25 replies to this topic

[Markinspace]

I’m stumped.  With sharpcap in live view and then later in pixinsight, my darks are not subtracting out the amp glow from my QHY 183m. All settings and temperatures are the same. I have tried just the master dark and an individual dark, no go.

 

Any suggestions.  It was working just fine two weeks ago.

[CurtisPorter]

Make sure your gain on the camera is the same as when you took your darks. I moved to a different computer not long ago and missed the fact that the gain setting was different by default. The result was amp glow after it should have been subtracted.

[Markinspace]

Gain, offset, temp, exposure duration all equal.  

[bobzeq25]

I used both 183s for years and never had a problem. Here's precisely what I did using PixInsight. Nothing more, nothing less.

Take many bias at the same offset and gain as everything else. Lowest possible exposure. Integrate (stack) them into a master bias. Using ImageIntegration.

Take a number of darks at the same offset and gain and exposure time as the lights. Integrate them into a master dark. Using ImageIntegration.

Take a number of flats at the same offset and gain. Exposure time such that the linear histogram was approximately centered (that need not be precise). Calibrate each flat with the master bias. Using ImageCalibration. Integrate the calibrated flats into a master flat. Using ImageIntegration.

Calibrate each light with the three masters, bias, flat, dark. Using ImageCalibration. Integrate the calibrated lights into a stack. Process.

All this is well described in this book. Common mistakes (generally involving checking the wrong boxes) are highlighted.

https://www.amazon.c...y/dp/3319976885

Edited by bobzeq25, 25 March 2025 - 12:49 AM.

[rj144]

Don't you need to take dark flats to remove amp glow?

[jml79]

How long are your flat frames and does your master flat have any amp glow? If you stack the flats without calibration is there amp glow in the master flat?

[bobzeq25]

Don't you need to take dark flats to remove amp glow?

NO. Dark flats, unless you have an old 294 or older 1600, are essentially an "urban legend". With modern low thermal noise and linear cameras they have NO advantage over the simpler bias.

The 183 definitely does not need them. I use nothing but the simpler bias. I had no amp glow in my years of using both 183s.

I'm not an outlier. Look at these tables for 183s in a thread from the experienced imagers forum from 7 years ago. Note the reference to bias.

https://www.cloudyni...olour-versions/

Did I mention dark flats are pretty much an urban legend? <smile> They do work as well as bias, but no better. And you need only one bias for all kinds of flats. MUCH simpler.

Edited by bobzeq25, 25 March 2025 - 02:27 PM.

[Yomamma]

how many darks do you use?

[Markinspace]

20.

[jml79]

NO. Dark flats, unless you have an old 294 or older 1600, are essentially an "urban legend". With modern low thermal noise and linear cameras they have NO advantage over the simpler bias.

The 183 definitely does not need them. I use nothing but the simpler bias. I had no amp glow in my years of using both 183s.

I'm not an outlier. Look at these tables for 183s in a thread from the experienced imagers forum from 7 years ago. Note the reference to bias.

https://www.cloudyni...olour-versions/

Did I mention dark flats are pretty much an urban legend? <smile> They do work as well as bias, but no better. And you need only one bias for all kinds of flats. MUCH simpler.

This has me incredibly curious. The 183 is an older camera which has some amp glow. It is established that the 294 needs longer flats with matching dark flats and in some cases the 1600 requires dark flats as well. Why wouldn't flats taken with a 183 that are long enough to have some level of amp glow also not benefit from dark flats? I completely understand that under most circumstances the 183 can get by with bias frames but as a technician there is a reason I asked the questions I did. If there is some amp glow in the flats, they would transfer to the image and the best way to get rid of it is to use dark flats. The best way to see if this is the source of the issue is to inspect the master flat. I could be completely wrong but if nobody even checks, you can't rule it out.

[dx_ron]

The 294 cannot use bias because the sensor is non-linear at very short exposure times.

 

The 183 can produce valid bias frames and, in most cases, a master bias is an acceptable substitute for calibrating flats. But the 183 has a lot of amp glow. If your flat exposures are multiple seconds long, I would recommend dark-flats to be on the safe side. Dark-flats will never be worse than bias for calibrating flats.

[bobzeq25]

This has me incredibly curious. The 183 is an older camera which has some amp glow. It is established that the 294 needs longer flats with matching dark flats and in some cases the 1600 requires dark flats as well. Why wouldn't flats taken with a 183 that are long enough to have some level of amp glow also not benefit from dark flats? I completely understand that under most circumstances the 183 can get by with bias frames but as a technician there is a reason I asked the questions I did. If there is some amp glow in the flats, they would transfer to the image and the best way to get rid of it is to use dark flats. The best way to see if this is the source of the issue is to inspect the master flat. I could be completely wrong but if nobody even checks, you can't rule it out.

You don't need to characterize the amp glow in the bias. That's done (fully) when you calibrate the lights with darks.

This is not a close call.

But, as I've said, technically dark flats are not worse than bias. If one wants to go to the extra trouble of shooting dark flats, it won't hurt your images.

And, it won't help them either.

Again, that is not a personal deal, it's just a fact.

I'd argue that bias is better for amp glow in the 183 because the imager is less likely to make a mistake with them. <smile>

Edited by bobzeq25, 26 March 2025 - 01:47 PM.

