Flats with IPAD screen

[Scooot]

I've created a plain white image on my ipad, which is illuminated and fills the screen when I open it. Is there any reason I couldn't hold this in front of my scope to take flats?

Here's a single image I took through the scope as an example.

[NigeB]

Hi Richard,

In my opinion:

I think the answer is yes...but. The iPad display is a non-lambertian source; the uniformity with viewing angle is pretty good (compared to old LCD display technology), so your results would probably be OK, but you can improve the situation by putting a sheet of white paper over the front. This acts as a diffuser and gets you closer to the characteristics of a true lambertian source.

I've done something similar in the past, though using an electroluminescent panel in place of an iPad / laptop screen, for assessing the gain uniformity of CCD cameras - but the panel was about the size of an iPad screen. In this case, I placed a thin opalescent glass diffuser over the front to get rid of very subtle small scale patterning in the panel. It was good enough to measure pixel-pixel gain variations.

So I suspect your iPad solution plus diffuser will be fine, as long as you can fill the entire aperture of your telescope with the source. So looking at your sig, I'd assume you're doing this with your FS-60 and not your larger telescopes - is that right?

One point - I downloaded your white panel image and according to my Gimp (!), all the pixels have the same value. Maybe it's the way the jpeg is processed by the SGL board system, but just to check: I assume you know that the flat should contain no saturated pixels?

Thanks

Nigel

[Scooot]

Hi Richard,

In my opinion:

I think the answer is yes...but. The iPad display is a non-lambertian source; the uniformity with viewing angle is pretty good (compared to old LCD display technology), so your results would probably be OK, but you can improve the situation by putting a sheet of white paper over the front. This acts as a diffuser and gets you closer to the characteristics of a true lambertian source.

I've done something similar in the past, though using an electroluminescent panel in place of an iPad / laptop screen, for assessing the gain uniformity of CCD cameras - but the panel was about the size of an iPad screen. In this case, I placed a thin opalescent glass diffuser over the front to get rid of very subtle small scale patterning in the panel. It was good enough to measure pixel-pixel gain variations.

So I suspect your iPad solution plus diffuser will be fine, as long as you can fill the entire aperture of your telescope with the source. So looking at your sig, I'd assume you're doing this with your FS-60 and not your larger telescopes - is that right?

One point - I downloaded your white panel image and according to my Gimp (!), all the pixels have the same value. Maybe it's the way the jpeg is processed by the SGL board system, but just to check: I assume you know that the flat should contain no saturated pixels?

Thanks

Nigel

Thanks Nigel. I was wondering about the screen, maybe if I just put a white cloth over it?

Yes i'd just use it with my fs60 or maybe just the camera sometimes, so would easily fill the aperture.

With that jpeg I just took a 5 second exposure through the scope, whilst it was lying on a table indoors to help explain my intentions.

I don't know much about flats other than I need to take them, and that I need to take them at the same exposure time and scope angle as the lights etc. If the pixels of that image were saturated, maybe its because my ipad screen was too bright. In the field so to speak, the exposure is likely to be longer,so even brighter, but a white cloth (like a cotton hanky) over it would diffuse it?

[Scooot]

Whoops, just re-read a bit about taking flats, not the same exposure time.

[NigeB]

Hi Richard,

That's right - matching exposure time isn't necessary, select whatever time gives you good signal levels but without saturation. I think white cloth would be fine if it is close weave, and free from any creases etc. Get your iron out!

I suspect you will probably need to change the colour on the screen from white to grey (maybe dark grey) to get the illumination levels low enough to prevent saturation even at very short exposure times. It'll be a good thing to try out on a cloudy night, and it seems that there are plenty of those coming up...

Nigel

[Davesellars]

I do this just opening a sketch pad on the iPad which has white "paper" and turning up the brightness to max.  I drape a cotton shirt over the scope to diffuse the light.  Seems to work OK!

[Scooot]

I just took these in AV mode. First without a cloth second with,

I can clearly see a difference in light levels around the edge, which is presumably what adding these to my images will help to remove

[NigeB]

Hi Richard

Loks good; yes, that's what you're after. The idea is to illuminate your optical system with a light field that is as close as possible to perfectly uniform. As the light passes through your optical system, there will be losses (due to e.g. vignetting, rings caused by dust on the sensor, etc). The slightly uneven illumination you've recorded on your CCD is effectively a map of that  non uniformity and loss. 

The same losses are present in your images. So when you divide your flat field into your image in your image processing software, those losses are essentially cancelled out, and you should see e.g., much improved background uniformity, and elimination of features that come from dust on the sensor. 

I've just taken the last flat field image you posted and stretched the histogram to provide maximum contrast - see below. You can see vignetting around the perimeter (always present - quite normal) plus the presence of a particle just below right of field centre (also quite normal - don't get the cleaning kit out!). So your raw images of e.g. star fields will be subject to the same effect. A 5th magnitude star in the middle of the field, will look brighter than a 5th magnitude star near the corners because the end-to-end sensitivity of your system is lower in those regions. But once you divide through with the flat field, they should appear to be of equal brightness.

Nigel

[Scooot]

Thank you Nigel, that's a great explanation and thanks for taking the time to analyse the image etc.. It's very helpful to get a better understanding of the process rather than doing it blindly.