Can PI evaluate a flat frame's "flatness"?

[MineralMike]

I'm building an LED light box to make flat frames. I did this once in the past, and its flats compared favorably with dawn flats.  The new box is not as deep, and I'm experimenting on LED placement to get even illumination on the rear panel.

Does PI have a process that can inspect a flat frame taken with an unfocused DSLR (no telescope, no filters, not my astro camera), and report the flatness, e.g., with percent variation or similar?

I took an intentionally unevenly illuminated image with my DSLR, then opened it in PI. I ran the ABE process several; times with None, Subtraction, and Division target image correction. I did an STF on each image, and I can see some variation, as expected.

But that's all visual. Is visual inspection of a stretched image my best shot, or does PI have a way to report the flatness of the image?

Thanks for any advice.

--- Mike

[pfile]

there is a script called FlatContourPlot that's under Scripts > Image Analysis. you can give that a try. it should also have documentation (click the little page shown on the script window at the bottom)

rob

[Niall Saunders]

Also,

Try taking two sets of Flats where the only change is the orientation of the LightBox with respect to the optical axis (i.e. rotate the LighBox 90° clockwise, for example, and then shoot your second set of Flats).

Apply exactly the same PI processing to both ssets of images and see if you can tell that the seconf set of Flats had been 'rotated'. If you can see the rotation, then you have optical issues. It is then up to you to decide whether 'flat' is 'flat enough'.

Hope this helps.

[MineralMike]

there is a script called FlatContourPlot that's under Scripts > Image Analysis. you can give that a try. it should also have documentation (click the little page shown on the script window at the bottom)

rob

Thanks, Rob. That looks like it will help.

Also, Niall, I have a camera rotator, so it will be easy to do what you suggest. During actual imaging, I'll need to take flats at the camera angles used for the light frames.

Thanks again.

--- Mike

[rbotero]

Thanks, Rob. That looks like it will help.

Also, Niall, I have a camera rotator, so it will be easy to do what you suggest. During actual imaging, I'll need to take flats at the camera angles used for the light frames.

Thanks again.

--- Mike

Mike

Are you checking your new lightbox "flatness" or how well it corrects?  I believe what Niall is suggesting by rotating the lightbox (not the imaging train with your rotator) is to check how uniformly it illuminates.  By rotating the lightbox and subtracting one flat from the other you will see if there is a difference in illumination inside the lightbox.  Rotating your imaging train will necessarily need different flats for each angle of rotation.

Roberto

[Niall Saunders]

Hi Mike,

Yes, Roberto is absolutely correct. If you change ANY part of your imaging train during the Flats evaluation process, then you will need to 'start again'. After all, the exercise was not to demonstrate how much flat-correction you need, but whether the Flats that you propose using for image calibration are actually any good for the task that you need them for.

So, no, the only thing that can be changed is the rotational orientation of the light-box with respect to the primary optical axis of your imaging train.

How you achieve this really depends on whether you are 'hanging' the light-box vertically and pointing your OTA horizontally, or whether you are fixing the light-box to the end of the OTA in some way (for example, laying a light-box horizontally across the end of an OTA which is pointing vertically).

Give us, or remind us of, the details of your light-box and proposed installation, and we can try and help further.

Hope this helps.