New Flats Question :)

[fatboy1271]

I had some issues with PI, that turned out to be somewhat self-induced... Anyway, I'm able to get back to my Processing lessons!

My last session I took Flats three ways using NINA's Flat Wizard:
1 Double T-shirt
2 Single T-shirt
3 No T-shirt

First question is which of the three looks "best?" They are single Debayered images with Calibration only in BPP; meaning no ImageIntegration. Bias, Darks, and Flats applied.

Second question is how much should my Flats be doing for that beautiful dust mote at the bottom?

fat

[Baskevo]

I hope you don't mind me asking on your post, and you'll consider this a bump in the forum

What was your problem in PI? my flats on my camera and in APT look great, histogram is half-way, but look super dark in PI. Did you have that issue?

[Stuart]

Hi Fat
It doesn't matter what the screen stretch looks like. Just that it's about 50% of your max ADU as we've talked about before. I believe you apply the flats before debayering the lights. (Although I could be wrong on that last point--it's been years since I've dealt with OSC before)

And yes, the flat should deal with the dust mote--it should pretty much eliminate it.
Stu

[fatboy1271]

I hope you don't mind me asking on your post, and you'll consider this a bump in the forum

What was your problem in PI? my flats on my camera and in APT look great, histogram is half-way, but look super dark in PI. Did you have that issue?

Hey James, as you can see from Stu's comment this has been a contentious issue for me for quite a while... I can't get PI to give me the same Statistics readings that other get when they look at my CR2 files, so that just adds to my confusion.

Hi Fat
It doesn't matter what the screen stretch looks like. Just that it's about 50% of your max ADU as we've talked about before. I believe you apply the flats before debayering the lights. (Although I could be wrong on that last point--it's been years since I've dealt with OSC before)

And yes, the flat should deal with the dust mote--it should pretty much eliminate it.
Stu

Hey Stu... Sorry, that you once again have to say the same thing over again to me It's annoying, I know!

You are correct that the Flat Master is created in BPP before Debayering. Once again when I look at the Statistics I am not getting anything close to what you've seen with my past Flats.

If I stretch the Master Flats, just to see what comes up, I do see the very obvious gradient that I can see in my Lights; however, I never see the dust mote(s).

===

In the midst of typing this reply I did a search for "how to check flat file histogram." I found this page: https://www.myastroscience.com/proper-flats-with-dslr

I feel like this explains some things to me I wasn't getting! My only question to this author, which I posted, is that he suggests that Bias should be shot at the same ISO as Flats. Any thoughts on that?

[Stuart]

More importantly, does the flat work?

[Stuart]

And in terms of the statistic number, what ADU number do you get if you saturate the chip? Take that number, divide by 2, and that's what you're shooting for, more or less.

[fatboy1271]

And in terms of the statistic number, what ADU number do you get if you saturate the chip? Take that number, divide by 2, and that's what you're shooting for, more or less.

Hey Stu,

Sorry but what's the best way to saturate the chip? That will definitely be helpful to know

Here M45 from my last session. Flats look like they helped with the gradient, but the dust mote is quite prevalent. This not a full process and it's the best I've ever gotten from M45:
BPP: Darks, Bias, Flats, Lights
StarAlignment
ImageIntegration

[Mike_Lewis]

Stu helped me find my way through capturing good flats with my DSLR years ago (as well as a host of other issues) so I can fill this blank in for you. When I was dealing with this I captured a series of flats going from very short to ridiculously long, like several minutes, then evaluated the ADU from each exposure in a spreadsheet. It was obvious from the results that there was a point where the ADU value flattened out and did not increase regardless of an increase exposure time. That max value was my saturated chip value. Then I just found the exposure time that corresponded to as close as possible to half the saturated chip value and I had my target exposure time.

One thing that was a very positive change in my flats capture routine was finally spending the money and buying an electroluminescent panel. That put an end to spending time trying to find the correct exposure length to get to the correct ADU. WHile trying to find that value with natural lighting, IE sky flats, the brightness of the light source is changing. With an ELP you have one source value to deal with, you figure it out once and use the same value every time. It's nice and repeatable.

