Debayering a DSLR's Bayer matrix.

[spokeshave]

I saw the same thing. I used the DCRAW "document" mode for decoding, which does no bebayering at all. I've been scratching my head about what is the best thing to do with a monochrome image encoded in a RAW file designed for color images. I just tested the sensor late last night and have had no time for experimenting with the file. I gave the TIFF a very careful examination, but that was just to look for scratches and dead pixels. I found no evidence of either.

I'll post a link to the RAW file when I get home.

Tim

[Xplode]

Great. 

Could you also take a flat and post it as RAW?

[spokeshave]

Here is a link to the CR2 file:

https://drive.google.com/file/d/0B1MOziksBQm2XzRubGxaMVAtR3c/view?usp=sharing

Let me know if it doesn't work. I *think* I have the permissions set properly.

I think I have the correct dcraw switches:

dcraw -v -d -r 1 1 1 1 -4 -T path_and_filename.extension

I'm using dcraw64 and this worked very well. No bayer matrix visible, completely monochrome and mapped into 16-bit space.

I might have a flat from my experiments yesterday. I'll take a look and post it if I have it. Otherwise, I'll take a flat tomorrow.

Tim

[spokeshave]

Here's a link to a flat RAW file:

https://drive.google.com/file/d/0B1MOziksBQm2OTQ5d0NsOU83Rjg/view?usp=sharing

I stretched the flat heavily after decoding with the dcraw string above. Here's a jpeg. There are a few artifacts near the edges and corners indicating that I could do more polishing, but they won't affect imaging if I get decent flats. Still, I will continue to experiment to strive for perfection. I'm very encouraged by the results so far.

Tim

[Xplode]

Thanks for sharing the RAW files.

They open perfectly in Pixinsight with the option "create RAW bayer CFA image"  (dcraw -4 -o 0 -D)

[flolic]

I ordered the hydrogen peroxide (it is conveniently available in 35% strength as a nutritional supplement) and picked up some sulfuric acid from the auto parts store (battery acid).

I ordered the hydrogen peroxide (it is conveniently available in 35% strength as a nutritional supplement) and picked up some sulfuric acid from the auto parts store (battery acid).

Regular battery acid will not work, you need concentrated stuff. You can concentrate sulphuric acid by boiling it in a glass beaker, but you must be VERY careful and do that outdoors or under the fume hood. Concentrated sulphuric acid is very nasty stuff, be extremely careful handling it. Also remember to NEVER add water (or hydrogen peroxide) to the acid, always (slowly) add acid to the water. Otherwise you risk thermal explosion because reaction is exothermic.

[mellowyellow]

So I was trying this on my 550d. I put the sensor back in to see how it was looking but my images are only coming out black, even when Im using live view. Anyone have any idea? Ive taken the camera apart down to the sensor and put it back together several times looking for a connection I missed but I cant find anything. Thanks in advance.

[Herra Kuulapaa]

So I was trying this on my 550d. I put the sensor back in to see how it was looking but my images are only coming out black, even when Im using live view. Anyone have any idea? Ive taken the camera apart down to the sensor and put it back together several times looking for a connection I missed but I cant find anything. Thanks in advance.

Your ticket didn't win and you have a questionable honor to join the club of DSLR killers.

It's a dead sensor

[Gina]

Welcome aboard   I'm afraid Herra is right   Maybe caught one of those tiny fragile gold wires.

[mellowyellow]

Lol thanks guys. I just figured that out. I got a spare sensor in the mail and swapped it out. Im going to hopefully do some imaging tonight and rip it back apart and try the new sensor tomorrow. I was pretty aggressive trying to get the glass layer off. Turns out it only needed heat lol. This is a great thread though btw. And gina, Ive gone through a few of your threads as well. Good stuff! And thanks for the welcome!!

[Herra Kuulapaa]

Unmodified DSLR is practically blind to near IR (>950nm).

Here's a handheld picture pair showing how sensitive the modified sensor is (no IR filter, color picture is there just for fun).

It's a good idea to filter IR out...

[Dave_D]

itching to have a go at debayering but there seems to be a complete lack of 1100d sensors available and the only place i found that *could* supply them wanted more than the cost of the camera when it was new

stupid for a discontinued model methinks.

[spokeshave]

I can't speak for the 1100D, but I found that looking for broken bodies is much more fruitful than trying to find the actual sensors. I recently bought a set of 5 broken 55D cameras off of eBay for less than $200. All had intact, working sensors. In fact, I was able to reconstruct 3 working cameras from the lot.

