Debayering a DSLR's Bayer matrix.

[rinaldop]

Hi...

This is my new method (video):

http://youtu.be/wASEdgtkpb8

And this is my first image, after 5 minutes of work:

https://dl.dropboxusercontent.com/u/34091483/IMG_4289.JPG

And these are my new tools (dental silicone polishing tips):

https://dl.dropboxusercontent.com/u/34091483/IMG_5180.jpg

Rinaldo

[rinaldop]

Hi...

This is my new method (video):

http://youtu.be/wASEdgtkpb8

And this is my first image, after 5 minutes of work:

https://dl.dropboxusercontent.com/u/34091483/IMG_4289.JPG

And these are my new tools (dental silicone polishing tips):

https://dl.dropboxusercontent.com/u/34091483/IMG_5180.jpg

Rinaldo

In the first image (IMG_4289.JPG) there are a lot of dust. Sorry.

Rinaldo

[russellhq]

Hi...

This is my new method (video):

http://youtu.be/wASEdgtkpb8

And this is my first image, after 5 minutes of work:

https://dl.dropboxusercontent.com/u/34091483/IMG_4289.JPG

And these are my new tools (dental silicone polishing tips):

https://dl.dropboxusercontent.com/u/34091483/IMG_5180.jpg

Rinaldo

I like this a lot. I was concerned you were going to scratch the sensor but it looks good so far.

I would suggest though instead of moving the head top to bottom (towards the wires). You move it side to side, parallel with the wires. You should have less chance of hitting them doing it that way.

Where did you get the drill?

[rinaldop]

Well, after a dead sensor, I would show pictures taken with my Canon 450D after correction at the first sensor (the removal of the Bayer filter layered on top and bottom).

https://dl.dropboxusercontent.com/u/34091483/IMG_4362-RAWcomposite.jpg

Well, I can say that the end result left me satisfied, despite some minor dirt and traces of Bayer filter on top and margin. Now, the next step is to add a Peltier cooling system.

Rinaldo

[rinaldop]

Another image, Some dusts on the sensor in F22 (ISO 100).

https://dl.dropboxusercontent.com/u/34091483/IMG_4366-RAWcomposite.jpg

Rinaldo

[rinaldop]

Finally, my last image of this night.

https://dl.dropboxusercontent.com/u/34091483/IMG_4392-RAWcomposite.jpg

Rinaldo

[pixueto]

That's very good Rinaldo, thank you.

I have lots of questions if you don't mind:

-the ceramic dental tools, how to you attach them to the Black and Decker?

-the ceramic tools? Aren't they too hard? Did they scratch the sensor?

-Did you end up with any scratched area? I can't see any in your flat frame

-What F is your flat frame?

Many thanks. I am already inspired by this. Maybe it's time to go back to the epoxy resin to protect the sensor Gina? However, I'm still concerned about the amount of damaged sensors when the resin was applied. Maybe some black mask tape over the edges to nicely cover it would do the trick?

Edited February 22, 2014 by pixueto

[rinaldop]

That's very good Rinaldo, thank you.

I have lots of questions if you don't mind:

-the ceramic dental tools, how to you attach them to the Black and Decker?

-the ceramic tools? Aren't they too hard? Did they scratch the sensor?

-Did you end up with any scratched area? I can't see any in your flat frame

-What F is your flat frame?

Many thanks. I am already inspired by this. Maybe it's time to go back to the epoxy resin to protect the sensor Gina? However, I'm still concerned about the amount of damaged sensors when the resin was applied. Maybe some black mask tape over the edges to nicely cover it would do the trick?

Pixueto,

Thank you. Sadly, I lost my second sensor because I broke only one gold filament. However, I have a good sensor (the first one) that is perfect to me. I used the silicone polish tip for remove the Bayer filter in the edges.

The dental tool is not ceramic. It is a hard silicone tool and connects perfectly to Black & Decker machine (by mandrel).  A previous inspection of the surface reveals apparent small scratches that don't affect the image quality (as seen in the flat obtained in F/22).

Rinaldo

[pixueto]

Thank you Rinaldo. Sorry to hear the second sensor is lost.

I had ordered the same Black&Decker  you are using and it arrived today. I'm still waiting for the cotton and felt tips to arrive from China -that may take a while, though.

I take that the scratches were caused by the silicone dental tool and they only happened in the edges? Is the cotton and felt tool smooth enough to avoid scratching the sensor in the main central area? Inspecting more carefully your photos, there seems to be a few little scratches only in the edges -I presume this is where you used the dentist tool? Have you though about using the tool with a bit of fluid to lubricate the process?

