72 replies to this topic

[Jon Rista]

 

 

 

 

This page is always helpful when in doubt. The term "dark flats" is proper.

 

http://deepskystacke...lish/index.html

I completely disagree. "Dark flats" is even worse! That sounds even more like you are talking about flat frames that are dark, which is completely wrong and confusing. Probably the main reason why people seem to be so confused about the term "dark flat" or "dark flats" when it comes up. 

 

That's why I've been rather pointed lately about using the term "flat darks", as it is a far clearer phrase. 

 

It's right from the source and as intended. You kids go right ahead..

 

Why is DeepSkyStacker the authority rather than somebody else?

 

See what I mean? Lets see. Because it's right from the author and as intended? What's next "dark masters" instead of "master darks"? Enough bickering already move on people.

 

The author of DSS did not invent the term. DSS is also far from the only tool that can calibrate and integrate data. I wouldn't consider DSS to be "the" authority on the term. 

 

I've seen confusion on this issue. I think using proper terms that properly describe what's what can help alleviate the confusion. It's not bickering, we were having a decent, reasonable conversation until your post here. 

[freestar8n]

I get confused on it also but I prefer dark flat. It is an exposure identical in all ways to the flat but it was kept dark during the exposure.

Dark modifies flat as a special kind of flat exposure.

But flat dark implies it is a dark that has somehow been flattened.

I have used both but I much prefer dark flat. It’s just like the flat exposures but without light.

Frank

[NorthField]

Truck tire flats = suckage

[Michael Covington]

I get confused on it also but I prefer dark flat. It is an exposure identical in all ways to the flat but it was kept dark during the exposure.

Dark modifies flat as a special kind of flat exposure.

But flat dark implies it is a dark that has somehow been flattened.

I have used both but I much prefer dark flat. It’s just like the flat exposures but without light.

Frank

Yes, that is the case for going the other way.  Both of the 2 words are nouns converted from adjectives.  If you convert one of them back to an adjective, it makes sense with "dark" as the first word.

So now I really don't know what to do!

Thanks for providing the interpretation for "dark flat."  I had it in the back of my mind but had not fully thought of it.  A dark flat is a flat exposure with the lens cap on, which makes it dark.

[freestar8n]

Yes. Jon feels it is fundamentally nonsensical to have a flat that is dark. Of course he does have a point. But when you consider that the whole point of a dark flat is to have exactly what the flat has except for the photons - it does make sense.

I don’t know of a definitive reference to cite.

Frank

[Jon Rista]

My point is that it's a dark frame, like any other dark frame. It's not a flat frame. It's just a dark frame that matches the exposure characteristics of a flat frame, rather than matching the characteristics of a light frame.

From the standpoint of proper grammar, it's "flat dark" or perhaps "flat's dark", "the flat frame's dark frame." Colloquially it seems people prefer "dark flat", maybe more for aesthetic reasons than anything else. When it comes to writing, personally I'd choose the grammatically correct phrase, but again...that's just me.

Edited by Jon Rista, 08 October 2017 - 09:52 PM.

[Sean13]

Maybe it's just me, I do believe Jon is right on the phrasing, but in my mind, I see it like Dark is describing the flat. As a dark frame is really a dark light frame, dark is describing the light frame. A light dark frame doesn't make sense, and to me flat dark doesn't make sense because flat is describing the dark in this phrasing, which is backwards, "dark" should be describing the "flat". The dark isn't flat, the flat is dark. As an example, add "A" in front of it to make a sentence. A Red Truck. Red describes the truck. A Truck Red doesn't make sense. A dark flat, or A flat dark?

 

Again, I do fully believe I am wrong, and that Jon is correct, but it's just the way my mind see's it and why I personally say it like that.

Edited by Sean13, 08 October 2017 - 10:04 PM.

[freestar8n]

When you take a flat you set all this stuff up and expose for x seconds. A dawn flat is a flat taken at dawn. Ok now you have taken dawn or twilight flats. You need matching dark flats to go with them. Ok. Cover the scope and take the dark flats the same way you took the dawn flats.

