Debayering a DSLR's Bayer matrix.

[Stub Mandrel]

2 hours ago, zfedoran said:

I imagine it would work since it is older. If you do try it out, the trick is to get started with a small scrape, just enough to lift a few CFA pixels. Then apply the tape, pressing firmly, then peal back slowly. You'll need to take a lot of care to ensure the static cling of the tape doesn't make it jump somewhere you didn't intend (like the gold wires).

Take a look here: https://www.cloudynights.com/topic/486810-thinking-about-removing-the-cfa-from-my-20d-thoughts/

 

Just a side note though, I've done this on two canon 20d's and honestly, it isn't worth the effort. The camera is quite dated and really annoying to work with software wise. Other than that, it also suffers from an outdated ADC process that introduces noise.

True, I haven't used the 10D since getting a 450D; being 12 rather than 14 bits probably loses more signal than debayering gains.

Still it might be fun to try.

[zfedoran]

1 hour ago, Stub Mandrel said:

True, I haven't used the 10D since getting a 450D; being 12 rather than 14 bits probably loses more signal than debayering gains.

Still it might be fun to try.

If you have the dexterity to successfully debayer a sensor, then maybe take a look at building the Cam86 (which has a 16 bit ADC with a real CCD sensor). You don't even need to sacrifice a camera to get the sensor, you can buy old new stock (about $10-20 per sensor).

 

 

[Stub Mandrel]

21 minutes ago, zfedoran said:

If you have the dexterity to successfully debayer a sensor, then maybe take a look at building the Cam86 (which has a 16 bit ADC with a real CCD sensor). You don't even need to sacrifice a camera to get the sensor, you can buy old new stock (about $10-20 per sensor).

 

 

Too many projects, too little time, at least for the present...

[benzomobile]

On 01/11/2019 at 14:50, alterelektroniker said:

You are right about that! Don't do that. 

This said, loooong time ago I work in an electronics factory and we did wire bonding Chip on Board modules, or COB as we call them. Bond wires where sealed with an special epoxy that had no thermal movement. Brand was Hysol then, now owned by Henkel, here is the link to what they have for that application now:

https://www.henkel-adhesives.com/us/en/products/potting-encapsulating-injection-molding-compounds/glob-top-material.html

And be careful, this not removable.

Alternately, google for Glob Top Adhesive-Sealants.

I suggest Uhu plus endfest 300  (not the faster compound), because it is glass solid and very stable over time. I've always used that product with with no problem at all.

Beppe

[Noobulosity]

Well.......  I guess I just couldn't leave well enough alone.  I picked up another sensor from eBay (my 4th one...) and went at it again.  This time, I'm really pleased with the results.

My method:

• After dismantling the sensor a bit, I was able to remove the cover glass relatively easily using a torch-style lighter to break the epoxy and a hobby knife to very gently lift the glass.  (Add a touch of heat to help it release if it doesn't want to lift.)
• Once the cover glass is removed, I reapply the plastic frame that held the UV filter (I removed it).  This acts as a guide for my scraper and helps me avoid hitting the tiny gold wires.
• Use a sharpened oak dowel to begin the removal process.
• Once I start seeing decent wear, I added some Maas Metal Polish to the tip of the dowel.  This helps to wear through the Bayer matrix to the bare photosites.  DON'T OVERDO THIS!!
• Once I start opening up areas through the filter layer, I switched to a cotton swab/Q-tip with more of the Maas Metal Polish.  Gently rubbing and swirling to help erode and remove the epoxy Bayer matrix.  This worked REALLY well!  I was able to get a nice, uniform removal over the majority of the sensor, only leaving behind the border areas to avoid hitting sensitive electronics.
• As-needed during the process, and to check progress, I used 91% isopropyl alcohol to clean off the metal polish, then a dry swab to clean up most of the alcohol and swarf.

Once I reassembled the camera, it worked like a charm!  It takes nice, clear, sharp images.  I can't wait to test it using my telescope.  Since it's a really bright moon right now, I may test out some H-alpha narrowband imaging and see what I get.

Flat:

Converted flat:

Converted daytime image:

Highly-stretched to inspect for defects:

[michael8554]

Congrats, you seem to have found the winning formula.

What does a Torch Lighter look like? I've googled but still not clear which one you used?

Michael

 

[Gina]

Well done - seems you've cracked it! 

[Noobulosity]

On 12/03/2020 at 17:12, michael8554 said:

Congrats, you seem to have found the winning formula.

What does a Torch Lighter look like? I've googled but still not clear which one you used?

Michael

 

These lighters have more of a jet flame than a lazy candle flame. Like a small propane torch. It's just a style, not a brand.

[dunners19871]

Hi

I recently tried mono modding my 700D... that sensor is dead now. I think i damaged the blue data bus getting the glass off, which seems to be the main issue with that camera and the 600D. I then got a 450D as the glass looked like it pretty much drops off those, which it did, happy days.

I removed the CFA using glass polish and a cotton swab, wrapped some cotton around a toothpick to get closer to the edge, but wanted to keep a good distance from the bus and gold pins.

i have attached my result and have  a few questions maybe someone can help with

1 is it meant to be red? i have seen other pics like this but as there is not matrix shouldn't it be grey? is there still a layer to polish off? 

2. there seems to be a cross of dead pixels that has appeared, any ideas on what could have caused this? 