Tim

[ultraviolent]

Apropos, does anyone have a 1000d body left over? My sensor didn’t die, but now I can only swap between the original and the debayered sensor for one body. Despite the sensor all parts should be in a working conditio.
 
I made a picture of the moon but the resolution of my MTO-11CA (f=1000mm Ø=100mm) is not high enough to see a positive effect in removing the Bayer layer: 
 full size

[mellowyellow]

Pretty happy with the outcome on my t2i. Only killed one sensor. This one I scratched with the razor blade taking the top glass off but its cropped out

[Herra Kuulapaa]

It's most likely a some sort of reflection, but have you read about the cases of eye cancer detected by photos?

[mellowyellow]

She blind in both eyes. She was bit by a big dog when she was a few weeks old and it took out her eyes. The vet put them back in but thanks for looking out!

[bitnick]

Here is someone in the US selling what they claim is 1100D sensors for $75 (about £50 before shipping & VAT).

Also, I've "only" read the first 50 pages or so of this thread, so maybe this has been said already, but I'm wondering whether some of the failures of Gina's sensors might be due to ESD. The work area did not look ESD safe to me (maybe I'm wrong though, difficult to see), and putting unprotected semiconductors into "generic" (non-conductive/non-ESD-safe) plastic containers is usually a very bad idea.

Here is a link to a paper by ON Semiconductor on handling image sensor devices. It starts out

ESD events can cause immediate damage to a device so that it is no longer functional. The effect may also not be noticed until a considerable time has passed, with the unit operating to specifications for some time. ESD events also show up as shifts in device characteristics.

ESD events occur by improper handling of the image sensor. Improper handling includes any operation that creates an electrostatic discharge; for example, handling the device without a wrist strap.

ESD is also a danger to the rest of the camera, of course (one of Gina's camera houses stopped working right for no apparent reason, if I remember correctly). I suggest using a properly grounded ESD mat when disassembling electronics like cameras - something like the 3M 68xx maybe? (Avoid the vinyl mats though!)

There are of course a thousand things that could go wrong with this mod, but the more a reason to eliminate as many of those things as possible!

Anyway, very cool of you to attempt this - kudos!

[Gina]

Thank you - that's a very good point, though I did take some ESD precautions it could be that I wasn't careful enough.  Good point about plain plastic containers - I was careful that no sensitive parts touched the container, but a good point none-the-less.  I was interested to see that the damage could take some time to show up.  Of course, most of my failures were due to very obvious causes but I admit that ESD could have caused one or two.

[Herra Kuulapaa]

This case is a Nikon D7000 sensor with a glass shard unconfortably close to the bonding wires.

Shard was fixed with epoxy and sensor works fine, but a close call nevertheless. See the microscopic wiring on the blue area? No scratching those if you wish to use sensor to imaging anymore..

[Gina]

WOW!!     Glad you kept it working

[Herra Kuulapaa]

I've noticed that they're using two different adhesives in Nikon D5100/D7000 models. Sourcing of the production materials I guess.

One is tight and the second, it's insanely tight. There's a heat release difference between them. That picture is the second type of adhesive. It tends to shatter glass also horizontally instead of just vertical fractures of "regular" tight glue.

[Gina]

Seems these sensor manufacturers are ganging up on us poor experimenters

[chimera15]

This is a huge thread.  I just found it, it seems like the only one I can find on the internet that might have a solution.  I saw someone earlier in the thread had damaged their sensor wires.  I had a sensor that I defiltered/deglassed and later cleaning it I destroyed one of the wires.  I can solder to the frame edge contacts, but the sensor itself refuses to adhere any solder.  Anyone have a fix for this...maybe I have to grind down to get to the actual pads or something...but they seem to be electrically conductive, just wont take any solder.  I might try to use electrically conductive tape or something to fix them, thats the only thing I can think might work.   Anyway, any tips?

[astromerlin]

This is a huge thread.  I just found it, it seems like the only one I can find on the internet that might have a solution.  I saw someone earlier in the thread had damaged their sensor wires.  I had a sensor that I defiltered/deglassed and later cleaning it I destroyed one of the wires.  I can solder to the frame edge contacts, but the sensor itself refuses to adhere any solder.  Anyone have a fix for this...maybe I have to grind down to get to the actual pads or something...but they seem to be electrically conductive, just wont take any solder.  I might try to use electrically conductive tape or something to fix them, thats the only thing I can think might work.   Anyway, any tips?

It's a long shot but you could try using electrically conductive paint/glue used to repair heater matrix on car rear windows.