[Stuart]

What Mike said! :-)

[fatboy1271]

Stu helped me find my way through capturing good flats with my DSLR years ago (as well as a host of other issues) so I can fill this blank in for you. When I was dealing with this I captured a series of flats going from very short to ridiculously long, like several minutes, then evaluated the ADU from each exposure in a spreadsheet. It was obvious from the results that there was a point where the ADU value flattened out and did not increase regardless of an increase exposure time. That max value was my saturated chip value. Then I just found the exposure time that corresponded to as close as possible to half the saturated chip value and I had my target exposure time.

One thing that was a very positive change in my flats capture routine was finally spending the money and buying an electroluminescent panel. That put an end to spending time trying to find the correct exposure length to get to the correct ADU. WHile trying to find that value with natural lighting, IE sky flats, the brightness of the light source is changing. With an ELP you have one source value to deal with, you figure it out once and use the same value every time. It's nice and repeatable.

Hey Mike,

Thanks for that info! I purchased an LED light panel made for tracing things like artwork. It has variable brightness, but you have to hold the power button to "scroll" through the brightness levels. It's a nice feature but really limits you to knowing exactly what brightness you're at. If you hold long enough it goes full bright or full dim. Anything in the middle is a crapshoot. Other than that it seems to work well; however, who am I to judge when I have so many questions With that in mind, is this how I should approach finding these values?

Put t-shirt on OTA; with a light panel do I need the t-shirt?
Put light panel on OTA
Shoot exposures from very quick, I've been told to try to keep them at a minimum of 0.5 seconds, to very long
In PI look at the Statistics and record the Mean value
Once the Mean has flattened out that is my max value
Find as close as possible to half max value, which would be my desired Flat value

I can't tell you how excited I am by this!!! Something so seemingly simple shouldn't be so difficult. I can be a moron though, so there's that...

This is the light panel I purchased:

fat

[JayTee]

Hi Patrick,

I use an EL panel for my flats. It is quick, easy, accurate, repeatable, and efficient. Specifically, this is the one I use. https://www.ellumiglow.com/astrophotogr ... circle-kit

Cheers,
JT

[fatboy1271]

Hi Patrick,

I use an EL panel for my flats. It is quick, easy, accurate, repeatable, and efficient. Specifically, this is the one I use. https://www.ellumiglow.com/astrophotogr ... circle-kit

Cheers,
JT

Awesome, thanks JT!!!

[Mike_Lewis]

Hey Mike,

Thanks for that info! I purchased an LED light panel made for tracing things like artwork. It has variable brightness, but you have to hold the power button to "scroll" through the brightness levels. It's a nice feature but really limits you to knowing exactly what brightness you're at. If you hold long enough it goes full bright or full dim. Anything in the middle is a crapshoot. Other than that it seems to work well; however, who am I to judge when I have so many questions With that in mind, is this how I should approach finding these values?

Put t-shirt on OTA; with a light panel do I need the t-shirt?
Put light panel on OTA
Shoot exposures from very quick, I've been told to try to keep them at a minimum of 0.5 seconds, to very long
In PI look at the Statistics and record the Mean value
Once the Mean has flattened out that is my max value
Find as close as possible to half max value, which would be my desired Flat value

I can't tell you how excited I am by this!!! Something so seemingly simple shouldn't be so difficult. I can be a moron though, so there's that...

This is the light panel I purchased:

fat

Fat,

you do have some variables to work with there. I'd just do it empirically. Set the panel at the very repeatable Low setting and capture a series between .6 seconds and 2 minutes. Repeat the series with the panel set to High. Record all of the results. After all this new fangled digital film is so cheap. If you find that the Low setting doesn't come to a peak in 2 minutes then work with the High settings. If the High setting peaks before .6 seconds then work with putting different dimming media, like T-shirts and sheets of paper over the panel to reduce brightness. When you find the right combination be careful to have that exact same material or combination on hand to avoid any more serious frustration with this aspect of the hobby after you've collected a good set of data. After all there are so many more aspects to be frustrated by, you'll want to leave room for all the rest!