Great. Now you have darks to go with the lights. And you have dark flats to go with the dawn flats.

I don’t see it as bad grammar or inconsistent. It’s just a natural way to describe the different exposures and how they are captured and used.

There is nothing at all flat about a flat dark. But there is something dark about a dark flat.

Frank

[Jon Rista]

Guess it depends on how you look at things. 

 

The red truck analogy is interesting, but it's also kind of perpendicular. The way I look at it, flats are photonic frames, like light frames. They contain signal primarily derived from photons. Darks, regardless of what their intended calibration target is, are not photonic. They are a different kind of frame with a different purpose, distinct in their own nature. So it isn't like saying "Truck Red"...it would be more like saying "Truck Spot", say to describe the parking spot a truck is parked in. In contrast to say "Car Spot", to describe the parking spot a car is parked in, or just "Spot" or "Parking Spot" to describe "normal" parking spots not intended for trucks. 

 

However Frank's description isn't wrong. So clearly there are different ways of looking at things.

 

I guess it's all up to Mike now. Your choice, buddy! You could easily go either way. 

Edited by Jon Rista, 08 October 2017 - 11:02 PM.

[Sean13]

Yeah now I can see it your way as well. I guess for me it boils down to whats easier to say.

 

Geeze and to think all I wanted to know was if I still need bias frames.

[Michael Covington]

That was too easy, so we answered it right at the start and found something to argue about! 

[sws626]

I am an author myself and am trying to choose between two rival terms for use in something I'm writing. I know that both terms are in wide use. Merely finding one of them used somewhere doesn't settle the issue.

I completely agree with your and Jon’s reasoning. “Flat darks” is the expression that makes grammatical sense. They are dark frames taken under conditions that make them suitable for calibrating flats.

[freestar8n]

Some uses and abuses of terminology really bother me but this isnt one of them. Either way works and I don’t know if there is a professional convention. At the same time I disagree that this is a grammar issue. In terms of grammar the dark is not in any way flat.

But I don’t mind either way and I don’t think it matters much.

But when people talk about what noise is and isn’t. Or when they say collimation is wrong and it should be align instead. Or they say “C14 Edge” - those are all entirely different matters - but with plenty of references to refer to for correct usage.

As for the op’s main question. I do use bias and flats and I do sometimes use dark scaling with asi1600. I’m about to find out just how bad it is - but if it does work ok then it greatly simplifies compared to matching dark times with light and matching flat times to dark flats. Especially if flat times vary across filters.

Frank

[keithlt]

I have been using the unlighted flats and dropped the bias.

just use the same exposure as you used for flats keep the temp same and cap your scope.

of course continue to use all the other calibration frames except for bias. experiment

Edited by keithlt, 09 October 2017 - 07:58 AM.

[Jon Rista]

As for the op’s main question. I do use bias and flats and I do sometimes use dark scaling with asi1600. I’m about to find out just how bad it is - but if it does work ok then it greatly simplifies compared to matching dark times with light and matching flat times to dark flats. Especially if flat times vary across filters.

Frank

Frank, one thing to keep in mind is, if you have written your own software that you use to calibrate, if dark scaling works for you (i'm skeptical, but it may depend on exactly how you do it), that doesn't mean it will work for everyone. I've tested extensively, and the default dark scaling features of many programs definitely do not work properly with data that has amp glow (and that doesn't just go for the ASI1600). I think the main reason for this is a lot of dark scaling algorithms evaluate noise, and use that noise evaluation to generate a scale factor, rather than simply scaling based on exposure lengths. Noise-based scale factors are usually wrong.

 

Since a lot of beginners use this camera, I've tried to keep my recommendations simple and consistent for them: Don't scale. If you don't scale, and keep the darks well matched, then calibration always works. Whenever beginners deviate from that, they have problems with remnant or inverted glows.

Edited by Jon Rista, 09 October 2017 - 09:25 AM.

[Stelios]

From an "impact of terminology" viewpoint, the first word tends to define the commonality and the second acts as specifier. Blue pants, blue sky and blue horizons--the commonality is that they are blue.

