Jump to content

Narrowband

Debayering a DSLR's Bayer matrix.


RAC

Recommended Posts

you can buy toxic solvents on ebay :grin: (and some of the related internet chemical shops that advertise there)

Is that a hardware debayer pic ie you removed the filter?

I'm still waiting for an old camera at the right price point before having another go but I'm now tooled up and ready to go

Link to comment
Share on other sites

  • 2 weeks later...

Here's a video showing how the CFA scrapes away.

I used the metal pick which, although can scrape the layer below, is a lot finer so it's easier to see what's happening (and I was only using the broken 'practice' sensor).

Hi Dave, Gina,

This thread is riveting. I've gone through it again and I got a 350d off ebay, and have a source for some spare 350d sensors, also ordered a microscope so am going to try removing the cfa with a plastic tool.

had a few questions

was that the 350d sensor you used in the video?

did you just use the raw output for your images after the mod?

Gina, did you have to do anything like disable sensors to use the camera electronics in a custom body?

will it work bare bone as in without assembly?

if it works, and i'm guessing i'll need two or three sensors, i'll end up cooling it as well.

i'll record a video and post it up here if i succeed.

Cheers

Alistair

Link to comment
Share on other sites

Its good to see you guys still having a go with this. I still have a few 1000d's but have not used them after getting an Atik 383.

Keep going its fun watching.

Link to comment
Share on other sites

Hi Dave, Gina,

Gina, did you have to do anything like disable sensors to use the camera electronics in a custom body?

will it work bare bone as in without assembly?

Alistair

I carefully removed a bit at a time, making sure it worked afterwards - I managed to ruin two cameras in my efforts :D I didn't disable anything in software. There are many hardware parts of the camera that aren't used in imaging so I removed as much as I could to reduce weight and size. All the info is in my DSLR modding and cooling threads.
Link to comment
Share on other sites

thanks

I'm planning to get started removing the cfa on a 350d as soon as I get my microscope.

spare sensors are about $60 on ebay.

i'm hoping to stay away from the edges as much as possible.

I'll post my progress as I go.

I just got a qhy8L, so it would be great if I can get the L channel data from a mono dslr and rgb from the qhy8L.

thanks

Alistair

  • Like 1
Link to comment
Share on other sites

was that the 350d sensor you used in the video?

did you just use the raw output for your images after the mod?

Gina, did you have to do anything like disable sensors to use the camera electronics in a custom body?

will it work bare bone as in without assembly?

Hi Alistair,

Yes that's a 350D in the video. I've got a few of them from ebay that I've revived from the dead over the years so I know them inside out.

After the mod just use set the camera to raw mode (jpeg settings will do weird things with the colours and you'll lose the gained resolution increase). You can then use the dcraw program to convert them to 16 bit tiff files with full resolution if you're not using something with built-in support (i.e. MaxIm)

As for disabling other bits, and bare bones, you might want to look at this thread, where I did the same as Gina and just removed a bit and tested it. Some things are obviously not needed for astronomy (assuming rehousing) such as the autofocus circuit, lcd, flash, and viewfinder - but certain bits are still needed such as the hotshoe as the camera gets upset if it doesn't detect them.

Link to comment
Share on other sites

Thanks Dave, Gina, and others,

I finished removing the cfa on a working 350d under a microscope.

I've checked all the wires, none seem to be damaged.

that was a good tip Dave, to add a bit of tape on the top glass. it didn't want to come off in one piece. I didnt really have the patience either.

I didn't want to use heat.

anyways, as mentioned, the microlenses came off easily.

the CFA needed a bit more pressure but once there was a break in the layer, it was easy.

i did most of the centre without the microscope, but used it for the edges so I didn't cut any wires.

overall, i think i made only one small scratch on the layer below the cfa which I'm sure will show up.

but weird thing is that under the microscope lights, i can still see red, blue and green reflections. its like they're offset.

could this be a result of refractive index of that layer that's left?

i used the other end of a wooden paintbrush. worked quite well.

took me about 2 hours, now to test it out.

problem is i bought the camera of ebay but it didn't come with a charger. i've ordered one, but it'll take a few more days.

i'll post results once I get the battery charged.

i also managed to break the plastic clip on one of the ribbon cables. luckily the cable goes in a bit so can be held with tape. this is the cable from the back screen.

this is with most of the microlenses removed

post-12882-0-69386300-1373154984_thumb.j

with most of the cfa removed

post-12882-0-58025300-1373154994_thumb.j

all of the cfa removed

post-12882-0-58025300-1373154994_thumb.j

i'llt the video from under the microscope shortly. but its similar to Dave's.

