Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

Debayering a DSLR's Bayer matrix.


RAC

Recommended Posts

Found my scrapped 1100D sensor assemblies and applied hot air gun to the one with the broken glass. Applied until it almost caught fire! The PCB was singed and the sensor fell off the PCB BUT the glass was still firmly attached to the frame! No turning white or anything. So the 1100D sensor cover glass cannot be removed with heat :(

I also tried removing the CFA with a sharpened piece of wood - wouldn't touch it! It was softwood though - I'll see if I can find some hardwood and sharpen that.

Update :- Now it's cooled down I've been able to remove the pieces of glass :Envy: So it might just be possible to remove the glass without destroying the sensor depending on which happens first.

I would like "to put this one to bed" I don't like being defeated by an inanimate object :D It's a matter of principle.

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

Nearly bought a 1000D off ebay today until I noticed it was collection only. Missed another at just over a ton. Been looking at prices of sensors as spare part - equal to or more than a new camera!! If another 1000D comes up at around £100 I might see if I can win it - just to have a play. It would be a pretty careful play though as I wouldn't want to blow a hundred quid toooo soooon :D

Gina, What about trying with a 350D ? This may interest you:

http://www.ebay.co.uk/itm/Canon-EOS-350D-Digital-SLR-Camera-Body-Battery-1gb-cf-Free-UK-Post-/300943689216?pt=UK_CamerasPhoto_DigitalCameras_DigitalCameras_JN&hash=item4611a44600

Link to comment
Share on other sites

Gina, What about trying with a 350D ? This may interest you:

http://www.ebay.co.u...=item4611a44600

That looks interesting - thank you :) I must read back through this thread and see how this compares with the 1000D. I've looked at the general specs and the only real difference seems to be live view. The pixels of the 350D are larger so should be a bit more sensitive. I don't think live view is so very important - a repeating 10s imaging loop should do for focussing.
Link to comment
Share on other sites

£69.99 or offers - I offered £50 - rejected - then £60 - ditto - then £65 and it was accepted. So I guess I'm now the owner of a 350D body for £65 :D The fun will be starting soon :D

  • Like 2
Link to comment
Share on other sites

£69.99 or offers - I offered £50 - rejected - then £60 - ditto - then £65 and it was accepted. So I guess I'm now the owner of a 350D body for £65 :D The fun will be starting soon :D

I'll be following with interest your progress with this camera as I might try that model if I fail with my 1000D. Luis is obtaining cracking images with the 350D.

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

hi Gina

do you think scoring the glass with a glass cutter would work?

if something is attached to the centre of the glass and a rectangle is scored on the inside, we should be able to pop it upward.

I had hopes on the hot air technique for my 600d, but I prefer breaking the glass rather than trying to get it off.

you might have to do the same with the 350d

Alistair

Link to comment
Share on other sites

Thank you Alistair. Are there any free options to Maxim where I can treat the raw files?

Thanks

hi

you should be able to save as fits with canon's DPP

believe others convert to tiff.

Perhaps Luis can explajn how he stacked his eagle nebula images.

Alistair

  • Like 1
Link to comment
Share on other sites

There remains a BIG problem with a filter wheel of spacing - I can only just about get enough with CCD cameras with 13mm back focus so a Canon EOS would need a lot more stripping down than I've done as yet.

Hi

The depth of the sensor for the 350d is 45mm compared to the 55mm of most other cameras.

I use an OAG so need to fit the filter behind the OAG.

I opened the 3 screws of the bayonet to T adaptor and that reveals an M48 thread that accepts 2" screw on filters. this remains recessed in the bayonet adaptor and can potentially act as a moisture seal.

Although it'll be a pain to change filters, it should help if you don't have sufficient back focus by keeping the filter inside the bayonet adaptor as it doesn't add to the length.

Other option is to get the bayonet adaptor from Teleskop express that fits to the TS9 OAG with a T2 thread. that could potentially fit filter wheels, then the OAG and then the coma corrector if you're using one.

for me, I was using the MPCC mark 3, but looks like I will have to get the RCC1 instead as that provides over 90mm of back focus.

if I go with a motorized filter wheel, I'll have to move the mirror up.

Cheers

Alistair

Link to comment
Share on other sites

Thank you Alistair. I've been learning DCraw to get the raw files without demosaicing. That's a very powerful programme, love it, but I had to enter commands through the MS-DOS console like the old good times.

I took some photos today and I'm showing some close-ups so you can see the effect of the removal of the bayer array. My light metering hasn't been ideal, sorry. I didn't pay much attention to that. The darkest areas are squashed microlenses on top of the bayer matrix.

