is this a good flat example?

[Atreta]

Hi,

I've been avoiding doing flat ever since i began with AP because i thought it would be another hard thing to do, but now i want to learn how to make a proper flat frame.

this example below i made with apt and using a white t-shirt pointing at the ceiling. Would it be an ok flat frame to use? 

attached is also an apt stretched version of the flat frame. Also does it indicate a proper mirror alignment?

thanks again for all the help so far

[Demonperformer]

Looks reasonable to me - this histogram looks about right ... but my experience with flats is "mixed".

The most important thing to remember is that they must be taken using precisely the same setup as your lights. Unlike darks (with a temperature-regulated camera) you cannot build up a library that you can use repeatedly. Nothing should alter in the optical train between the lights and flats.

[Atreta]

42 minutes ago, Demonperformer said:

The most important thing to remember is that they must be taken using precisely the same setup as your lights.

Thank you, DP.

i'm planning to do the flat frames after every session. It is much easier and faster than I thought and well worth doing. 

[newbie alert]

I'm no expert on flats but I thought it needed even illumination..

[Demonperformer]

That is the aim of the white t-shirt (diffuser) pointing at a flat surface.

[carastro]

The stretched flat looks OK to me.  They are a must for me, I find it impossible to process an image without flats applied.

Carole 

[tooth_dr]

2 hours ago, newbie alert said:

I'm no expert on flats but I thought it needed even illumination..

This is what an evenly illuminated flat would look like!  It highlights the uneven illumination of the optics.

[newbie alert]

7 hours ago, tooth_dr said:

This is what an evenly illuminated flat would look like!  It highlights the uneven illumination of the optics.

Yes but the ceiling isn't being illuminated from anything other than natural daylight or unnatural lightbulb..surely both will cause shadows?

[Atreta]

12 hours ago, carastro said:

The stretched flat looks OK to me.  They are a must for me, I find it impossible to process an image without flats applied.

Carole 

Thanks, Carole. I hope it will make things easier to process. Tonight looks clear,  I'm just waiting for the sun to set to give it a try. 

4 hours ago, newbie alert said:

Yes but the ceiling isn't being illuminated from anything other than natural daylight or unnatural lightbulb..surely both will cause shadows?

Yes, there is a bit of shadow over the ceiling , i first though that it wouldn't work or it would make an uneven flat, but i think the white shirt spreads the light evenly as Demonperformer said. 

I used this site as a reference: https://astrobackyard.com/how-to-take-flat-frames

 

[wimvb]

The proof is in the pudding:

1. Stretch a calibrated (flat corrected) light frame untill you clearly see the background. There shouldn't be any visible dust motes, and the background should be uniform.

2. Use the master flat for calibration of a single flat frame (dark or bias subtracted). Stretch it. If the master flat works, the single flat frame shouldn't show any dust motes or vignetting.

[Atreta]

4 hours ago, wimvb said:

The proof is in the pudding:

1. Stretch a calibrated (flat corrected) light frame untill you clearly see the background. There shouldn't be any visible dust motes, and the background should be uniform.

2. Use the master flat for calibration of a single flat frame (dark or bias subtracted). Stretch it. If the master flat works, the single flat frame shouldn't show any dust motes or vignetting.

Thanks,  I'll do it the next time  I'm imaging. Tonight seeing is not very good so it's not worth setting up outside. 

 

[wimvb]

5 hours ago, Atreta said:

Thanks,  I'll do it the next time  I'm imaging. Tonight seeing is not very good so it's not worth setting up outside. 

 

You can do the test with data you already have. That's how I use cloudy nights.

[Atreta]

I don't know why, but flats yesterday didn't work at all. The stretch makes them almost pure white even though the not stretched ones look ok.  

Also is there another program that can check collimation? My trial of ccd inspector expired. 

Thanks for all the help. 

[wimvb]

3 hours ago, Atreta said:

The stretch makes them almost pure white even though the not stretched ones look ok. 

Flats, when stretched, should show the vignetting and dust motes. If you're in PixInsight, I would guess that you have a thin edge that is still black. This prevents the (non-clipping) stretch from showing the vignetting. It's quite easy to see if this is the case. Just crop a very thin edge from the flats you want to check, then do a new stretch. Don't save this flat as a master flat.

If an unstretched flat looks ok, and shows very little vignetting, this is good news. It means that your sensor's illumination is the way it should be: flat.

[Atreta]

1 hour ago, wimvb said:

Flats, when stretched, should show the vignetting and dust motes. If you're in PixInsight, I would guess that you have a thin edge that is still black. This prevents the (non-clipping) stretch from showing the vignetting. It's quite easy to see if this is the case. Just crop a very thin edge from the flats you want to check, then do a new stretch. Don't save this flat as a master flat.

If an unstretched flat looks ok, and shows very little vignetting, this is good news. It means that your sensor's illumination is the way it should be: flat.

Thanks a lot Wim,

i don't have pixinsight but i opened it with startools just to check it out with an autodevelop  and it shows more vignetting and the dust motes  than with apt.

this is a print screen of it and below a cr2 example of the same image:

edit: i made this one by laying the notebook screen over the telescope opening and the white t-shirt.

