Debayering a DSLR's Bayer matrix.

[Thalestris24]

I believe you can put AR glass in front of (on?) the sensor but don't ask me how it's done!

[GTom]

30 minutes ago, Thalestris24 said:

I believe you can put AR glass in front of (on?) the sensor but don't ask me how it's done!

That doesn't help the reflection of the silicon surface. Of course with an AR window you won't get ghosts back but you still loose the light what's reflected.

[Thalestris24]

3 minutes ago, GTom said:

That doesn't help the reflection of the silicon surface. Of course with an AR window you won't get ghosts back but you still loose the light what's reflected.

I think the glass can be bonded to the silicon surface but don't quote me!

[GTom]

10 hours ago, Thalestris24 said:

I think the glass can be bonded to the silicon surface but don't quote me!

That's very well possible. If I decide for a commercial conversion (Brent most likely), I'll certainly ask in general about sensor reflectance. Actually my worry is more the ghosting on the rear lens of my flattener than the actual light loss. I have a tsflat2.5, which might be a great glass but its coating looks too shiny. I've also seen complaints on the Riccardi reducer.

 

Edit: Ouch, You got quoted,  sry. Anyway, I am more fan of a coating like-solution, instead of bonding the glass.

Edited December 8, 2016 by GTom

[Adam J]

11 hours ago, GTom said:

That's very well possible. If I decide for a commercial conversion (Brent most likely), I'll certainly ask in general about sensor reflectance. Actually my worry is more the ghosting on the rear lens of my flattener than the actual light loss. I have a tsflat2.5, which might be a great glass but its coating looks too shiny. I've also seen complaints on the Riccardi reducer.

 

Edit: Ouch, You got quoted,  sry. Anyway, I am more fan of a coating like-solution, instead of bonding the glass.

The picture I posted of the mono 1000D  m45 is just a bare sensor.

The first optical element that light reflected from the sensor encounters is my CLS clip filter followed by the Baader MPCCMKIII (which is fully multi-coated) then a threaded 2 inch UV/IR block filter (the CLS is not a CCD vershion). I would not say that I had any horrible reflections especially given how bright M45's stars are, reflections were comparable to RGB camera for an identical exposure.

I think reflections are actually more likely to be an issue if you place an optical element very close to the sensor.....I just left the sensor bear with the wires protected by a plastic shim within the old filter mount. 

Others may had a different opinion but I would leave the sensor 'naked' as the process creates allot of debris and to be honest its useful to be able to clean the sensor using mirror lock up / manual sensor cleaning. You will end up with left over debris around the margins of the sensor as you cant get too close to the gold wires when cleaning the sensor after the mod. If debris shifted on the sensor and you had a clear glass element in front of it then you will almost certainly end up having to full disassemble the camera to clean the sensor.

The best thing to reduce relections from the sensor is to keep the nearest glass element well away from it. Also having poked, heated, scraped and mechanically polished the surface of the sensor....do you really think it needs a glass layer to protect it? A clip filter will do the job of keeping dust out just fine and will be significantly further away from the sensor. You can get clear MC glass clip filters for EOS APS-C from modern astronomy....not so sure about a full frame cameras though. 

Dont try to bond anything to the surface of the sensor!

Edited December 8, 2016 by Adam J

[GTom]

There are clip filters for FF Canons from Astronomik. However, this case I'd raise the bet and vote for something airtight and fill the housing with argon (I have access to) in order to eliminate any kind of internal fogging.

Regarding optical elements: I'll have a Baader filter maybe ~50mm from the sensor and a flattener with concave rear glass (=enhances the reflected light) around 90mm from the sensor. Besides ghosting, I would like to capture those photons too :D.

[Thalestris24]

Um, I might have been thinking of the Baader BCF filter or similar clear glass which I think just goes in the sensor carrier?

[Stub Mandrel]

Surely reflected or not, it's going to be lost light anyway as it's only photons that miss the light-wells.

You can't ARE coat the wells themselves without stopping them working.

[GTom]

45 minutes ago, Stub Mandrel said:

Surely reflected or not, it's going to be lost light anyway as it's only photons that miss the light-wells.

You can't ARE coat the wells themselves without stopping them working.

What about a plastic/polymer type coating? (Don't ask which one, I am still seeking options here). Obviously the coating has to be a strong electrical insulator and shouldn't use heat, etc for curing.