Thanks

post-18331-0-72535900-1375587216_thumb.j

post-18331-0-99275000-1375587229_thumb.j

post-18331-0-34401600-1375587239_thumb.j

post-18331-0-27003800-1375587258_thumb.j

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

Thank you Alistair. I've been learning DCraw to get the raw files without demosaicing. That's a very powerful programme, love it, but I had to enter commands through the MS-DOS console like the old good times.

I took some photos today and I'm showing some close-ups so you can see the effect of the removal of the bayer array. My light metering hasn't been ideal, sorry. I didn't pay much attention to that. The darkest areas are squashed microlenses on top of the bayer matrix.

Thanks

nice work. is this the 1000d?

you should be able to clean up the edges by scraping width wise under a microscope very carefully. needless to say, one wrong move and the golden wires come off.

just needs patience.

I'll take a video and show how the stubborn cfa areas come off with repeated scraping and without damaging the sensor.

cheers

Alistair

Link to comment
Share on other sites

hi Gina

do you think scoring the glass with a glass cutter would work?

if something is attached to the centre of the glass and a rectangle is scored on the inside, we should be able to pop it upward.

I had hopes on the hot air technique for my 600d, but I prefer breaking the glass rather than trying to get it off.

you might have to do the same with the 350d

Alistair

I guess it might but I think with glass cutting it works better from an edge. Mind you, if the crime progs on TV are right, it's possible to cut a round hole in a window...

The trouble with breaking the glass (as I found with one 1100D) is that the bits tend to drop on and break the fragile fine gold wires :(

Link to comment
Share on other sites

Hi

The depth of the sensor for the 350d is 45mm compared to the 55mm of most other cameras.

Great - that's a big bonus :)
I use an OAG so need to fit the filter behind the OAG.
I see. I haven't decided which scope I shall be using it on. With the MN190, I currently use an OAG. With the ED80 pair I have a choice - I'm currently using OAG plus EFW2 with Atik 460EX on one and home made FW with 314L+ on the other so the 350D might replace the 314L+. Lots to think about :D
I opened the 3 screws of the bayonet to T adaptor and that reveals an M48 thread that accepts 2" screw on filters. this remains recessed in the bayonet adaptor and can potentially act as a moisture seal.
That's interesting - must check mine but I think it's a plain surface.
Although it'll be a pain to change filters, it should help if you don't have sufficient back focus by keeping the filter inside the bayonet adaptor as it doesn't add to the length.
That's true :)
Other option is to get the bayonet adaptor from Teleskop express that fits to the TS9 OAG with a T2 thread. that could potentially fit filter wheels, then the OAG and then the coma corrector if you're using one.

for me, I was using the MPCC mark 3, but looks like I will have to get the RCC1 instead as that provides over 90mm of back focus.

if I go with a motorized filter wheel, I'll have to move the mirror up.

Cheers

Alistair

I think I shall probably build a completely new casing for the camera for cooling and sealing and include a DIY filter wheel in the design. Of course, the FW would have to be outside the sealed part so that filter changes don't mean breaking the seal. Since my 1100D cooling project, I have gained further info and had more ideas of how to do it. Edited by Gina
Link to comment
Share on other sites

nice work. is this the 1000d?

you should be able to clean up the edges by scraping width wise under a microscope very carefully. needless to say, one wrong move and the golden wires come off.

just needs patience.

I'll take a video and show how the stubborn cfa areas come off with repeated scraping and without damaging the sensor.

cheers

Alistair

Thanks Alistair, that video would be lovely as I'm now stuck trying to remove those leftovers.

Yes, that's the 1000D. What I like about this camera is that it's relatively cheap, it has live view and the glass covering the sensor comes off intact relatively easily (the 450D wasn't difficult either). However, the cfa is quite difficult to remove completely, just as with my 450D practice sensor. If don't succeed with the 1000D monochrome mod, I might try with a 350D before giving up.

I'll eventually clean up the edges if/when those leftovers go away. What I'm trying to avoid now is hitting those golden connectors and having to buy another sensor without even knowing if this mode can eventually be pulled off.

Many, many thanks for your help.

Edited by pixueto
Link to comment
Share on other sites

I think I shall probably build a completely new casing for the camera for cooling and sealing and include a DIY filter wheel in the design. Of course, the FW would have to be outside the sealed part so that filter changes don't mean breaking the seal. Since my 1100D cooling project, I have gained further info and had more ideas of how to do it.

would be great if you could share info on how to run the camera in a housing like the central ds housing or filipo's housing.can you get away without using a shutter emulator?

i think a sealed custom enclosure is the way to go.

Alistair

Link to comment
Share on other sites

I shall certainly post info and photos of any progress I make - I believe in sharing all my ventures/adventures :) I have been wondering about the shutter but don't have any info as yet. One thing I did think of was to move the shutter mechanism away from the light path and out of the way. I must check up on what (if any) feedback there is from the shutter to the electronics. I would be risking fewer squid with a 350D than I did with all the 1100Ds :D

Edited by Gina
Link to comment
Share on other sites

I thought I didn't make justice to the camera with the photos I posted before so, here I go again, this time paying attention to light metering to avoid overexposure. The photos have been taken with a 50mm 1.8 nifty-fifty lens on a tripod. Again, the images were in raw format converted to jpeg without any demosaicing applied (apart from the overexposed) using DCRaw.

