I shared this in an earlier thread, but here is a simple method (courtesy of John Upton) to evaluate uniformity of your flat panel:
Flat Frame Evaluation in PixInsight:
- Set up telescope and camera as when taking Flat frames.
- Shoot 10 Flat frames with similar median brightness at about 30,000 ADU. (Set #1)
- Rotate the Flat light source by 90° leaving the camera and telescope in the same position.
- Shoot 10 more Flat frames with similar median brightness at about 30,000 ADU. (Set #2)
- In PixInsight, integrate Set #1 using Image_Integration as you would for creating a Master Flat.
- Name the Set #1 integrated result Flat_00_Deg.
- In PixInsight, integrate Set #2 using Image_Integration as you would for creating a Master Flat.
- Name the Set #2 integrated result Flat_90_Deg.
- Use Pixel_Math to divide one integrated result by the other and normalize it:
Flat_00_Deg / Flat_90_Deg / 2.0
Be sure to select Create New Image and select the output Sample Format to be 32 bit Floating Point.
- Name this new image Flat_Diff.
- Run the Flat_Contour_Plot script on the Flat_Diff frame.
Select the script parameters as: 50, 2.5, 15, 50
- On the resulting Contour Plot, note the numbers shown at the top (Max) and bottom (Min) along the right side.
- Calculate the total non-uniformity in the Flat as:
total_non-uniformity = (Max - Min) / ((Max + Min) / 2)
- You can optionally divide this by 2 to get the ± deviation from the mean value.
You are looking for uniformity numbers under 10%, preferably under 5%. The Spike A Flat claims numbers lower than 1%; unfortunately, they are now exiting the astro business so the panel is no longer available new. The Gerd Neuman Aurora flat panel, which is an EL panel, has also been evaluated using this method and is ~1%. I don't know the metrics for other panels.
The general challenge with LED based panels that are edge lit is that you need an engineered light guide, reflector sheet, and good diffusers to achieve uniformity. Simple tracing pads would not bother with this since it is overdesigning for the intended application.
)
