Debayering a DSLR's Bayer matrix.

[Zakalwe]

You lot are nuts!

This is a fascinating thread (though you collectively must have spent more than a couple of CCDs?)...keep it up. it's great seeing what a determined modder is capable of!

[Gina]

I was wrong - it's not light level. I have just been taking darks with the lens cap on. Firstly 8s at ISO 200 then 30s at ISO 1600. It's warm today and must be around 20-25C in the obsy. I'll do some temperature tests later (fridge and freezer). Or maybe I ought to get the proper cooling working

These are JPEGs just resized with no other processing other than what the camera does.

Edited August 19, 2013 by Gina

[fwm891]

....

I don't think so unfortunately. The microlens layer has a gel-like consistancy under the microscope and comes off very easily, but doesn't ever seem to come off it large chunks. The filter layer I would describe as almost crumbly - it doesn't seem like the coloured pixels in the layer itself are held together, only stuck to the layer beneath.

That said the idea of scraping it off with something very thin by getting underneath might work. But the trouble would be getting something to a sharp enough edge without it being made of metal or something that would scratch the glass. I'll happily experiment with my broken sensor.

...

OK I've only got to page 15 ATM but a wooden lolly stick cut from one side with a blade will give a 'soft' chisel edge - do not sand as abrasive grains will embed themselves in the wood.

If this has been covered later by someone - apologies

Francis

[JamesF]

OK I've only got to page 15 ATM but a wooden lolly stick cut from one side with a blade will give a 'soft' chisel edge - do not sand as abrasive grains will embed themselves in the wood.

The stuff about heating the glass to remove it and using epoxy to protect the connecting wires is definitely something you should read, too. And there's a post from "Rottweiler" explaining how they have done it on some sensors.

James

[Gina]

I've been trying the latest version of APT. It won't work properly with this extremely modded 1100D It opens and closes the shutter (you can hear it working) but it won't save any images and just says busy after the exposure. Only powering down the camera will clear this. Whether it's due to the sensor or something else that's happened with the camera, I don't know. All I do know is that this camera was working with APT v1.94 a while back in a different box with full set point cooling or cooling off. But it's been pulled about considerably since then and I could have upset something.

EOS Utility takes and saves images perfectly.

Edited August 19, 2013 by Gina

[flolic]

Failure is not always a total failure

[Gina]

OK I've only got to page 15 ATM but a wooden lolly stick cut from one side with a blade will give a 'soft' chisel edge - do not sand as abrasive grains will embed themselves in the wood.

If this has been covered later by someone - apologies

Francis

I am writing up my progress on my particular 1100D camera and sensor in a separate thread but I'm not up-to-date yet. Just as far as removing the cover glass so far.

http://stargazerslounge.com/topic/192357-debayered-and-cooled-1100d/

Edited August 19, 2013 by Gina

[Luis Campos]

Hi all,

Ok, one more sensor (350D) and allmost there this time, I made a very small scratch on the surface but it was quite deep and reached the silicone damaging some photodiodes, there are two dead pixels collumns but the sensor works ok for "normal" photography, the worst part is that in long exposure there is terrible noise and the sensor is totally useless, it's a sea of noise and all the detail in the images is gone, i guess the scratch damaged more then just the two collumns, maybe the sensor overheats when set to long exposure causing the dredfull noise....so no go for astro use I'm afraid...still have one sensor to go, waiting for it now...:/

Here's an image i took from my kitty, just check out the detail on the cat's eye fear and nose, it's mindblowing!:

Cheers,

Luís

[flolic]

What I've learned so far:

-cover glass comes down quite easily with the help of hot air (350 to 390C), carefully applied around the edges through small nozzle (3mm).

-golden wires and blue border of the sensor are very fragile, slow setting epoxy applied to those parts protect them excellent.

-only scraper hard enough to attack CFA, and soft enough not to scratch too deep is wooden one. Any metal is (imho) too risky.

-when you have your sensor's glass removed, epoxy applied and hardened, and CFA scraped off with wooden tool, any remains of CFA

can be easily cleaned by polishing. Because all sensitive parts of sensor is protected with epoxy, just put a small drop of polish paste to the

sensor surface and very gently polish it with soft cloth and your finger. I've used Xerapol paste for polishing polycarbonate plastics, but I am sure

there are better products with finer and smaller abrasive particles. Polishing takes only minute or so, you have to remove only CFA remains and

not underlying layers! So, wipe few times with polish and then clean and inspect surface. Every progress will be quite obvious, because you can see

the difference if you look at the surface against the light under various angles. My advice is to polish, clean and then install the sensor into camera

and take few test shoots to see progress. And then repeat that until chip is clean.

[alistairsam]

Hi,

I'm going to get a 1000D sensor and have a go in a few days.

with Gina's sensor issues, could it be the heat that's damaged a few SMD's on the underside? I recall you mentioned the cold finger dissipated most of the heat, but I still wonder.

although flolic's success with the 450D does question this theory.

great work guys, keep at it.

Cheers

Alistair

[Gina]

Thank you Filip Good summary

Have you tried cooling yet?

[pixueto]

What I've learned so far:

-cover glass comes down quite easily with the help of hot air (350 to 390C), carefully applied around the edges through small nozzle (3mm).

