Nikon D5100 Monotec narrow bands

[Adam J]

18 hours ago, GTom said:

Sorry to revive an old thread. I am thinking on the ideal travel-astro-camera. Normal IR modification is relatively painless and reversible. However as I see, removal of the Bayer array is not for the faint-hearted, I would certainly not risk any 200++€/$/£ worth of camera with that. However, the big question is, how much does the process really yields? Has anyone made a direct comparison, same object, same exposure time, same filter+same scope?

I am interested both on H-Alpha and O-III regions, how much exposure time can I really save on a B&W conversion, if I am aiming a decent H-alpha, Oiii, H-beta set?

If the gain reaches 1EV or adds 1 to the limiting magnitude I have an effectively scrap-priced e-pl5 as astrocam candidate, which could in theory deliver better results than any color APSc camera.

I would not bother with H-Beta its more for visual use, its effectively showing you the same stuff as the H-A filter but at a much lower luminosity. I am going to do my comparison with H-A and then OIII and then look at luminance. I need a reasonably faint target....am at work so I am not sure but I think that the soul nebula is available at the moment. 

[Thalestris24]

11 minutes ago, Adam J said:

I have two camera one delayered the other one just astro-modified. As soon as we get clear skys ill be able to preform a H-A test and get a definitive answer on sensitivity / noise. Well at least for my model of camera. However I suspect there will be a degree of read across to others. 

Hi

I think it depends on what exactly you are comparing. There is probably little difference in sensitivity between a debayered camera and an astromodded one. If you are comparing narrowband performance i.e. via an Ha filter, then you have to bear in mind that an undebayered camera is only using 1 in 4 of the pixels for detecting red light. For that reason you wouldn't ordinarily bother to use a colour camera for narrowband although you can get an image. Also, the raw files are pre-processed differently. I've converted my cr2 files from a debayered 550d to mono via IRIS which lets you save as fit format. You can then stack and treat them as pure mono files.

Louise 

[Adam J]

30 minutes ago, Thalestris24 said:

Hi

I think it depends on what exactly you are comparing. There is probably little difference in sensitivity between a debayered camera and an astromodded one. If you are comparing narrowband performance i.e. via an Ha filter, then you have to bear in mind that an undebayered camera is only using 1 in 4 of the pixels for detecting red light. For that reason you wouldn't ordinarily bother to use a colour camera for narrowband although you can get an image. Also, the raw files are pre-processed differently. I've converted my cr2 files from a debayered 550d to mono via IRIS which lets you save as fit format. You can then stack and treat them as pure mono files.

Louise 

Louise, I am aware of the increase in the number of pixels available and I have been using IRIS in exactly that way to convert the test files for full resolution. For that matter I have also been leaving them in RGB sometimes to test that result as on a mono camera is akin to 2x2 averaging and so lower noise. In any case I have been performing comparisons based on converted raw files. 

I have had some great images from my cooled undebayered camera using H-A and OIII filters. The problem is that my comparison results are not totally adding up despite my being very very careful with my methodology between the two cameras. I suspect that there is some sort of calibration / balancing being applied by canon within the camera which may be tripping up when the camera sees a pure red image (as in the case when using a H-A filter). For example I already know that the red channel in any DSLR is boosted in comparison with the other channels to acheive a normal color balance. That being why the red channel tends to carry more noise in comparison to the blue and green (although the green gets two pixels so its naturally lower noise) its also why dark frames tend to have a red tint to them. You can also see it when converting an undebayered image to mono as the red pixels will always appear brighter when you use any of the normal white balance settings. 

When I balance the channels on the mono camera by using a custom white balance things get interesting. I suspect that when performing this balance the camera is drawing the red channel down to match the green and blue channels rather than boosting the green and blue up to the level of the red channel (I am going to guess that the camera balances channels with respect to the level of the green channel.  This leaves the mono image deficient in 'amplification' relative to the red channel in the RGB image, in effect its as if a lower ISO has been used within the red channel.  Ultimately I believe that this is leading to short H-A exposures (in daylight conditions) performing better under the RGB camera as the mono image is not being amplified out of the read noise while the red channel of the RGB image is. However, for long exposures within a dark room this reverses as the lower noise of the mono image comes into play and the mono image moves out of the read noise. However this is still not the full story as when I compare like for like using a daylight white balance on both camera I still get some strange results most likely indicating that another layer of processing is going on within the camera.   I am hoping that my real world test will be revealing. 