 

The commonality between darks taken for flats and other darks is in the *darkness*. There is no commonality between darks taken for flats and flats--there is no 'flatness' involved. So it should be dark flats--dark frames to match flats, as (if the second term was not obvious) it should have been dark lights (darks for light frames) rather than light darks (which has an entirely different meaning). 

 

My reasoning is sort of the same as Jon's, but I reach the opposite conclusion. Exactly because it's an oxymoron to have a flat that is dark, you should call them dark flats. 

[Michael Covington]

Interesting.  What we have going on is that each of the 2 words could be an adjective (referring to a quality) or a noun (referring to a frame with that quality).  The second one needs to be a noun because this is English.  So the possible interpretations are:

N N -- "flat darks" -- flat-frame dark-frames -- dark frames that go with flat frames -- what Jon had in mind

 

Adj N -- "flat darks" -- flat dark-frames -- dark frames that are flat -- ??? not sure what it means

 

N N -- "dark flats" -- dark-frame flat-frames -- flat frames that go with dark frames --- ??? not what we want

 

Adj N -- "dark flats" -- dark flat-frames" -- flat frames that are dark -- what Stelios had in mind

 

So each of the 2 ways of saying it is defensible, with a different grammatical analysis.

 

Thank you all for your input!

[freestar8n]

Well the other day I was taking flat exposures and trying to get the settings right.  I started out with long exposure and got a really bright flat.  I said - "Woah - that is a bright flat!"  Then I went too far the other way and said, "Woah - that is a dark flat!"  Then I realized that if I took the flat exposures with no light at all, I would get completely dark flats - and I could use them to subtract off the dark current and bias present in the normally exposed flats.  So I now take a series of dark flats and they work great to subtract from the flat exposures to get the pure photon signal.

 

Later I was taking some darks to go with the lights and they looked really noisy and rough.  I wanted to make them look better but I was just stuck with what I had.  There was no way I could turn a normal dark into a flat dark.  So I gave up trying to make a flat dark.

 

Any grammar errors?

 

Jon - I'm still in the process of studying the camera and hope to post soon.  I see some anomalies but I don't know yet if they are a real factor.  I'll be able to say more when I have results.

 

Frank

[Michael Covington]

You (and Stelios) have almost convinced me to stick with "dark flat," which is what I was calling it before I asked the question.

A dark flat is like a flat, but darker. 

[Xshovelfighter]

Sorry to bring up an older thread. If using PI’s batch preprocessing script, would I load the dark flat frames into the bias category? There is no category for dark flats using BPP, but my understanding is that it is replacing the bias and serving the same function.

Thanks for any insight!

[Jon Rista]

Sorry to bring up an older thread. If using PI’s batch preprocessing script, would I load the dark flat frames into the bias category? There is no category for dark flats using BPP, but my understanding is that it is replacing the bias and serving the same function.

Thanks for any insight!

No, a dark flat is not a bias, it is a dark. You skip the master bias entirely (uncheck it), and put the mater dark flat (or master flat dark, whichever you prefer) into the master dark area in PI.

[Michael Covington]

 

Sorry to bring up an older thread. If using PI’s batch preprocessing script, would I load the dark flat frames into the bias category? There is no category for dark flats using BPP, but my understanding is that it is replacing the bias and serving the same function.

Thanks for any insight!

No, a dark flat is not a bias, it is a dark. You skip the master bias entirely (uncheck it), and put the mater dark flat (or master flat dark, whichever you prefer) into the master dark area in PI.

 

I disagree.  Dark flats are bias frames.  Do not put the flat dark in place of the dark -- it will result in grossly incorrect calibration.

 

I have compared bias frames and dark flats taken with my DSLR, and they are indistinguishable.  Like bias frames, dark flats are darks taken with an exposure short enough for exposure-time-dependent dark current to be negligible.

Arithmetic, in full, here:
http://www.covington...html#Arithmetic

 

I thought we had been through this, much earlier in the thread.  In PixInsight Batch Preprocessing, I put in dark flats where bias frames are requested.