Dave, have you done any astro testing with the debayered 350D?

Thanks to all the info here, it was fairly easy. took me under 2 hours and I don't have any screws left.

if this works, I'll then test with a bigger sensor, but I'll stay off the 1100D. :-)

my objective is to debayer a sensor that's a decent size and has between 4.5 and 6 micron pixels to match my 10inch F4 Newt.

Cheers

Alistair

post-12882-0-15212300-1373155001_thumb.j

  • Like 1
Link to comment
Share on other sites

oops, I should've previewed my post before posting. i've uploaded the same pic twice and incorrectly referenced it. anyone know how i can edit my post? also how I can embed the video here rather than a link?

here is a video showing the removal. it was very useful as I could do the edges under magnification.

thanks

Alistair

  • Like 1
Link to comment
Share on other sites

I was getting impatient so I connected my 1000D batteries to the 350D battery with two wires, that charged up the 350D battery and i was able to take a few test images.

The results look good to me, although hard to say cause I didn't take the colour images.

looking closely, it obviously needs a bit more cleaning, but I'm happy with the way its turned out.

at this stage I still have the factory IR filter in place. I will replace it with the UV/IR filter and try out some images with my telescope, but since I don't have the timer for the 350d, i'll be stuck to 30s exposures. so might try a few globular clusters as they don't need the IR filter mod.

i've ordered a timer and will order the IR filter or a clear glass.

since this is more of a test, my next go will be with a newer camera other than the 1100D or the 20D.

I will also add a cold finger.

Maxim opened the CR2 files and I was able to save it as FITS and jpg's.

so no issues with processing the images.

Cheers

Alistair

post-12882-0-09396200-1373174812_thumb.j

post-12882-0-17206700-1373174819_thumb.j

  • Like 1
Link to comment
Share on other sites

Well done. Just for fun I went and had a play scrapping the cfa layer off and it was very easy so this seems the best way by far.

Link to comment
Share on other sites

and another one at a lower mag showing how i handled the edges

Cheers Alistair,

These are great news ;)

I also have a spare 350D sensor and I could try my luck on it, just one question though, what is that sracping tool you are using? Is it wood thing...?

Thanks!

  • Like 1
Link to comment
Share on other sites

Hi Luis,

so am I.

the canon 350D can't be controlled with Backyard EOS for long exposures and I don't have the handheld timer yet. ordered that on ebay, on its way.

but till then, I've built an xp laptop so I can use the old version of canon's software called eos capture i think and that works only in xp.

so hoping to give it a go tonight.

I also just bought a 1000d, so looking forward to debayering that.

i'm going to remove the IR filter, mod it to B&W, and cool it. so it'll be my narrowband camera. although with a QE of 32%, it'll need longer integration times. but atleast I can combine Ha data from the mono with RGB from the QHY8L.

so i'll use an "L" filter with the mono dslr and RGB with the QHY8L for RGB henceforth.

looking forward to that.

Gina, have you done any tests with your modded 1000D you started this thread on?

there are some test images by Guy with his b&w nikon somewhere here I think.

I will post results soon. clouded out here at the moment.

cheers

Alistair

Link to comment
Share on other sites

Hi Alistair,

Those are exactly my thoughs, I have a TEC cooled 350D and it works like a champion, I've just bought a lovely 450D to go also for the cooling method (cold finger) this one has so little noise that will be just beatifull with cooling, the 350D will be my choice also for NB and deep Lum. work...that is if I succeed with the mono conversion :grin:

I had a play with my spare 350D sensor, a good friend of mine offered me that for me to have a go on debayer, first tests with a plastic needle resulted on sratches (bad ones...) on the underlayer, but with yout technic (woden paint brush) it came out easilly except for one part near one edge where the layer was really hard to remove, i just made a couple of very litlle scrathes, nothing compared with the other methods I used, so this is the way to go, I will do this tomorow with my beloved 350D and hope not to kill it...let's see. I would love this to be successfull, there's so many people that say that's impossible that I would just love to say: Here you go, nice hu...??!!

BTW, I use APT and a home made cable to control my 350D on my Windows Vista laptop...and yes,Vista sucks...