The image has been reduced to 26% (because of file size) but I'm showing too a crop at 100% showing the central region. I also show the original image that the camera took with its automatic light metering. You can see it was badly overexposed. The removal of the CFA has increased dramatically the sensitivity of the sensor. The camera estimated a exposure of 1/25 and ISO 800 but I had to go to 1/40 and ISO 400 to get the composition right. Thus, I had to go to manual, lowering the exposure time and ISO significantly in order to achieve a correct histogram. All photos were taken at F8.

I can't see any scratches coming up in the photo. As I said, I also run a test for dead pixels with a number of dark frames at different exposures and ISO 1000. The test was both visually and using software which returned 0 dead pixels so the sensor seems intact despite the scratches created with the wooden tool?

Thanks for looking. Any comments welcomed.

Edited by pixueto
Link to comment
Share on other sites

I thought I didn't do justice to the camera with the photos I posted before so, here I go again, this time paying attention to light metering to avoid overexposure. The photos have been taken with a 50mm 1.8 nifty-fifty lens on a tripod. Again, the images were in raw format converted to jpeg without any demosaicing applied (apart from the overexposed) using DCRaw.

The image has been reduced to 26% (because of file size) but I'm showing too a crop at 100% showing the central region. I also show the original image that the camera took with its automatic light metering. You can see it was badly overexposed. The removal of the CFA has increased dramatically the sensitivity of the sensor. The camera estimated a exposure of 1/25 and ISO 800 but I had to go to 1/40 and ISO 400 to get the composition right. Thus, I had to go to manual, lowering the exposure time and ISO significantly in order to achieve a correct histogram. All photos were taken at F8.

I can't see any scratches coming up in the photo. As I said, I,ve also run a test for dead pixels with a number of dark frames at different exposures and ISO 1000. The test was both visually and using software which returned 0 dead pixels so the sensor seems intact despite the scratches created with the wooden tool?

Thanks for looking. Any comments welcomed.

post-18331-0-37790800-1375648530_thumb.j

post-18331-0-50130400-1375648559_thumb.j

post-18331-0-71945500-1375648584_thumb.j

post-18331-0-81533100-1375648607_thumb.j

post-18331-0-99889400-1375649557_thumb.j

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

Hi,

Those pics look good. its odd about the light metering, my 350D seemed to expose correctly, I was mainly experimenting with landscape mode that's mostly automatic and the pics seemed to turn out fine. could it be because the cfa leftover is still covering a decent portion of the sensor. anyway it shouldn't matter cause for astro images, you'd be using manual mode anyway.

did you remove the IR filter or replace it with clear glass?

with my 350D I removed the IR filter so the sensor is pretty naked!!. that affects auto focus, but its not a big deal. I should ideally get a replacement UV/IR glass, but the 350D glass is a lot more expensive than the 1000D, probably because of the thickness. I might just get some clear microscope slide glass cut to size for basic protection.

look forward to your astro testing.

cheers

Alistair

Link to comment
Share on other sites

Hi,

Those pics look good. its odd about the light metering, my 350D seemed to expose correctly, I was mainly experimenting with landscape mode that's mostly automatic and the pics seemed to turn out fine. could it be because the cfa leftover is still covering a decent portion of the sensor. anyway it shouldn't matter cause for astro images, you'd be using manual mode anyway.

did you remove the IR filter or replace it with clear glass?

with my 350D I removed the IR filter so the sensor is pretty naked!!. that affects auto focus, but its not a big deal. I should ideally get a replacement UV/IR glass, but the 350D glass is a lot more expensive than the 1000D, probably because of the thickness. I might just get some clear microscope slide glass cut to size for basic protection.

look forward to your astro testing.

cheers

Alistair

Thanks Alistair.

The sensor is completely naked as yours. I have the cover glass intact but I'm struggling to remove the CFA leftovers. Did you say you have a method to clean those bits under the microscope? This is what is holding me know. If I could remove them I would order a H-alpha clip filter so I could stat testing. Everything else seems fine... if it wasn't for the array leftovers...

Because the plan is to use the camera with a H-alpha filter, there is no need to add anything to the sensor, right? That filter should block IR light.

The filter you are referring to... is it the Baader filter? Bear in mind that the 350D came only with a filter so I would expect the Baader filter for that model to have some sort of anti-aliasing built on it -that will blur your images slightly in order to avoid moire.