[leveye]

I am an author myself and am trying to choose between two rival terms for use in something I'm writing.  I know that both terms are in wide use.  Merely finding one of them used somewhere doesn't settle the issue.

I didn't merely find it used just "someplace". It was written by the author on his DSS description page. Silliest thread ever.

[leveye]

 

 

Sorry to bring up an older thread. If using PI’s batch preprocessing script, would I load the dark flat frames into the bias category? There is no category for dark flats using BPP, but my understanding is that it is replacing the bias and serving the same function.

Thanks for any insight!

No, a dark flat is not a bias, it is a dark. You skip the master bias entirely (uncheck it), and put the mater dark flat (or master flat dark, whichever you prefer) into the master dark area in PI.

 

I disagree.  Dark flats are bias frames.  Do not put the flat dark in place of the dark -- it will result in grossly incorrect calibration.

 

I have compared bias frames and dark flats taken with my DSLR, and they are indistinguishable.  Like bias frames, dark flats are darks taken with an exposure short enough for exposure-time-dependent dark current to be negligible.

Arithmetic, in full, here:
http://www.covington...html#Arithmetic

 

I thought we had been through this, much earlier in the thread.  In PixInsight Batch Preprocessing, I put in dark flats where bias frames are requested.

 

In DSS you can use Bias frames,Flats with Dark flats or all three if you so wish. I see nowhere in the authors figures or instructions of (DSS at least) where you would use dark flats for bias frames. Not sure why dark flats would be used so in any another program either but I do admit I only use DSS for stacking. Hope you all can figure it out. Happy New Year.

 

http://deepskystacke...english/faq.htm

Edited by leveye, 31 December 2017 - 04:24 PM.

[Jon Rista]

 

 

Sorry to bring up an older thread. If using PI’s batch preprocessing script, would I load the dark flat frames into the bias category? There is no category for dark flats using BPP, but my understanding is that it is replacing the bias and serving the same function.

Thanks for any insight!

No, a dark flat is not a bias, it is a dark. You skip the master bias entirely (uncheck it), and put the mater dark flat (or master flat dark, whichever you prefer) into the master dark area in PI.

 

I disagree.  Dark flats are bias frames.  Do not put the flat dark in place of the dark -- it will result in grossly incorrect calibration.

 

I have compared bias frames and dark flats taken with my DSLR, and they are indistinguishable.  Like bias frames, dark flats are darks taken with an exposure short enough for exposure-time-dependent dark current to be negligible.

Arithmetic, in full, here:
http://www.covington...html#Arithmetic

 

I thought we had been through this, much earlier in the thread.  In PixInsight Batch Preprocessing, I put in dark flats where bias frames are requested.

 

Sorry, this is incorrect. The reason to get a dark flat is because there is non-trivial dark signal, such as amp glow, that must be removed. That is distinctly different than a bias frame, which is so short that they have no meaningful dark signal at all, not even from glows. If you have no dark signal, then it is not a dark flat, it is a bias. That is the distinction, and a critical one.

 

A dark flat is definitely a DARK frame, and most definitely NOT a bias frame. They should be treated as such, and handled appropriately.

 

So NO, do not put dark flats where bias frames are normally put, because by definition they contain dark signal! FTR, this is the glows that can appear in a half-second ASI1600 dark flat:

 

 

Exposure time has nothing to do with whether a frame is a dark or a bias. Dark signal has everything to do with whether a frame is a dark or a bias.

 

FTR, not every camera is the same. What happens with your DSLR is not the same as what happens with a CMOS astro camera. I've been using dark flats in a master dark flat and calibrating my flat frames with that master in the master dark area of PI's tools for well over a year, and it has always worked, and never produced "grossly incorrect" calibration. In fact, it did exactly what it was supposed to do: fix the glows that were resulting in two slightly darker bubbles in my flat-calibrated lights!! I would also offer that you could very well have forgotten to disable dark optimization in BPP, which would most likely, as it does with pretty much all cameras that have glows, resulted in an incorrect noise-evaluation based scale factor, resulting in incorrect correction.

Edited by Jon Rista, 31 December 2017 - 06:31 PM.