Best wishes,

Luís

Edited by Luis Campos
Link to comment
Share on other sites

Cheers guys,

Been working on my 350D sensor this weekend, I succeded on removing the bayer matrix with the wood end of a paint brush as explained by fellow Alistair (you are the man!!!! :))

The most difficult part for me was removing the glass cover from the sensor, this one was crazy hard to remove, I was hopping to remove it in one piece to use it again, but it ended up in pieces even with extra care...

Here goes my first test images shot through the ED80 aiming at a street lamp at aprox. 25 meters, I'm not sure if the RAW image is obtained the correct way though...using IRIS software (my favorite of all times) I still get the quadriculated pattern from the matrix, using DSS on the other hand and disabling the bayer matrix under the RAW / FITS configurantion and saving as FITS and then convert to TIF under FITS LIBERATOR, gives me a clean image on Photoshop with no sign of the matrix, I'm not sure if this is the correct way...it look darn sharp to me !!! :)

I'm not sure about the sensitivity, at a first glance it look like a bit less then with the original micro lenses and bayer matrix...can't wait to test this under the stars but it's cloudy here now... :mad::eek:

Oh don't forget I have this and TEC cooling so if this works it will be a blast, fingers crossed guys :tongue:

Please take a look on my Flicr galery, the first one is with the matrix straight from IRIS:

http://www.flickr.co...ith/9285240985/

Best wishes,

Luís

Link to comment
Share on other sites

Hi Luís.

Your pictures look great, but there's a couple of things:

It's odd that you can still see the matrix pattern in IRIS, as this shouldn't be possible. Could you maybe upload the raw file somewhere?

Also, what are you selecting in the RAW / FITS settings in DeepskyStacker? As far as I can see there's only four options, none of which seems to give an un-interpolated monochrome output.

Link to comment
Share on other sites

Hi Dave,

Yes, I dont get it either, not only that but i also get color images straight from the camera, I don't get it...???

Here goes the RAW file converted to JPEG:

CFA converted under IRIS: http://www.flickr.com/photos/37419943@N08/9285907537/

COLOR converted under IRIS: http://www.flickr.com/photos/37419943@N08/9285906799/in/photostream/

This is the sensor with the bayer removed, but with the Baader filter in place: http://www.flickr.com/photos/37419943@N08/9285962295/in/photostream/

This is the "training" sensor badlly damaged: http://www.flickr.com/photos/37419943@N08/9285958751/in/photostream/

I'm I doing something wrong? Why I'm I getting color???

Cheers,

Link to comment
Share on other sites

Hi luis

Well done

for testing, you can even set the format to jpg and open it. it will look reddish, that just needs a custom white balance. so open paint in your pc, set the background to white, full screen, take an image and use this in the camera's custom white balance. after that the jpg images will look b&w

btw that option in dss is for fits files, not dslr cr2's. its the first tab. try this option as attached

the cr2 files open as mono in maxim without any change

also, can you post pics of your cooling mod. how did you avoid sensor frosting issues?

do you have an Ha filter to test narrowband?

cheers

post-12882-0-20758700-1373843919_thumb.p

Link to comment
Share on other sites

Hi Alistair,

Thanks mate, I was starting to think I goofed somewere... :)

It's very confusing to me to see color in the images out of the camera, although the colors are muted they are there, it's so weird lol.

The scraping method went very good, But it's a nightmare to clean the sensor afterwords, I cleaned after each pass with a qtip embeded in pure acetone, this works exceptionally well as it drys instantly and leaves no residue.

I will post latter at night some images of my cooling method and some more of the heating system, I refined this heating system now that I opened the camera :)

Greets

Edited by Luis Campos
Link to comment
Share on other sites

I hate to be the one to say it, but could it be that only the microlenses were removed and not the filters? :huh:

I processed the second image you posted a bit, and there seems to be colour in there, although muted.

Perhaps take a photo of this image https://upload.wikimedia.org/wikipedia/commons/0/0e/Red-green-blue_flag.svg - you should only see different shades, not different colours.

Alistair, the Bayer Drizzle still seems to output a colour image. Although it doesn't interpolate, it will give each pixel a colour and a saved FITS file will have 3 planes for RGB. Since DSS uses DCRAW, it should be pretty simple to add the option we need though.

I think most of the time the Super Pixel method will be fine, unless you plan to display/print your photos at full resolution (a rarity for me at least).

post-970-0-09650200-1373881649_thumb.jpg

post-970-0-55898600-1373881659_thumb.jpg

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.