Debayering a DSLR's Bayer matrix.

[Gina]

There is no doubt about it Gina, you will have to run a very controlled test. Same subject, same night. Short subs at 3200 and longer subs at 1600. The processing should reveal the differences if any. If your theory is correct then both pics should look the same with equal processing.

Yes, I could certainly do that - best would be to inlerleave exposures - pairs of alternating short and long subs with associated ISO setting. No problem in APT. The problem is the weather - no telling when we might get a clear night

[Cath]

I guess it also depends on whether the ISO setting is adjusting the analogue gain BEFORE the ADC or whether it (at some point) the ISO setting is adjusting digital gain (AFTER the ADC). It might be that the analogue gain is adjusted up to a certain ISO setting, then after that it uses digital gain. A bit like they tend to do with optical zoom/digital zoom on some cameras. I don't know.

[Gina]

I guess it also depends on whether the ISO setting is adjusting the analogue gain BEFORE the ADC or whether it (at some point) the ISO setting is adjusting digital gain (AFTER the ADC). It might be that the analogue gain is adjusted up to a certain ISO setting, then after that it uses digital gain. A bit like they tend to do with optical zoom/digital zoom on some cameras. I don't know.

I don't know either. I've conducted a noise test run, now I need to do a signal test run. Problem is getting a continuous and steady test signal. There is also the problem of different DSOs having different characteristics, particularly different dynamic range, though I think we can say that they all have more dynamic range than any camera can cope with.

So the big question is, does using a high ISO limit the dynamic range? The next question is, how do we find out? A crude test on a familiar object with say 1m at ISO 6400 and 4m at ISO 1600 alternating should give an indication at least. There may not be a big enough difference with only a 2:1 ratio so I'm thinking 4:1 may be more revealing

[Ags]

I've been doing some tests with my 1100d with an image analysis app I wrote, and it seems to me that the 1100d applies the ISO gain to the analog signal right up to ISO 6400. Based on my results I'm pretty certain it is not just multiplying the digital signal at any ISO. Another interesting thing I've noticed is that the number of noise pixels stays constant up to ISO 1600, but shoots up drastically at ISO 6400. I think there is a low level of false signal in the chip that is quantized to zero, but over ISO 1600 the background murmer gets quantized to a positive value.

[Gina]

I've been doing some tests with my 1100d with an image analysis app I wrote, and it seems to me that the 1100d applies the ISO gain to the analog signal right up to ISO 6400. Based on my results I'm pretty certain it is not just multiplying the digital signal at any ISO. Another interesting thing I've noticed is that the number of noise pixels stays constant up to ISO 1600, but shoots up drastically at ISO 6400. I think there is a low level of false signal in the chip that is quantized to zero, but over ISO 1600 the background murmer gets quantized to a positive value.

Do you know what the sensor temperature was when you took those figures, Agnes? I've certainly noticed a disproportionately greater noise with ISO 6400 with an uncooled camera in the warmer weather when the EXIF T reached 20 or 30C. But my noise tests at around 2-3C with cooled camera shows a low noise level even at ISO 6400. eg. the noise with a 10,000 second exposure produces the sort of signal I get with a bright DSO at much shorter exposures. I think I'll go through my data and see if I can find some related examples.

[Gina]

Here's a quote from my near ambient cooling thread :-

Now have results from last night's darks sequence run. The computer on the Blue cooler camera crashed before the end of the run but I have full results for the Silver one. Here is an image and histogram from the last dark - 10,000s at ISO 6400. That's 2.77 hours or over 2 hours 45 mins. EXIF T showed 8C but I think this is reading about 3C high - ambient estimated at about 2C

- ->

[Gina]

Oh That quote didn't work - I'll attach the images again here.

- ->

This is the JPEG image scaled to 1024px wide in PS CS5 no other procesing and the histogram from the original JPEG image. The RAW CR2 image produced by the 1100D is not readable by PS.

Edited November 2, 2012 by Gina

[Ags]

That wo

Do you know what the sensor temperature was when you took those figures, Agnes? I've certainly noticed a disproportionately greater noise with ISO 6400 with an uncooled camera in the warmer weather when the EXIF T reached 20 or 30C. But my noise tests at around 2-3C with cooled camera shows a low noise level even at ISO 6400. eg. the noise with a 10,000 second exposure produces the sort of signal I get with a bright DSO at much shorter exposures. I think I'll go through my data and see if I can find some related examples.

That would make a lot of sense - I was using room temperature for reference. It also makes sense that the background murmur is thermally related - I would expect that. I'm getting my work colleagues to shoot test data on their cameras and I need an easy (but rough) standard temperature.

