Flats - some clarification please

[IvanT]

Hi Folks,

I've been reading about flat frames and there seems to be a difference of opinion that hopefully one of you nice folk will clarify for me.

Flats - Some people claim that flat frames can be taken any time and do not need to be at the same temp as the lights. Is that correct? My understanding is that a flat should be taken at low ISO, fast shutter speed and can be taken any time, i.e. temp is not important but ensuring no change in the optical path is critical. So as I set-up & tear down each night, I'd need to take my flats after each imaging session without changing anything in the optical path?

Bias - Am I correct in saying that a stock of bias frames can also be taken any time and are simply acquired with the camera cover on, at varying ISO settings and fastest shutter speed?

Darks - Need to be taken at the same temp, ISO & duration as the lights. In the absence of a cooled CCD that means AFTER the imaging session?

I have one of those light panels for the flats and image at prime focus on my SWE 200P with a coma corrector and LP filter using a modded 1000d.

Apologies if it's already been covered, I did search but am getting myself confused with all the "options".

Thanks in advance.

[Psychobilly]

DSS (if thats what your using) will throw up a warning if the ISO of your flats doesn't match that of the lights...

I use the same ISO for everything..

Temp matching - just for the darks and lights...

Peter...

[Bizibilder]

Take everything at the same ISO.

Darks must be at the same temperature BUT you can build up a "bank" of them. Take a series for each of your commonly used exposure times for each temperature and save them. I take every 4-5°C (and use the nearest), 1min, 2min and 5min subs. Gives the camera something to do on cloudy nights!! (Mine are taken in the garage as it is more convenient than the observatory - I use a remote timer but if I didn't have that I would set up the laptop to do the job).

Flats are taken immediately after the "light" frames without moving the scope or changing the optical path in any way - DON'T switch off the camera or the vibrating cleaner will operate and shake off any dust! They are quite short exposures so don't take long - just set the camera to "AV" mode and click away.

I usually take around 24 each for darks and flats.

I may be wrong, but I don't take bias frames as I believe the data is "contained" within the dark frames - But someone wiser may well know better than me on that one.

[IvanT]

I take every 4-5°C (and use the nearest), 1min, 2min and 5min subs.

I guess that's only an option with a cooled CCD that has temp set point, if so then darks would need to be taken after lights on the same night?

[Psychobilly]

A lot of people "comandeer" the fridge to do the runs I used to use a cool box when I was doing "thermal" tests... the camera will take a while to change temp and you can use software to sort the darks by temperature...

So even though the temp is changing over time if you do enough runs you will eventaully build up a "library" of darks...

Here's a link to the software a lot of people are using

http://darkframe.net/~cdndob/downloads/darklibrary_v1_7.zip

Peter...

[IvanT]

Thanks for that Peter, should have thought of the fridge myself Useful link too, much appreciated.

[Psychobilly]

NO worries...

If you can pull together all the darks you have to date they can form the basis of the library the softare will go through them and "sort" them...

I found the cool box easier.. I used to freeze down freezer packs and bottles of water and place them under a few layers of paper .. The fridge was too busy with people after food, milk etc ...

Have fun

Peter...

[Bizibilder]

Sorry - I didn't make it clear - I take my darks over time so that I slowly build up the ones I need. Cloudy winter nights are used for low temperature darks and Summer nights for higher temperature. You soon build up the ones that you need. I don't keep them for ever - just retake them continuously over the course of the year. You can't just keep using the same ones as the camera sensor changes with time (eg pixels may fail on the sensor).

Just to keep my sanity I normally use just ISO800 for deep sky photos - otherwise I would need darks at every ISO as well!!

[IvanT]

Ok, think I know where I've been going wrong, and when I say wrong, I mean WAY wrong. I was taking the flat frames using bulb, same ISO as light and (wait for it, bulb was your clue), yes you guessed it, the same blumming duration as the light frame. Over saturated I hear you say

This is a test flat using Av 1/200, ISO 800, light panel on the end of the scope with my modded 1000d (LP filter & coma corrector in place). I found that 1/200 was the closest setting to get the exposure histogram in the middle of the graph. Does it look anything like how you guys would expect a flat to appear?

