove the CFA.  Would you be willing to share what steps you take and which chemical(s) are involved?

I can agree about the debris...  that stuff is hard to clean off the sensor after scraping it up.

I will trade the process for sony sensors for the process for canon ones. My composition/times/proportions does not work on 6/5dm3. BTW Heat is ok for canon glass removal, Sony glass on the imx sensor ceramic base will laugh at you until you burn it to hell. 

I can buy the d600/d610 sensor on ebay, remove the cfa, put one 1mm glass back (no infitity) and send it to your address.  Changing sensor in a camera is very easy, you can make it yourself. Details can be discussed, sorry for my russian english..

Link to comment
Share on other sites

8 hours ago, YellowPacman said:

I will trade the process for sony sensors for the process for canon ones. My composition/times/proportions does not work on 6/5dm3. BTW Heat is ok for canon glass removal, Sony glass on the imx sensor ceramic base will laugh at you until you burn it to hell. 

I can buy the d600/d610 sensor on ebay, remove the cfa, put one 1mm glass back (no infitity) and send it to your address.  Changing sensor in a camera is very easy, you can make it yourself. Details can be discussed, sorry for my russian english..

I appreciate that you want to get something back from investing in your time and efforts.  But I already have a monochrome conversion on my Canon 7D now.  I just want to clean up the CFA around the perimeter, if possible.

Also, this thread was started with an aim at experimenting, trading ideas, and coming up with a way to do this ourselves.  It's fun and interesting.  I'm not interested in making any money on this, and I'm not looking to pay anyone to do it for me (or pay for the formula, or pay for sensors already converted).  If you're not willing to share your chemical process, that's okay.  But it seems you're more interested in getting business from this thread than contributing to it.  You've mentioned multiple times now that you know how to accomplish CFA removal, but you haven't mentioned any of your methods.  There are multiple places here in the US where I can send my camera to have it converted.  The whole point here is doing it ourselves.

Edited by Noobulosity
  • Like 4
  • Thanks 1
Link to comment
Share on other sites

Across the web there are multiple solvent options for Sony sensors including acetone and MEK.

The Canon ones are epoxy based and the most you can expect to achieve is softening of the matrix..

 

After 102 pages is there any consensus on the safest approach to Canon debayering? I might have a go with my 1000D

 

 

Link to comment
Share on other sites

I've come across references to using 'photoresist remover' which I think is based on a solvent called N-Methyl-2-Pyrrolidone (NMP). I imagine it requires heating too. I've no idea where you'd get NMP from to try it, unfortunately.

Louise

Link to comment
Share on other sites

Doing a Google search just for "photoresist remover" popped up an interesting hit on the first link; a PDF from Micro Chemicals GmbH describing different options for removing photoresist material.  In the white paper, they mention NMP, as well as alternatives (because NMP is apparently quite toxic).  Looking at another link to their site, I'm curious if the "Image Reversal" or "cross-linking" (positive or negative) lift-off options might be viable.  Though, as Louise said, I don't know where you'd get small quantities of this stuff.  Looks like typical sales volume is 6x 2.5L bottles or larger sizes.  Maybe they'd be willing to ship samples for a small fee?

Link to comment
Share on other sites

I think NMP is potentially toxic because it's easily absorbed through the skin so gloves a must. It has a low vapour pressure so fumes not a problem. I came across some discussion here: https://www.dpreview.com/forums/thread/4075910 The thing is there's no telling exactly what compounds a particular manufacturer uses to make their CFAs and it might vary quite a lot even for the same maker. So it's probably a matter of *trying some solvents and seeing. I think they must be designed to be pretty resistant to most common solvents. You could try signing up to Researchgate and asking a specific question: https://www.researchgate.net/post/Which_solvent_can_be_used_to_dissolve_the_cured_polyester_epoxy_resin

Also, could maybe just try some common hydrocarbon solvents (aliphatic - no benzene rings!): http://www.ccc-chemicals.com/blog/hydrocarbons-aliphatic-hydrocarbon-solvents-petroleum-hydrocarbon-solvents/

Louise

* that's what damaged sensors are for!

 

 

Edited by Thalestris24
  • Like 2
Link to comment
Share on other sites

Picked up some dimethyl sulfoxide (DMSO), based on the info I found. Applied it to the sensor and face it some time. Unfortunately, it doesn't seem to be working. But I'll let some of it sit and soak on the CFA for now and see if it gets any better. But right now it hasn't really softened at all. I think we can probably rule that one out.

Link to comment
Share on other sites

Came across this:

https://www.microtonano.com/TIN-Removing-Cured-Epoxy.php

You can get dichloromethane online but probably via carrier, not through the post. I think it's used in products like tank cure or fertran that dissolve epoxy seals in petrol tanks. Xylene and paint stripper are possibilities. You probably need to include heating to soften the coating.

Louise

  • Thanks 1
Link to comment
Share on other sites

I've seen DCM mentioned multiple times, but it sounds like that stuff is pretty bad for you. I tried DMSO because it's actually pretty safe, and even used as a topical pain reliever.

I contacted the chemical company I found earlier to see if they have a recommendation.

But, yes, it's possible that dissolving the coating could mean dissolving protective layers and even attacking the tiny electronics. Guess we'll see. That's why I'll practice on the damaged sensors.

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.