Good luck and clear skies!

[fatboy1271]

Awesome, thanks Mike! Yeah, I definitely need to have some frustration because you know, otherwise it just isn't a very fulfilling activity...

I will take to heart your suggestions. This sounds like a great to way to work this out!

[Baskevo]

Fat! I just read something that might be why I"m not getting good flats... You are much better at this than I am, so you've probably already done this, but just in case you and or someone as silly as me doesn't know:

Turn off the automatic sensor cleaning, so it doesn't clean the dust off in between subs and flats

[UlteriorModem]

I ended up building a light box with some tape led lights, and a dimmer. It runs on 12 volts. It was easy to build from foamboard some white translucent plexiglass and a bunch of hot glue. The box is about 18" 'deep' so that the tape light is a good distance from the plexiglass panel. I also added a piece of acrylic prismatic lens from an old light fixture a few inches below the tape lights to diffuse the light avoiding hot spots.

With that arrangement I can adjust the light level and get flats with good duration that are not saturated. Also if using NB filters you can turn it up some and not have to endure 12 second exposures

One of the problems I found with sky flats is that the duration was so short, in the order of 1/8 of a second, it was not really a good sample. I like the flat duration to be around 2-3 seconds.

[Mike_Lewis]

I also added a piece of acrylic prismatic lens from an old light fixture a few inches below the tape lights to diffuse the light avoiding hot spots.

The prismatic lens is a nice touch. A lighter weight solution might be using theatrical lighting diffusion media http://www.leefilters.com/lighting/diffusion-list.html. This stuff is light weight and very durable. If you look at the linked site consider that the diffusion patterns shown are for one light source. A flat light source like UM's has multiple lights so the diffusion should create a very uniform spread.

It's not all that difficult to acquire, just do a search for Theatrical Lighting Gel suppliers and you should be able to order single sheets online.

[Stuart]

I ended up building a light box with some tape led lights, and a dimmer. It runs on 12 volts. It was easy to build from foamboard some white translucent plexiglass and a bunch of hot glue. The box is about 18" 'deep' so that the tape light is a good distance from the plexiglass panel. I also added a piece of acrylic prismatic lens from an old light fixture a few inches below the tape lights to diffuse the light avoiding hot spots.

With that arrangement I can adjust the light level and get flats with good duration that are not saturated. Also if using NB filters you can turn it up some and not have to endure 12 second exposures

One of the problems I found with sky flats is that the duration was so short, in the order of 1/8 of a second, it was not really a good sample. I like the flat duration to be around 2-3 seconds.

Hey UM. I have a very simple solution--it's what I use. Put a couple sheets of plain white printer paper over your objective.

[UlteriorModem]

I also added a piece of acrylic prismatic lens from an old light fixture a few inches below the tape lights to diffuse the light avoiding hot spots.

The prismatic lens is a nice touch. A lighter weight solution might be using theatrical lighting diffusion media http://www.leefilters.com/lighting/diffusion-list.html. This stuff is light weight and very durable. If you look at the linked site consider that the diffusion patterns shown are for one light source. A flat light source like UM's has multiple lights so the diffusion should create a very uniform spread.

It's not all that difficult to acquire, just do a search for Theatrical Lighting Gel suppliers and you should be able to order single sheets online.

I actually got some of that stuff but never 'installed' it. Might have to get around to it. As it is right now the dimmer is barely turned on. If the dimmer is too low the led's will flicker and strobe.

[fatboy1271]

Fat! I just read something that might be why I"m not getting good flats... You are much better at this than I am, so you've probably already done this, but just in case you and or someone as silly as me doesn't know:

Turn off the automatic sensor cleaning, so it doesn't clean the dust off in between subs and flats

Ugh... Well, I never thought about this and yeah, of course it's turned on Thanks James! Always learning