[Stub Mandrel]

This is interesting:

https://courses.cs.washington.edu/courses/cse467/08au/pdfs/lectures/07-cmos-ccd-imagers.pdf

 

[Adam J]

23 hours ago, GTom said:

What about a plastic/polymer type coating? (Don't ask which one, I am still seeking options here). Obviously the coating has to be a strong electrical insulator and shouldn't use heat, etc for curing.

The problem is applying a very very even coating. If you get a variable thickness you will have all sorts of distortions in the image and a horrible flat frame. 

[HyperCams]

I can add that for anyone trying the polishing method, I was told by an engineer from CDS that polishing removes a layer that he called the "optical pad" between the bayer and the sensor and that would lose some sensitivity...
I have no data to back this up other than my personal observations of comparing the polished sensors vs the scraped sensors and I could easily see a difference. I split a sensor once, half polished and half scraped, and could easily see the difference and the scraped was "brighter". By brighter I mean the actual images from the sensor...

I have to look to see if I kept the images or the sensor (probably not as I have used all of the dead ones as testing beds for technique over the years)...
someone may have already covered this here, if I missed it I apologize.

And now, a word on scraping. LOL

Oak. Oak is your friend. I tried a bunch of different materials when developing the scraping method. I went to the local hardware store and bought a bunch of 1/4" wood dowels of different wood types. Everything they had. Also some plastic/composite and metal ones but those either did not work at all or went straight through to the silicon sensor and were useless *for me*.

If you take an oak dowel and sharpen one end to a chisel point/end (it needs an angle, not a curved edge) it should work for you. getting a good edge on it is not easy but doable with some practice, luckily you can get it cheap...I bought a 1/4"x24" dowel over two years ago and am still using the original piece I cut off of it at ~6" (same length as a pencil/pen, also used a foam pad for ergonomics or a bunch of rubber bands wrapped around it to make a comfortable handle)..

But a key note, I do not push the material off in a back to front or away from me motion. I move side to side like you are coloring in a picture with a crayon. It is a more natural movement for me and worked best for my efforts.

The tool will actually work best with the bayer material stuck to it. You may have to move the tool around a while on the Bayer to build up a small layer..it gunks up pretty quick though. 

Do NOT press "hard".  Move the tool very lightly from side to side and slowly add more pressure until you see gold. Then keep doing the same thing a million times over and over and over LOL

Always start in a corner so if you do scratch it you will not damage the whole sensor by having a "bad" area right in the center of the frame...

Oh....and NEVER CLEAN YOUR SENSOR *with chemicals* BEFORE YOU START!!!!!! Never put ANY chemicals, even rubbing alcohol on it. I only ever use a bulb blower or a dry swab to remove particles before beginning work...which is needed as they will scratch your sensor.
Never leave the sensor to sit for a long period of time after removing the cover glass. Never partially remove the bayer and leave it for even a day.... Finish once you start and never use chemicals to clean while you are working. Blow whatever you can off as you go and also you can go over the same area multiple times with the dowel to make sure you got everything. 

WHY? I swear it seems like Canon's bayer actually hardens after being open to air for extended times, or (especially) when soaked with a chemical like alcohol...Maybe it is just me though..I have damaged quite a few sensors this way because the amount of force to remove the bayer after exposure was too much and would damage the sensor...

Also, use a microscope to do this kind of work. It is hands down the best investment I have ever made for this work. I own this one here: LINK to Amazon

This is all just "what has worked for *ME*" and YMMV.
Anyways, just some (hopefully useful) info off the top of my head while I had it.

I really need to finish some tutorial videos etc as I have most of them already and have the ability to record through my microscope as well. It might help some understand what I am saying and newcomers who may not know much about the processes etc...and really the techniques needed to do this well and have a very clean result.

Clear skies!
Brent

Edited December 9, 2016 by HyperCams

[GTom]

Thanks for the great deal of info you are sharing us Brent. The existence of the "optical pad" is true, you can clearly see them on the chipworks electromicrograms.

Is it difficult to get the oak "singing scratch free"? If possible I am aiming to have a clean flat, that I can use the cam for normal daylight shots too.

Videos are very welcome! There is not much on youtube, a few monents of old, 350d-450d conversions, but nothing on the "real thing" how e.g. a mono 6d comes to this world:). Once I get there i will document every move at least with photos.