-golden wires and blue border of the sensor are very fragile, slow setting epoxy applied to those parts protect them excellent.

-only scraper hard enough to attack CFA, and soft enough not to scratch too deep is wooden one. Any metal is (imho) too risky.

-when you have your sensor's glass removed, epoxy applied and hardened, and CFA scraped off with wooden tool, any remains of CFA

can be easily cleaned by polishing. Because all sensitive parts of sensor is protected with epoxy, just put a small drop of polish paste to the

sensor surface and very gently polish it with soft cloth and your finger. I've used Xerapol paste for polishing polycarbonate plastics, but I am sure

there are better products with finer and smaller abrasive particles. Polishing takes only minute or so, you have to remove only CFA remains and

not underlying layers! So, wipe few times with polish and then clean and inspect surface. Every progress will be quite obvious, because you can see

the difference if you look at the surface against the light under various angles. My advice is to polish, clean and then install the sensor into camera

and take few test shoots to see progress. And then repeat that until chip is clean.

Filip, Can you show us a flat frame say at F8 with your 450D? Do you think car glass polish will do the job?

Thanks

[Gina]

I now have the test camera in the fridge with power and USB cables brought out past thr door seal. It hasn't had much time to cool down yet but a time exposure of 10s at ISO 800 is fine but with 20s I get the exposure then err70. 15s OK... 19s OK... strange. Camera power now off and letting camera cool down.

Edited August 19, 2013 by Gina

[russellhq]

Filip, Can you show us a flat frame say at F8 with your 450D? Do you think car glass polish will do the job?

Thanks

I think if you're going to use a polish, you will want to use one that breaks down as you apply it so you don't polish away the silicon! Something used for car paint for example (not T-Cut as it doesn't break down) may work.

[JamesF]

I now have the test camera in the fridge with power and USB cables brought out past thr door seal. It hasn't had much time to cool down yet but a time exposure of 10s at ISO 800 is fine but with 20s I get the exposure then err70. 15s OK... 19s OK... strange. Camera power now off and letting camera cool down.

That's quite odd. If you set the exposure for 30s does it give the error after 30s, or terminate the exposure at 20s and throw the error?

James

[Gina]

That's quite odd. If you set the exposure for 30s does it give the error after 30s, or terminate the exposure at 20s and throw the error?

James

The 30s exposure completes (can hear the shutter) and then I get err70. Now cooled down and I'm getting err70 immediately on starting time exposure.

I now have one of my ambient cooled 1100Ds (Rebel T3) stripped down to sensor cover glass ready to apply heat. Before that I tested it in both EOS Utility and APT latest version and all was working fine - took 10 x 10s subs in APT. Sensor was running at 25C.

[pixueto]

I think if you're going to use a polish, you will want to use one that breaks down as you apply it so you don't polish away the silicon! Something used for car paint for example (not T-Cut as it doesn't break down) may work.

It's for removing the CFA leftovers while removing some scratches at the same time.

[JamesF]

The 30s exposure completes (can hear the shutter) and then I get err70. Now cooled down and I'm getting err70 immediately on starting time exposure.

That seems a very odd failure mode. If the exposure completes then presumably it's not a problem with the exposure per se. But if shorter exposures work ok then presumably there's no issue with getting data off the sensor either. Might be interesting to try at different ISO settings too? Perhaps there's some different execution path in the firmware for exposures greater than some limit at some ISO setting and the longer one finds a problem that the shorter one doesn't.

James

[Gina]

That seems a very odd failure mode. If the exposure completes then presumably it's not a problem with the exposure per se. But if shorter exposures work ok then presumably there's no issue with getting data off the sensor either. Might be interesting to try at different ISO settings too? Perhaps there's some different execution path in the firmware for exposures greater than some limit at some ISO setting and the longer one finds a problem that the shorter one doesn't.

James

I can do some more experiments but it seems the bottom line is that this sensor is duff as far as astro is concerned. Now I hsave the other camera stripped down I could put the debayered sensor in that and see what happens. That camera is unmodified apart from the colour balance filter and the LCD display being removed.

[JohnC64]

Following this thread with interest - although its !!!!!WAY!!!!! out of my comfort zone.:-) After some googling around, came across this company doing professional conversion.

http://www.maxmax.com/b&w_conversion.htm

Looking at their prices though, I suspect that the "buy a duff body route" off ebay would be cheaper even if you had a few failed attempts. But they do show some interesting images showing how good the camera is after cfa removal.

Good luck everyone doing the conversion

John

[Gina]

Well... I've put the debayered sensor in the other camera and APT is working with it. Next test is to put it in the fridge and try cooling.

Sorry about the "red herring"

[Gina]

I have APT running with the camera in the fridge and netbook on the kitchen worktop. Not very cold yet - the EXIF T is showing 21C. Of course, the sensor and other components are producing heat which isn't being dissipated very efficiently. Currently running 30s at ISO 800. I've got 5s pause between subs. I think I'll increase this to let it cool down more. I might look at using Peltier TEC cooling tomorrow.

[JamesF]

So it looks like some other part of the camera that may be at fault?

James

[russellhq]

If you can get it in a freezer you'll find it cools down far quicker! Well that's what I found anyways

[Gina]

I've had to stop - the camera is dripping with condensation!