F_0196_ISO400_1-13s__25C.CR2

[Demonperformer]

I think that is just a brighter version of the one in the first post. Your first one had darker vignetting on the left side and that is still clearly there. The fainter vignetting is still just visible and (most obviously) the dust bunnies in the bottom left corner and in the lower half of the right hand side are still there in the same place. I would say that it might be worth shortening your exposures a bit when you use the notebook screen for your illumination - but I'm no expert on flats and others may have different advice.

[Atreta]

35 minutes ago, Demonperformer said:

I think that is just a brighter version of the one in the first post. Your first one had darker vignetting on the left side and that is still clearly there. The fainter vignetting is still just visible and (most obviously) the dust bunnies in the bottom left corner and in the lower half of the right hand side are still there in the same place. I would say that it might be worth shortening your exposures a bit when you use the notebook screen for your illumination - but I'm no expert on flats and others may have different advice.

you're right, the same artefacts are showing in both pictures but the last one is brighter.  i need to get my head around this, but it's nice to learn new things.

[Demonperformer]

AP is a very steep learning curve ... it really is a case of 'Enjoy the journey'. There are very few Damien Peachs around and some of us (e.g. me) are never going to achieve that level. But if you enjoy taking each step in that direction, there's a ton of fun to be had along the way.

[astrosathya]

The best flats I've taken have always been using the Av mode of my DSLR. I stretch a white vest over the front of the scope, point the scope to the Western sky at dawn (6 am ish) and shoot at the same ISO I used for the Lights. 

For my refractor I sometimes use my laptop, open a blank notepad file, point the scope flush to the screen and shoot Av mode flats. These don't work well sometimes.

Sky flats are always good.

[wimvb]

Here's my method (sorry for the large format. Can't edit on my mobile phone)

The light source is a FLOALT led panel from IKEA, at its dimmest setting. Ekos/kstars automates exposure time to get an average pixelvalue of 25000 (of max 65000). At lowest gain, this is 0.2 to 1.5 s, depending on filter.

[david_taurus83]

I use AV mode on my Canon for flats but I set the exposure value to +2. The histogram in APT isn't accurate. I've experimented with flats and inspected the results in the Iris software. With a fully saturated sensor, 30 seconds with a flats panel on max, Iris measures the max ADU as 15300. Using AV on the camera with exposure value at 0 the max ADU I could get was around 3000. By setting the exposure value to +2 I could get the ADU up to 6800/7000 which is just below half the full well capacity which I believe is recommended.  Not an expert on the subject but I struggled with flats at first and this method now works for me.

 

http://www.astronomyforum.net/astronomy-digital-cameras-forum/178452-dslr-histograms-flats-warning.html

[Atreta]

12 hours ago, astrosathya said:

For my refractor I sometimes use my laptop, open a blank notepad file, point the scope flush to the screen and shoot Av mode flats. These don't work well sometimes.

Thanks.

That's how i did the last time. 

8 hours ago, wimvb said:

The light source is a FLOALT led panel from IKEA, at its dimmest setting.

That's interesting, I'll try to find something similar here. 

4 hours ago, david_taurus83 said:

I use AV mode on my Canon for flats but I set the exposure value to +2. The histogram in APT isn't accurate. I've experimented with flats and inspected the results in the Iris software. With a fully saturated sensor, 30 seconds with a flats panel on max, Iris measures the max ADU as 15300. Using AV on the camera with exposure value at 0 the max ADU I could get was around 3000. By setting the exposure value to +2 I could get the ADU up to 6800/7000 which is just below half the full well capacity which I believe is recommended.  Not an expert on the subject but I struggled with flats at first and this method now works for me.

 

http://www.astronomyforum.net/astronomy-digital-cameras-forum/178452-dslr-histograms-flats-warning.html

I'm using it set to AV mode too,  have just downloaded iris to check the flats.   Gonna check that link. Thanks. 

[Atreta]

12 hours ago, Demonperformer said:

But if you enjoy taking each step in that direction, there's a ton of fun to be had along the way.

I do, but most of the time I'm too lazy

 

[Atreta]

Looking back i forgot to dim the laptop screen, i think that's why it looks much more different. 

[ollypenrice]

On 05/07/2018 at 08:01, newbie alert said:

I'm no expert on flats but I thought it needed even illumination..

The illumination in front of the scope must be even but if the resulting flat is even then you don't need it. In reality it never will be.

A good flat is not a flat which looks flat. A good flat is a flat which shows every irregularity of illumination in your optical system. My widefield rig flats have terribly dark corners, a bright middle and a smattering of dark dust bunnies here and there. For precisely this reason they are brilliant flats because they are telling the stacking software exactly what needs correcting - and it corrects accordingly.

Your system is going to have to be pretty disastrous if an unstretched flat doesn't look flat. Why do daytime photographers not need flats? Because they don't stretch their images. Why do we need them? Because we do. And as we do so we stretch the uneven field illumination inherent to any lens.

Atreta's flats look credible to me. Dark corners, not perfectly symmetrical, and a bright middle. They should work - but if they don't then the way they are taken will need to be refined. For reasons I have never understood I sometimes shoot flats which are not effective. I just shoot them again.

Olly