[Drothgeb]

Flat darks are just longer exposure bias frames, up to the exposure length of the flats. They are always best, but are usually not needed. 

 

I use two 294s. As long as I keep the flats and bias/flat dark exposures between 0.3s and 0.8s, they work fine. I typically use 0.5s bias/flat darks, and just keep my flats within the range. If you keep the flat exposure in that range, even a synthetic bias works well. 

[bobzeq25]

Flat darks are just longer exposure bias frames, up to the exposure length of the flats. They are always best,

Disagree. In the vast majority of cases they are NO better than bias. And, you can do one master bias, and use it for months for EVERYTHING (provided you don't change offset or gain). Whereas, if you change flat exposure (as is almost always true if you change filters), you have to do a new set of flat darks.

First of all that's work for no gain.

Secondly, it's risky. You could easily make a mistake.

Bias > flat darks. This is not a close call.

Flat darks started when non-linear 1600s and 294s needed them. And morphed into one of DSO AP's more persistent urban legends. To say they've outworn their welcome is an understatement.

The situation is similar to people thinking they have to precisely level their mounts, or work at getting polar alignment below 30 arc sec.

All useless. <smile>The discussion of flat darks in this well titled webpage is good (although its suggestion to use a fixed value for bias instead of actually taking bias is not).

https://siril.org/20...ith-dark-flats/

Edited by bobzeq25, 26 March 2025 - 07:13 PM.

[Drothgeb]

Whether you call it a bias or flat dark, they’re both subtracted from the flat, the same way a dark is subtracted from a light. The whole purpose is to remove noise. You match a dark’s exposure to a light. It only makes sense that you would treat the flats the same way.

 

The noisier a camera is, the more it makes a difference. Newer cameras have less noise than the older ones, but they are still far from noise-free. The longer the bias/flat dark exposure is, up to the exposure length of the flat, the more noise it will contain that it can remove from the flat. It may be a tiny, tiny bit, but it’s there. There’s no good reason not to remove it. 
 

You might make a mistake while working with flat darks. I have set exposure lengths for my flats, and have folders with flat darks that work with them that last for months too. With me, bias or flat darks, same difference, I treat them the same. 

[Andros246]

Hahahaha I think we should all quit the hobby

Cant come to an agreement with rule #1 of astrophotography…

Calibration frames


My reason for the inconsistency across camera sensors whether the “proper” darkflats are required over bias?

ZWO probably gets their sensors from Sony’s reject pile and they have small sensor defects that changes how things should and should not work. (You really think some Chinese manufacture is going to pay full price for a perfect working sensor directly from Sony? I doubt it!)




Also a “dark frame” always should match the exposure length of whatever its calibrating. This goes for flats too. Bias works cause who knows why (most cameras you can skip it entirely) but that’s what a dark frame is supposed to do.


I also know someone who uses bias with a mono 294 and color 294 works fine… sooooo mileage can vary

Edited by Andros246, 27 March 2025 - 07:38 AM.

[Yomamma]

Have you opened one of the dark frames and looked at it?

[Markinspace]

Yes, they all have the amp glow, except the master which does not.  

[Markinspace]

I had to have changed a setting on something because they were subtracting perfectly the prior session before they didn’t.  But I simply cannot figure this out.

[jml79]

Yes, they all have the amp glow, except the master which does not.  

Can you verify your master dark doesn't have amp glow? I don't have a 183 but that doesn't sound right at all. If the dark subs have amp glow then the master should too. If it doesn't have amp glow then how can it correct the amp glow in the lights. I know this is not a 183 but this is one of my master darks from a 294M which shows the amp glow needed to eliminate it from the light.

 

[bobzeq25]

Yes, they all have the amp glow, except the master which does not.

Then you've done something wrong in creating the master dark. That mistake is almost certainly why your lights do not have their amp glow removed.

The good news is that you have a clear path forward. What I'm hearing is the individual darks have amp glow, but the master dark doesn't. So, create a new master dark by simply integrating (stacking) the darks. Which is the proper method for creating a master dark.

Does the new master dark have amp glow? If so, use it.

Edited by bobzeq25, 28 March 2025 - 12:07 AM.

[Markinspace]

I tried that by loading the individual darks into astropixel and even doing that no joy. I’ll take another round and see if starting fresh is the ticket.

[Markinspace]

So I took another look at the master dark and may have found something.  The individual darks clearly show the amp glow (as in the image posted by Jml above).  I stretched the heck out of the master and got a glimpse of amp glow but ran out of room on the left to stretch any further.  So I think I’ll try setting my gain higher and play with the offset to move the next round of image (lights and darks) histograms further right. Worth a shot I guess.

Edited by Markinspace, 28 March 2025 - 05:54 PM.

[jonnybravo0311]

Your darks gain/offset/temp/exposure need to match your lights gain/offset/temp/exposure. There's something else happening if you've got a set of matching darks that individually show amp glow, but when combined into a master don't, even when stretched. Can you upload like 5 or 10 of these darks to somewhere like Google Drive and share a link? That way people can take a look to see if anything stands out.

[Markinspace]

Concur, and they do match. I’ll try and load some up.