My app does not yet read temperature from the image files - I'm working on the this weekend... :-)

Edited November 2, 2012 by Ags

[Auntystatic]

The RAW CR2 image produced by the 1100D is not readable by PS.

Oh yes they are you need to update the camera raw software to about v6.7 that will allow PS to see the file.

[Gina]

Oh yes they are you need to update the camera raw software to about v6.7 that will allow PS to see the file.

Thank you - I'll look into that

[Gina]

I've just done a run with alternating ISO 1600 and 6400 in both Ha and OIII of the Cygnus Loop. Subjectively, the ISO 1600 subs are smoother with less noise. I'll do a quantitative assessment tomorow. The ambient temperature is higher than when I did my noise run by about 3 degrees C. Present conclusion is what I found in practice in the past - that 6400 is going too far. I'll do a test with ISO 3200 and 1600 later. The Cygnus Loop has gone out of sight from here.

[Photosbykev]

Oh yes they are you need to update the camera raw software to about v6.7 that will allow PS to see the file.

v6.7 isn't compatible with cs5, v6.3 is the latest version for cs5 that works

Sent from my iPhone using Tapatalk. Blame Apple for the typos and me for the content

[Gina]

v6.7 isn't compatible with cs5, v6.3 is the latest version for cs5 that works

Thank you

[Cath]

http://stargazerslounge.com/topic/51020-using-iris-to-measure-performance-of-a-canon-450d/page__hl__cr2#entry1566287

[Ags]

Did Ags cool his cameras for the test like Gina did for hers?

No SHE did not :-) The sensor was at 25-26 degrees when the test was run. I'm doing the tests at room temperature because my photography friends are reluctant to put their cameras in the refrigerator in the name of science! Also, to get stable temperatures across the ISO range in the 60 second test shots, I first bake the sensor with a five minute exposure - so very long exposures + room temperature = perfect storm of noise. I will repeat the experiment tonight outside under field conditions with no sensor baking, and cool-down gaps between each ISO sub.

[JamesF]

No SHE did not :-)

I did wonder about that, but decided that perhaps starfox knew better than me and that Agnes might be a non gender-specific name in Dutch or something

James

[Chris]

No SHE did not :-)

Ups! how embarrassing !, my apologies Ags.

[Photosbykev]

I think it was an honest mistake, Ags was coming out with some reasonable opinions so naturally you assume she is a he ;)

I'll get my coat (chuckle)

[JamesF]

Never mind your coat, I think you should be looking for your flame retardant underwear

James

[Photosbykev]

Never mind your coat, I think you should be looking for your flame retardant underwear

James

I always wear that

[Auntystatic]

v6.7 isn't compatible with cs5, v6.3 is the latest version for cs5 that works

Sent from my iPhone using Tapatalk. Blame Apple for the typos and me for the content

Well I'm using v6.7 of camera raw with ps cs5 and an 1100D with no problems what so ever , I suggest you click on updates in the help menu and see what it gives you, I still prefer digital photo professional over camera raw anyway.

[Photosbykev]

Well I'm using v6.7 of camera raw with ps cs5 and an 1100D with no problems what so ever , I suggest you click on updates in the help menu and see what it gives you, I still prefer digital photo professional over camera raw anyway.

Sorry my error, it was 7.1 that was not compatible with cs5.

One advantage of the Camera raw is, imo, the lens correction and vignetting adjustments. They seem better in Camera raw than in DPP, but for general use I use DPP or Phase one Capture Pro

[pixueto]

Well, since there is no way of removing the CFA without damaging the micro lenses, that begs the question: what's the purpose of the micro lenses? How important a role they play? The images with the CFA removed (and without the micro lenses) seem to be alright, don't they?

Another possibility would be to find a camera model that has the micro lenses under the CFA (if any) so it can be modded for mono astrophotography.

Edited November 4, 2012 by pixueto

[RAC]

Personally I wouldn't worry about the micro lenes. You couldn't do this mod with saving them either way if they were on top or under the CFA. Getting an even amount polished off is the problem. You can't fix it with flats if its as bad as my version and thats makes it useless in my view.

I think im just going to have to save up for a mono ccd. Currently thinking about a Atik 383L

[Gina]

Well it was an excellent try and I congratulate you :icon_salut: Is the variation in grinding depth really that bad? Does it really ruin your images? I too have concluded that I shall eventually want a mono CCD but can't manage the 383L - I plan to get the 314L+. As a pensioner it takes me quite a while to save up for things but I don't want to wait too long at my age. Meanwhile, I have plenty I can do with my three 1100Ds including NB. I'm planning to get better NB filters first, they'll be useful when I get a mono CCD anyway, but will improve my DSLR imaging in the meantime. Oh for some more nice clear nights