3. the image is combined with flats, which i thought would remove the cross and the dust, but both are still there. any ideas on why that might be

Thanks in advance, and for all the previous posts, its a really interesting topic :) 

cheers 

matt 

[michael8554]

Why still Red ? Well the sensor is now mono, but the processing is still RGB, and is applying an appropriate White Balance.

A Custom White Balance should correct the caste.

Michael

[dunners19871]

8 hours ago, michael8554 said:

Why still Red ? Well the sensor is now mono, but the processing is still RGB, and is applying an appropriate White Balance.

A Custom White Balance should correct the caste.

Michael

Thanks Ill give that a go  makes sense.  

do you have any idea on what might have caused the cross? i assume i've touched something i shouldn't have but knowing what is the tricky bit. Now i have got a sensor to work mono modded, i want to get it spot on... well as spot on as you can, trashing a perfectly good camera 😜 

Thanks Michael.

 

Matt 

[Noobulosity]

Totally forgot to post an update here.  I finally had an opportunity to test my 7D mono mod, after months of nothing but clouds around here (and a few clear nights where I had no opportunities to image...  but mostly clouds).  I didn't get very far before more clouds moved in, but I managed to get some shots in on IC5070 Pelican Nebula with 600s exposure times.  I shot H-alpha only, using a Baader 7nm H-a filter, 600s, a mix of ISO400 (2 frames), ISO800 (6 frames), and ISO1600 (3 frames).  Last time I had a working sensor, someone suggested longer times were best with narrowband filters.  So, that's what I tested.  No calibration frames or dithering here, I just didn't have time.  And...  I was just plain lazy and didn't feel like it.  This was just a test anyway, not a serious imaging session.

In any case, this is what I ended up with.  Still pretty noisy, which I'm sure would improve with additional exposures and calibration frames to smooth things out.  Taken with my WO Zenithstar 103.

This one is just the straight stack, no editing at all:

With a little work done to it:

[Stub Mandrel]

There are some very effective actions for removing horizontal banding on canon images.

[Chriske]

Did not read the whole thread, sorry.
Could older camera's be used for debayering ..?

[scratlovesacorns]

Created a account to share my experiences.

I ruined my first sensor for my 450d. Thankfully there are some refurbished sensors going pretty cheap on eBay(US). I was successful on my second attempt. I learned from my first sensor that, at least  for the 450d, the sensor is surprisingly durable. I tried scratching the dead sensor with a wooden pick and found you would need to apply A LOT of pressure. The easiest way to remove the glass protecting the sensor is using a hair dryer at the lowest settings. I'm sure others have done the same but I can confirm this works beautifully along with an x-acto tool. I was able to remove the glass in one piece. I found that using a bamboo cocktail pick works wonders as a scrape tool. This along with the Meguiars PlastiX used by conehead, can all be bought on Amazon. I initially scraped away a small amount of the bayer filter with tweezers VERY gently. An opening will be enough to start scraping away with a bamboo pick sharpened to a blunt but also sharp tip. I then polished with the plastix with a q-tip very gently. I've attached my flat and it is far from perfect, however I will count my blessings as I think it is good enough. It's nothings flats and dithering can't take care of. Of course with the timing in everything in this hobby, I have cloudy skies the next few nights so I cannot test this out on DSOs. But I will share my results when I do get a chance.

Edited August 22, 2020 by scratlovesacorns

[Noobulosity]

On 21/08/2020 at 05:15, Chriske said:

Did not read the whole thread, sorry.
Could older camera's be used for debayering ..?

Sure. Pretty much all color cameras have a Bayer matrix. You could practice on cheaper, old cameras.

[terrance]

you can absolutely scratch to the very edge, there's a "non active" margin between "active area" and other circuit, which can buffer some damage.

My Rebel T3i(600D) , just need some fine clean up

[terrance]

(not so) flat field

[Gina]

Pretty good that!!  👍

[terrance]

Just never use any metal toolings, which leave residue. And do NOT use alluminium, the oxidized surface layer is extremely hard. Bamboo toothpick is quite ok but sort of too soft.

I'm trying IC packaging epoxy (just grab some dead ICs).  ...

[R26 oldtimer]

What is more expensive than astrophotography??

The answer is diy astrophotography equipment...

Unfortunately I've managed to cut the tiny wire, not from the start but after 4 hours of successful scraping, when I was almost there. 😱🤬

[terrance]

needs patient and steady hands...😜

EOS Rebel T3i(600D), and Nikon 1 J5. J5 sensor cover is extremely hard to remove. My hot air station melted glue and desoldered it with the cover completed unaffected,

Edited November 5, 2020 by terrance
add more photos

[Sven_160580]

Are you still working on debayering sensors?

 

[terrance]

I have  tried IC packaging epoxy, and failed, it's too hard and WILL damage metal layers.

[markleonardo]

I have a 300D which I was going to mod anyway by removing the IR filter. I have read in the original reviews from 2003 that the 300D is a plastic case version of the 10D, using the same sensor.

I guess most who originally contributed to this topic have moved on to other things, but has anyone tried the cover glass removal on this sensor and succeeded or failed?

The UV light baking to make the ML/bayer array filter epoxy brittle and flaky sounds good.  Sunlight is pretty rich in UV, although it takes a few months for the deterioration to show when I've exposed fixings outside.

 

Mark