If that is the case, we may be better off using the cotton and felt tip with the machine and either:

-go back to the epoxy resin (but I think this has already caused some damaged sensors)

-covering the untreated edges with a thin line of black masking tape over the glass covering the sensor (a similar tape that the one is used on the hot mirror. This way the sensor would be cropped just a bit, covering the colour array leftover.

What do you think? I think it's doable now with a perfect finish? Well I may be talking too soon; I still have to try your system Rinaldo.

Edited February 22, 2014 by pixueto

[rinaldop]

Pixueto,

I think felt tips (in rotary tools) should be used for the Bayer filter removal in Cmos sensors. Use the silicone tips (conical) only for the edges. I disagree with the use of liquids (lubricants) in the polishing process.

Rinaldo

[rinaldop]

Hello,

A new test image:

http://www.pbase.com/rinaldopsantos/image/154613182/large

Rinaldo

[russellhq]

For those using Isopropanol to clean up the sensor. Do you just use it on a sensor swab or do you bath the sensor in to?

[Tf2k]

Hello Rinaldo, 

nice to have another method to remove the CFA.

Can you tell us at which speed level you are using the black'n decker? at a rather high, mid or low speed? 

[melsky]

What are peoples thoughts on this camera? Its a monochrome Canon 15mp sensor that could be a dream come true, JTW Astronomy are hoping they can offer a cooled version.

http://stargazerslounge.com/topic/212124-canon-announces-m15p-cl-industrial-camera-with-monochrome-sensor/

Mel

[Psychobilly]

I wish you hadn't mentioned JTW Astronomy ... it just wandered over to their website - it  could be an expensive visit

[pixueto]

Dear all,

Finally, I have found some time to try Rinaldo's method. I, as Rinaldo did, went for a dry polish with the rotary tool. The results were very encouraging and I managed to remove all the CFA matrix. At first inspection it all seemed very clean without any scratches. However, a closer look under the microscope does show some scratches. Unfortunately, as the work progressed, I got too brave and confident and managed to break some gold connectors so I can't check how much of an issue these scratches are in the final images. Overall, a far better result than what I got with my working 1000D debayered sensor and this method shows promise in my opinion although there are some issues which need to be sorted.

These are the tools that I've used:

http://www.ebay.co.uk/itm/261211368960?var=560197392052&ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

http://www.ebay.co.uk/itm/191083563134?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

I am waiting now for my next sensor to arrive. I will now try applying an epoxy resin compatible with electronic circuits (chip on board, as one of the forum contributors suggested) to see if I can effectively protect these connections from a robust debayering process. This is the resin I have ordered from Maplin:

http://www.ebay.co.uk/itm/400609675218?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

Rinaldo, did you manage to get a complete clean debayered sensor? I also noticed that some parts are darker than others so I assume that the surface isn't evenly polished in my sensor. That flat F22 that you showed us looked pretty well, though.

I would also like to ask you what do you think about using the rotary tool with fine glass polisher or even with a mini silicone carbide grinding wheel to make the surface even? I'm thinking about something like this:

http://www.ebay.co.uk/itm/15-PIECE-SILICON-CARBIDE-GRINDING-WHEELS-DISCS-USE-WITH-ROTARY-TOOL-INC-DREMEL-/161151019601?pt=UK_Home_Garden_PowerTools_SM&hash=item2585596a51

Or fine glass polishing with this:

http://www.ebay.co.uk/itm/80-PIECE-13mm-SMALL-FELT-POLISHING-15-25-/161064344528?_trksid=p2047675.l4066

Thank you

Edited April 15, 2014 by pixueto

[pixueto]

Pixueto,

Thank you. Sadly, I lost my second sensor because I broke only one gold filament. However, I have a good sensor (the first one) that is perfect to me. I used the silicone polish tip for remove the Bayer filter in the edges.

The dental tool is not ceramic. It is a hard silicone tool and connects perfectly to Black & Decker machine (by mandrel).  A previous inspection of the surface reveals apparent small scratches that don't affect the image quality (as seen in the flat obtained in F/22).

Rinaldo

That's interesting Rinaldo. As I said, when I inspected my debayered sensor under the microscope, I did see some small scratches. Those scratches were invisible at simple sight when the surface looks perfectly smooth. After what you are saying, chances are that they wouldn't be visible in a flat frame at F22? That would be brilliant. I shall be more careful with my next sensor and I will install it back before finishing the edges to check this possibility.

