DSLR Exposure Time & Underexposed Pixels - Forums

Topic starter

Aug 31, 2021 · Edited: Aug 31, 2021

To start, my apologies for another post on exposure length, but sometimes it's hard to find precisely what you're looking for...

This is sort of a continuation of my previous post on underexposed red pixels, but the factors have changed slightly. I've now modified my camera (LPF2 removal) and am getting a lot more red data. It's much easier to bring the red channel's "skyfog" peak up above the bottom of the histogram even with an aggressive CLS filter. My question now is, should I be aiming for 0% underexposed red pixels when choosing an appropriate exposure time?

I've been using RawDigger to check the over/underexposure information on my RAW files. With the red channel peak well above the bottom of the histogram, I'm still getting 5-10% underexposed red pixels. Increasing exposure time brings that number gradually down, but at the cost of overexposing my green and blue channels slightly (burning-out brighter stars) and increasing my number of rejected subs (I'm shooting unguided, and I have to toss up to a third of my subs when I start shooting over 60 seconds).

To sum up, I'm wondering if it's common practice to tolerate a certain amount of underexposed pixels in order to properly expose stars in green & blue channels. If so, it would definitely lead to a greater number of useable subs and improve my star colour. Or is any amount of unexposed pixels so undesirable as to necessitate longer exposures?

Really curious as to how you all handle this. Thanks!

Clear skies,
Mark

PS: To be clear, this is only an issue with shots using a filter. I shoot broadband targets without a filter and have no issues with underexposure.

Well written Insightful Respectful Engaging

Björn Arnold

Perfect post

Sep 2, 2021

Hi Mark,

When I am shooting mono, I always have the same exposure times for the filters. No matter if narrow band or broad band. Therefore the behavior is the same as for your DSLR: blue and green wavelengths are brighter than red due to the lower spectral QE of the sensor.

With your light pollution filter, the signal transmission is likely imbalanced and therefore if it where a mono camera one would need different exposure times for the channels.

In any case, unless I am hunting the stars of our galaxy, I need to bring out the faint details of the target. Therefore, you will have to expose as long as necessary to catch the faint signal and get a good SNR. Of course, you certainly overexpose some brighter stars and they will be white.

Typically lowering gain/ISO (assuming most people start with higher values) will increase the usable full well depth of the sensor and therefore you can exposure longer before the bright stars have filled the buckets.
The problem will usually then be the limitations from the gear. In your case the tracking.

The alternative is to use the shorter exposure times with the higher gain and produce many subs which will increase the dynamic range to some extent of the integrated image.

In either case, the subs/channels shouldn’t be underexposed.

Björn

Helpful

Mark Germani

Topic starter

Sep 2, 2021

Thanks @Björn Arnold - this all makes sense.

I've actually been wondering recently whether or not I should do a separate run of shots for green & blue data, taking care not to overexpose them, but then even with some underexposure, I'm still overexposing cores of stars in the red channel (both overexposed and underexposed in the same exposures) so I think that lowering the ISO, as you suggested, would be better in the long run. So... when I can afford to start guiding, I can tame this beast, perhaps.

Thanks again & CS!
Mark

Björn Arnold

Perfect post

Sep 2, 2021

Hi Mark,

I am wondering how you overexpose stars in red while in green and blue they are not overexposed (of course, most stars are reddish but that’s not the point). I am just wondering which CLS you are using? With the modded cam, I guess you should use the CLS-CCD version from Astronomik.

Cheers,
Björn

Well written Insightful Respectful Concise Engaging Supportive

Mark Germani

Topic starter

Sep 2, 2021 · Edited: Sep 2, 2021

Björn Arnold:
Hi Mark,

I am wondering how you overexpose stars in red while in green and blue they are not overexposed (of course, most stars are reddish but that’s not the point). I am just wondering which CLS you are using? With the modded cam, I guess you should use the CLS-CCD version from Astronomik.

Cheers,
Björn

Sorry, I wasn't clear here - I am overexposing star cores in all channels, but red is also slightly underexposed. The point I was trying to make was that it probably wouldn't help to try and adjust exposure times for each colour until I wasn't overexposing anything - I'd probably just end up underexposing each colour to some degree as well.

Björn Arnold:
I am just wondering which CLS you are using?

I did the LPF2 removal only, which shouldn't require the CCD (LPF1 filters out some IR & most UV, from what I have been told). Practically speaking though (and Jerry Lodriguss stresses this as well) with refractors, a lot of IR still gets through the LPF1 so I'd suffer from less star-bloat with the CCD. I don't have access to the CCD version of the Astronomik CLS so I'll have to make do for now, but I intend on upgrading to the Optolong L-Pro which has UV/IR cut built-in, so maybe that will tame some of the overexposed star cores?

So many variables!

CS,
Mark

Helpful Insightful Respectful Engaging

Related discussions

Sub-frame rejection rate, exposure length and ISO

Hi, I only starting this amazing hobby one lunation ago - and joined this forum at the same time. I am pretty pleased with my photos to date, although very much in awe of what many are able to do on this site. In that vein, I hope the answers to thes...

DSLR exposure time discussion relevant to author's technical questions about settings.

May 20, 2020