No Red Data with CLS Filter

Not sure what is happening with your specific image, but I can tell you that with a Svbony CLS filter in a Bortle 7 sky using a Canon 800D (T7i), I was able to see the nebula in a 2-minute exposure at ISO 1600 in the red channel.  That said, my blue and green histograms were right of center during collection which is not ideal.   If you have the money, investing in a narrow band h-alpha filter or dual narrow band h-alpha/OIII (like an optolong L-enhance) will pay off.  It will remove most of the light pollution and make the h-alpha emission nebulae pop - even in badly polluted skies.

Clear Skies!

Here is the spectral response of the channels in my camera:
https://www.researchgate.net/figure/Spectral-response-of-a-Canon-T3i-APS-C-CMOS-image-sensor-with-a-Bayer-pattern-CFA-This_fig5_287249644

Mi Mark,
indeed your image doesnt look like i'd expect in the first place. But something different caught my attention! I'm a bit confused of the Canon T3i (aka Canon 600D) spectral response image from that paper. It shows peak sensitivity at 650 nm. To be honest, i doubt it, despite it has been published in Optics Letters. At least, it must be a modified camera, but i found no mention in the paper prosa. Usually i would expect to obtain a spectral response from that camera, that is somewhat resembling that of the human eye - having something like 5% response in that spectral region.
Capturing absolute spectra, including spectral responses, is not a trivial task. Usually you need a calibrated tungsten strip lamp for intensity calibration. I heard of rumors, it is possible using the sun as an alternate method, but i think it is very (to terribly) inprecise because of aerosols.

I see that the blue and green channels have a small amount of response at 650nm.

Couldn't this result from h-beta (486 nm), which is very close to OIII?

CS Horst

Thank you all!

Sorry for the delay in wading back in here, I've had a busy few days.

@frankszabo75 thank you for the insight regarding clipped stars. I might either have to live with that, or try doing shorter or lower-ISO subs and combining them with a starless version of the nebula. My histogram is at the 1/4 mark currently. Since I'm imaging un-guided, my stars become more and more eccentric past 2 minutes so using a lower ISO for the nebulosity is not an option for me at this point, sadly.

@normmalin thanks, I think a narrowband filter like the l-enhance or l-extreme will be in my future, but perhaps not until I invest in a guiding setup so I can image for the longer times it will take with those filters.

I actually discovered the red info hiding beneath the bottom of the histogram when viewing the raw files in Adobe Camera Raw. As soon as I changed the white-balance slider and the tint, I was able to balance the colours and retrieve the red info. I have NO idea why the histogram doesn't show red info below a certain point, perhaps it's not in the desirable range of red for daytime photography so they don't show it? Alternatively, was I taking blue and green data and redistributing it as red by sliding the temperature and tint sliders in Camera Raw? The histogram didn't really bear that out, it seemed like the bumps for red and green were just just moving laterally rather than decreasing or narrowing.

@Björn Arnold some great insight, thank you! I decided to increase my exposure by upping my ISO rather than time, as I need to be mindful of how many subs I have to toss. I modified my histogram to show RGB separately instead of averaged, and nudged my ISO up to 1600. I increased my exposure time until the red channel was no longer clipping at the bottom. I think this is a good balance between blown-out stars and underexposed nebulosity. In the end, 75-90s at ISO1600 was the sweet spot for NGC 7000.

@HR_Maurer thanks - that's reassuring. I guess I should be a little more skeptical. I remember seeing more accurate spectral response curves for the 600D/T3i elsehwere on the internet, but I had trouble tracking that graph down. 5% transmission at 650nm would account for my reduced red data. @Björn Arnold your comment on the green response makes total sense. Also one of the reasons I love processing in APP with bayer-drizzle. I get twice the SNR on the green channel!

With all your feedback in mind, I did manage to get 25 minutes of pretty decent subs during a break in the clouds on Thursday night, and I did a quick process which revealed plenty of nebulosity which I was able to colour-correct. There is SO much to learn in this hobby!

CS,
Mark

Björn Arnold:
For my approach, I'm turning the gain down to the lowest possible value, i.e., 1. For a DSLR that would be ISO 100. Blowing out stars will happen but also keep in mind: increasing gain decreases the available full well depth and reduces dynamic range. If you reduce ISO from 1600 to 800, you should be able to double exposure time without clipping more than before.

Thanks @Björn Arnold. You’re right - I did find that statistically I get a better SNR at ISO400 than at ISO800 for the same integration time, but below ISO400 I have read that banding becomes an issue on Canon cameras, and some of my early camera lens work confirms this. Without a CLS sensor, I’m able to shoot 60-90s before blowing out stars at ISO400, and I used this approach when imaging M101 recently.

I would love to be able to take the same approach with the CLS filter, but that might have to wait until I get a guiding setup.

This has been really helpful in terms of setting my priorities for future investments. Thank you all for your input!

CS,
Mark

Ok, I thought I'd just add a few things here:

I've been imaging with the Astronomik for a few nights now, same target. First two nights I shot at ISO1600 for 75 seconds. Last night I tried ISO800 for 120s, after seeing a few subs with good stars (spoiler alert: I had to toss 2/3 of them).

I have only stacked the first night so far, as a test. Once I corrected the colour balance, there was plenty of nebulosity despite being underexposed on the red channel. I'm going to add stack the rest of them just as soon as I get dark frames for last night's session.

One interesting little tidbit, though: I found a program called RawDigger because I was wondering if the Canon histogram was for a gamma-corrected image instead of the RAW file (which I think it is). RawDigger revealed some really interesting stats on overexposed/underexposed channels. As I suspected, Red is dramatically underexposed:

75s @ ISO1600: 13.4% underexposed
90s @ ISO1600: 9.7% underexposed
120s @ ISO800: 26.4% underexposed (but looks better than ISO1600 - better contrast in nebulosity too)
180s @ ISO800: 1.5% underexposed (wait, what?)

There is a big jump in red channel exposure beween 120s and 180s @ ISO800. Given that the histogram sits roughly at the same spot (on my camera at least) for "x" @ ISO1600 and "2x" @ ISO800, this means that lower ISO and longer exposures are key here, which everybody has been saying anyway!

Without guiding I'll stick to 75s or 90s @ ISO1600 for now. I really, really wonder how much astromodification could further reduce the underexposure levels without having to extend exposure times beyond what my mount is capable of, without guiding.

Cheers,
Mark

Hi Mark,
maybe i missed something - how do you define "underexposure"?
Is this the missing percentage of the brightest (non-defect) pixels, to reach full bit count?

CS Horst