Incidentally if people have had the white balance set to the default auto value when posting past results then it will give very misleading answers. 

 

 

[Thalestris24]

Not sure if what you're doing is worth the time and effort... I'm not sure that doing a custom white balance is meaningful. The sensor is mono... Sure, the camera thinks the file is colour, but it isn't really. Software thinks the raw file is colour until you tell it otherwise. Apart from conversion to mono, any processing of the raw file that is thought to be colour will be incorrect.

Louise

[GTom]

11 minutes ago, Thalestris24 said:

Not sure if what you're doing is worth the time and effort... I'm not sure that doing a custom white balance is meaningful. The sensor is mono... Sure, the camera thinks the file is colour, but it isn't really. Software thinks the raw file is colour until you tell it otherwise. Apart from conversion to mono, any processing of the raw file that is thought to be colour will be incorrect.

Louise

The sensor is mono, but the firmware treats subpixels with different boost factors - the boost factor is actually part of the white balancing. WB is still on place with a mono sensor (the camera doesn't know that you scraped off the Bayer array), I'd certainly stay away from auto-WB. I don't have a straight idea what to start with, but I'd start around a well-balanced 6000K preset.

The comparison might be more difficult as it seems, instead of luminance values we should probably look at noise levels and limiting magnitude for astrophotos.

[Adam J]

35 minutes ago, GTom said:

The sensor is mono, but the firmware treats subpixels with different boost factors - the boost factor is actually part of the white balancing. WB is still on place with a mono sensor (the camera doesn't know that you scraped off the Bayer array), I'd certainly stay away from auto-WB. I don't have a straight idea what to start with, but I'd start around a well-balanced 6000K preset.

The comparison might be more difficult as it seems, instead of luminance values we should probably look at noise levels and limiting magnitude for astrophotos.

Yes exactly. If you take a picture of a white peice of paper with a mono camera then view the non debrayered raw image you will see a regular pattern of sub pixels which different levels, like a chess board pattern. Its because the camera still thinks its RGB and is applying different levels of amplification or boost to the red, green and blue pixels so that the debrayerd image will appear white to your eye...only without the micro-filters its not balanced at all. Try zooming right into a mono camera picture without doing this and you will see what I mean. You fix this by making a custom balance after mono modification and hey presto the chess board patter disappears. But the levels have been drawn down in the process resulting in an apparently darker image.

Its making testing difficult: Also see my post in the original thread concerning the effects of focal length. 

[Xplode]

Weirdly it varies if the chess board pattern can be seen, i took a look at my R G and B of M33 and only the blue channel has the pattern.

Narrowband has the pattern too, but not luminance.

If flats are taken with the correct filters that should remove the pattern nicely without any messing with white balance (which will mess up the channels without the pattern)

[Thalestris24]

On 11/17/2016 at 14:13, Adam J said:

Yes exactly. If you take a picture of a white peice of paper with a mono camera then view the non debrayered raw image you will see a regular pattern of sub pixels which different levels, like a chess board pattern. Its because the camera still thinks its RGB and is applying different levels of amplification or boost to the red, green and blue pixels so that the debrayerd image will appear white to your eye...only without the micro-filters its not balanced at all. Try zooming right into a mono camera picture without doing this and you will see what I mean. You fix this by making a custom balance after mono modification and hey presto the chess board patter disappears. But the levels have been drawn down in the process resulting in an apparently darker image.

Its making testing difficult: Also see my post in the original thread concerning the effects of focal length. 

Hi

I thought it was only necessary to convert the .cr2 files to fits mono via IRIS? But maybe I've been doing it wrong!!

Louise