Edited December 9, 2016 by GTom

[HyperCams]

So my first message here to the group was apparently deleted or lost in translation as it is no longer here....my first post that I can see now, the one right above GTom'so post was a followup to my original post (now nowhere to be seen lol) where I introduced myself and gave a bit of history on me and the cameras I have done and all of that jazz...

Anyways...Sorry but I'm not gonna retype allllllll of that again with my 3year old son climbing all over me lol

So....Hi everyone! I'm Brent and I'm an addict.......I mean I modify cameras and love doing it! Or maybe I am an addict and am addicted to performing open heart surgery on cameras

 

Clear skies!

Brent 

[zfedoran]

Has anyone tried using liquid nitrogen to remove the sensor glass? 

Just ran across this video, it would be interesting to see if this or dry ice can be used to remove the protective glass over the sensor.

 

[GTom]

On 12/13/2016 at 21:31, zfedoran said:

Has anyone tried using liquid nitrogen to remove the sensor glass? 

Just ran across this video, it would be interesting to see if this or dry ice can be used to remove the protective glass over the sensor.

Wow, that's interesting . I need to grab a 1$ sensor for that test - however, that doesn't prove that a 6D sensor would survive the icing + thermal shock. Actually I could avoid icing by blowing argon on the sensor until it's room temp again, the question is just, if anything cracks in the sensor stack (electrodes, optical pad)?

Edited December 15, 2016 by GTom

[Marf1]

hello

where i can download example mono raw images vs Colour images?

(No astronomy  images)

Thanks

[Adam J]

1 hour ago, Marf1 said:

hello

where i can download example mono raw images vs Colour images?

(No astronomy  images)

Thanks

Weird thing is that very few people have done proper side by side comparisons. 

[Marf1]

thanks

I want debayer my sony nex 3...is it possible??

 

[GTom]

On 17/12/2016 at 16:08, Marf1 said:

I want debayer my sony nex 3...is it possible??

Heard about nex5, it should be doable. Grab a few spare sensors from ebay to stay on the safe side.

[GTom]

Is the 1000D a good camera to practice on, if the ultimate goal is a 6D? Shall I expect the same issues?

[Adam J]

On 22/12/2016 at 09:32, GTom said:

Is the 1000D a good camera to practice on, if the ultimate goal is a 6D? Shall I expect the same issues?

Its difficult to say....they are separated by a number of sensor generations. My experience of the 1000D was that it ended up only just slightly better following mono conversion.....my point being that if you do try and its successful then you may as well have something worth while for you efforts. On ebay you can often get cameras that are 'broken' in such a way as to not have any effect on their performance for astronomy..auto-focus for example or scratched view finder. So if your patient then you might be able to get something more up to date but still at a rock bottom price.

[StuartJPP]

I have got a Canon 1000D (Rebel XS) with a spare sensor that I have been meaning to mono modify but I realise that I just may not get around to it so may be putting it up for sale soon for cheap...it is currently astro modified (front filter intact).

[ccdfreak]

Some time ago I tried to debayer my Sony NEX-5 / 14MP sensor, let me share my experiences with you:

1. It was quite hard to remove the cover glass. The only thing that helped here was to crack it by shocking it thermally und then removing the shivers manually.
2. I fixed the bonding wires with UHU endfest, a 2 component glue, to protect them.
3. The tool I used to scratch was a simple wooden toothpick

Attached you can see the sensor at work in progress:

and the "First Light" of the debayered sensor. Unfortunately I touched some parts of the circuitry, so the left part is nearly dark with extremely low responsivity. On the outer right side there's still some CFA left, you can still see some colors in the pillow:

At the moment I'm stuck with the procedure, so I've decided to stop here and save my money for a ASI 1600 mono :-)

CS,

Andreas

 

[ErwinL]

On 8.12.2016 at 14:52, GTom said:

What about a plastic/polymer type coating? (Don't ask which one, I am still seeking options here). Obviously the coating has to be a strong electrical insulator and shouldn't use heat, etc for curing.

I tried this when I debayered a 1100D two years ago. The effect was hardly visible, however, much weaker than one would expect for a silicon surface. The reason, I believe, is that the surface index is much lower than that of silicon, while the high reflectivity originates in a buried index step. Such a buried index step can be compensated partially by AR coatings, but very special layer properties have to be met.