[ollypenrice]

I can't see any need to worry over where the tube is pointing when you take your flats, which should be one third to two thirds of the way to saturation. Look at the histogram. But nothing in the optical path must change. (The outside world is not in the optical path which is why you can point at the zenith for simplicity when placing your panel on the OTA.)

Flats need calibrating but the best way to do this is to take bias frames and treat these as darks for flats. Don't use bias frames in any other capacity because you have them contained in your darks. Bias do need to be taken at the same temp as anything you intend to calibrate with them.

The trouble with DSS is that it asks for all of these things and you don't know (or I don't) what it is going to do with them all. I use AstroArt where I have control over each step.

Olly

[RobH]

Bias do need to be taken at the same temp as anything you intend to calibrate with them.

Olly

Although that's right, in practice it doesn't matter as the exposure is so short that thermal noise isn't an issue.

Rob.

[ollypenrice]

Although that's right, in practice it doesn't matter as the exposure is so short that thermal noise isn't an issue.

Rob.

Would this also be true for DSLRs though? I don't know, hence my caution.

Olly

[IvanT]

Would this also be true for DSLRs though? I don't know, hence my caution.

Olly

I'd have thought it would, i.e. wouldn't the chip still be hot when taking calibration frames at the end of a long run?

PS thanks for resisting the temptation to poke fun at my over saturated flats guys :-)

Any comments on the sample flat?

Sent from my iPhone using Tapatalk

[RobH]

Would this also be true for DSLRs though? I don't know, hence my caution.

Olly

I can't see it being any different than with CCD's Olly....you're looking at minimum exposure time, and it would also mean that you need to build a temperature matched bias library, which I've never heard of anyone doing.

Rob.

[RobH]

I'd have thought it would, i.e. wouldn't the chip still be hot when taking calibration frames at the end of a long run?

Ivan, you don't need to do bias at the end of a run. You can do them any time and they will be good for at least 6 months.....after a while the characteristics of the chip change a bit so it's worth redoing them.

Rob.

[daz]

Looking at the flat, it looks too "clean". I would expect to see dust do-nuts on it at the very least - this has nothing!

[ollypenrice]

As above, suspiciously clean! What happens if you drop the screen brightness of this flat?

Olly

[IvanT]

Hmm, I'll try increasing the exposure time from 1/200 to see if that makes a difference. The image was the jpg taken straight from the camera. I used that setting as I thought the aim was to have an exposure that placed the histogram roughly in the middle.

Sent from my iPhone using Tapatalk

[ollypenrice]

It is, but the in camera screen stretch might be masking contrast, I don't know.

Olly

[RobH]

I opened it up in PS, and the histogram peak is fairly central.

I did a heavy contrast stretch but it's falt as a pancake....not a single blemish.

You either have the cleanest system in the known universe, or something's not quite right

Rob

[IvanT]

Well I know for a fact there is dust on the coma corrector. I did clean the primary and secondary end of last year and haven't had the scope out very much so I'd expect to see at least the dust. I'm simply placing the light panel on the end of the scope and taking the shot.

Just in case I'm missing something, light path is modded 1000d prime focused to coma corrector, light pollution filter, secondary, primary, then light panel. I redid the copulation yesterday and performed a star test last-night and it's spot on, so I can only guess the fault lies in the camera settings, Av ISO 800 at 1/200 WB set to tungsten I think.

As a flat, does it seems to you guys much like the flats you'd take in regard to brightness, it was a lot darker than I was expecting. But then my only experience of flats up to that point was of the completely white over exposed kind.

Sent from my iPhone using Tapatalk

[RobH]

I redid the copulation yesterday and performed a star test last-night and it's spot on

That must have been the most fun anyone's ever had doing a star test :D:D

[IvanT]

Lol, don't you just love iPhone auto-correct, now that's one to add to the list.

Sent from my iPhone using Tapatalk

[ollypenrice]

That must have been the most fun anyone's ever had doing a star test :D:D

I reckon I'd better re-do mine too! Hey, great, I 've got six telescopes!!!!

Olly

[IvanT]

But nothing in the optical path must change.

It just occurred to me that part of my problem may be that I remove the due shield before placing the light panel on the end of the OTA. Does that constitute a change to the light path?

Could the A3 EL Panel I'm using be simply too bright and need toned down somehow?