By the way, if the epoxy resin compatible with circuit boards does the trick, I plan to apply it all the way though the edges where the connectors are and I have no intention of placing the glass back over the sensor so I can increase the sensibility even further. Anyone sees any problem with that? I have been using my debayered canon 1000D without any protection on the sensor (as this is only a prototype) and it has been fine so far.

Edited April 15, 2014 by pixueto

[Gina]

I too will be interested if this epoxy resin works - I tried ordinary slow cure epoxy and that didn't.  I had one debayered sensor that was fine at 20C but failed when cooled.  Whatever is used it needs to be good down to at least -20C.  Sometime I'm planning to have another go with an 1100D sensor as this is a better model than the 450D I have working ATM.

The 450D isn't yet in use for astro imaging as I haven't finished the enclosure.  I'm waiting on a replacement control board for my 3D printer as I shall be printing plastic parts to my own design.  In order to reduce the optical path to accommodate a filter wheel and OAG I'm replacing the main plastic frame that connects sensor and shutter to the front of the camera.

[moxican]

Guys I have been following this thread for a while now. I think it is a great idea that some of you here challenges the elements I have also seen here that it is a little difficult to get your hands on another sensor if you screw the first one up (and it would seem that it is almost certain one would), therefore I decided not to do this on my 450D sensor. But now I have a 450D sensor which is redundant. I don't want to make any sale advertisement here as I am very sure that the moderators would not be pleased with it but if any one interested just send me a private massage. It'd be nice if it does not go to waste.

[pixueto]

I too will be interested if this epoxy resin works - I tried ordinary slow cure epoxy and that didn't.  I had one debayered sensor that was fine at 20C but failed when cooled.  Whatever is used it needs to be good down to at least -20C.  Sometime I'm planning to have another go with an 1100D sensor as this is a better model than the 450D I have working ATM.

The 450D isn't yet in use for astro imaging as I haven't finished the enclosure.  I'm waiting on a replacement control board for my 3D printer as I shall be printing plastic parts to my own design.  In order to reduce the optical path to accommodate a filter wheel and OAG I'm replacing the main plastic frame that connects sensor and shutter to the front of the camera.

Hi Gina,

I'll let you know how that goes. Thank you for the offer Moxican. I bought the last two canon 450D sensors for £15 and £20 so it would be actually more expensive to ship the sensor from Canada that anything else! Nevertheless, I didn't even check whether these sensors work before the debayering process but I'm too lazy to disassemble my working canon 450D to try it!

I'd like your input guys on my suggestion about not placing back the sensor cover glass after the debayering process. I am after avoiding reflection and increasing the sensitivity of the optical train by removing as much glass as I possibly can. The idea is to protect those connections from moisture by using that epoxy resin. Here you have a video where they are doing just that in an integrated circuit:

Let me know what you think. Gosh! this thread has gone cold, hasn't it?

Edited April 16, 2014 by pixueto

[Gina]

Thank you

I may warm this thread up again before long - I've just bought a duff 1100D to play with from ebay auction

[pixueto]

Well, I applied the epoxy yesterday night and today it's rock solid. I could now step on top of it and the gold connectors would still be protected. I have placed the sensor back and it's working fine. That resin seems to be the solution as it doesn't seem to affect the sensor -at least in the short term and at room temperature.

Gina, you wanted to test it at low temperatures. Any suggestions?

I'm planning to give it the dremel treatment tonight once the kids aren't around. Fingers crossed!

[libraryman]

Where does one obtain this type of resin?

Ray

[Gina]

Well, I applied the epoxy yesterday night and today it's rock solid. I could now step on top of it and the gold connectors would still be protected. I have placed the sensor back and it's working fine. That resin seems to be the solution as it doesn't seem to affect the sensor -at least in the short term and at room temperature.

Gina, you wanted to test it at low temperatures. Any suggestions?

I'm planning to give it the dremel treatment tonight once the kids aren't around. Fingers crossed!

I already had a camera modified with a cold finger and Peltier TEC cooling but failing that I guess you could put it in the freezer.  Probably the whole camera with the USB cable coming out past the seal.

[pixueto]

I already had a camera modified with a cold finger and Peltier TEC cooling but failing that I guess you could put it in the freezer.  Probably the whole camera with the USB cable coming out past the seal.

Where does one obtain this type of resin?

Ray

I got it from maplin online:

http://www.ebay.co.uk/itm/400609675218?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649

Only problem, though is that you have to mix the resin and the hardener in the boxes provided; you end up using 10% of it and discarding the rest! Anyway, it's about £8 the whole thing.