How do you intend to use your camera? If you are going to use it with a H-alpha filter you wouldn't need any other glass, am I correct? If you go for a clip filter that would give you some protection for the naked sensor. Just a thought.

Thanks

Link to comment
Share on other sites

hi

with my 1000d sensor, I just repeatedly scrape over with the wooden tool and it eventually comes off. its slow and you need to keep at it but it definitely comes off.

you could try a plastic tool and it comes off quicker but I think it scratches so my preference is the wooden tool.

my worry with the naked sensor is that even with a filter in front, you could havr moisture on damp and dewy nights that could cause a short. so I would replace the IR filter with clear glass.

I think thr bloated stars due to IR are only with refractors

clip in filter should be fine

Alistair

Link to comment
Share on other sites

I guess it might but I think with glass cutting it works better from an edge. Mind you, if the crime progs on TV are right, it's possible to cut a round hole in a window...

The trouble with breaking the glass (as I found with one 1100D) is that the bits tend to drop on and break the fragile fine gold wires :(

Hi Gina

how about using a dremel with a grinding stone and grinding away the edges. then if you stick tape, the glass inside the edges should break off without too many pieces.

do you have any more practise sensors to try the heat gun to apply low but constant heat and let it cool

it could weaken the bond without burning the pcb or maybe repeated heating and cooling.

Alistair

Link to comment
Share on other sites

Hi Gina

how about using a dremel with a grinding stone and grinding away the edges. then if you stick tape, the glass inside the edges should break off without too many pieces.

That was one of the things I tried but it seems that the vibration damaged the sensor :( I don't know what part - I couldn't see anything wrong with the gold wires. I didn't even get through the glass! It did take a lot of grinding with the grinding tool I had - could have done with a narrower one. I did have some thin cutting discs but they weren't up to it and broke.
do you have any more practise sensors to try the heat gun to apply low but constant heat and let it cool

it could weaken the bond without burning the pcb or maybe repeated heating and cooling.

I don't have any spare sensors that work but I still have that sensor unit with the glass half cut through. The only other sensors I have are in working cameras. I have 3 astro modified ones - one with Peltier cooling in a new case and two more with cold fingers brought out to the back with CPU coolers (just cooling to near ambient). I had these up for sale but got no takers. I also have an unmodified one with a zoom lens I use for general photography.

So I have three 1100Ds here already that I could work on if I have a reasonable chance of not destroying the sensor. Most of the spare parts are a reasonable price except the sensor unit which costs more than a second hand refurbished body. Two could be sold and recover the cost of the original body I guess. The Peltier cooled one is not ready to sell ATM so I'd be prepared to risk that one. I was considering playing about with that a bit more sometime anyway to improve the sealing and internal humidity control.

Edited by Gina
Link to comment
Share on other sites

hi

you should be able to save as fits with canon's DPP

believe others convert to tiff.

Perhaps Luis can explajn how he stacked his eagle nebula images.

Alistair

Cheers guys,

Very cool to see this thing growing, what an excitment :)

I use only IRIS for image calibration / alignment / stacking and set background black point, from here I save as 16bit TIFF and go for Photoshop CS5, nothing more ;)

IRIS can be a bit tricky to learn, but for me it's the holly grail of image callibration.

Best wishes,

Link to comment
Share on other sites

Hello all,

Ok, this time is a "real fire" test :)

Last Saturday off I went to my dark sky spot, very windy night with good calm periods though, good transparency and seeing, started my session with the filter modded Canon 450D, this one still has the LPF filter *1 in place so there's a little bluring going on, but nevertheless I wanted to test several things, sensitivity, resolution and S/N ratio.

I made 13x5 min. subs @ 400 ISO (15º C amb. temp) and then swaped cameras leaving the T2 adaptor and Baader MPCC on the 200mm f/newt. so that it remained the same position, with the mono 350D I made 15x5 min. @ 400 ISO and let me tell you that the difference is big, it's not just a small thin it's BIG...both sensitivity AND resolution, beautifull pin point stars with the mono sensor, and with bigger pixels (6.4 vs 5.2 micron).

I made 3 images for you to compare, first the pure mono data just from the 350D, then the color data from the 450D processed exactly the same way, then a blend of the two (that's the main goal) using "L" from the 350D and "RGB" from the 450D, guess what??? It works beautifully!!!!

Here goes the Cocoon nebula in Gygnus, first the MONO 350D image:

Cocoon 15x5 min 400 ISO mono 350D_to merge

The 450D color image:

Cocoon color 13 x 5 min 400 ISO 450D_to merge

And LRGB (350D "L" and 450D "RGB"):

Cocoon color 13 x 5 min 400 ISO 450D_merged

This was a quick process just to see if I could match the images in PS (not easy) I think I have some more data "burried" there, I'll try the screen mask invert method to pull some more dust out ;)

Best wishes,

Luís

  • Like 3
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.