Planetary Histograms

[Mr Ron Palgrave]

I have a novice question: I normally run my capture videos through PIPP and then AS! 3 to end up with a quality stacked image. This image I then run through Registax, but, before wavelet sharpening, I chop off the Black end of the histogram in the range from 0 out to say 30 or 40. [Experience says that the background sky contributes in this region. ]. I find this step clarifies the basic image quite a lot and then permits more scope for sharpening. I stumbled across this effect while processing images of Uranus, but have found it applicable to other planets, to a greater or lesser extent

However, my aim is to produce useful images as opposed to “astro-art”. Therefore, I wonder, is this histogram chop a valid step in the process, or have I taken the wrong fork in the road?

Ron Palgrave

[David Arditti]

I’ve not heard of this technique before, but it’s an interesting idea. I’m not sure why it would have much effect on the sharpening, as you are just eliminating some random background noise. I’d be interested to see examples of images processed with and without this step.

David

Director, Equipment & Techniques Section

[Mr Ron Palgrave]

Thanks for your response. I attach a comparison for you .You can see that the limb darkening appears to improve. I am nervous that I am tampering with the image too much, but you can understand the temptation.

[Grant Privett]

I would imagine the impact of the noise varies depending upon how the wavelet transform is implemented in Registax. A large chunk of the image being replaced with a single value would certainly change the frequency spectrum and has reduced the total dynamic range within the image.

The pixels on the planet will still contain readout/sky noise, but that is a lot less obvious when on top of the brighter background provided by the planet.

My first inclination would be to wavelet first and apply any threshold afterward, but I’m probably over cautious.

[Robin Leadbeater]

Hi Ron,

I don’t have much experience in planetary imaging but in slit spectroscopy the sky background, measured above and below the target spectrum, is similarly subtracted from the spectrum image.  When this is done however it is important that the software preserves any negative pixel values that are generated (due to noise in the image). If the negative values are clipped to 0 you do not get the correct total flux values. I think the same will apply here.  Are the two images in your document before any sharpening ?  If so I would say there is clear signs of clipping. (You could perhaps check by taking a profile through the two images before sharpening . The brightness of the planet limb should merge smoothly into the background due to the finite resolution of the imaging system). It depends to what use the image is going to be put but if there is any clipping, I would expect any measurement of total flux or diameter for example not to be reliable.

Robin

[Mr Ron Palgrave]

Thanks for your comment Grant. I have quickly tried the sequence you suggest, that is chopping AFTER wavelet sharpening. My first impression is that it produces a more useful image. I will pursue this a little further. 

[Mr Ron Palgrave]

Thank you Robin. You understand my point that the chopping process might be detracting from the usefulness of the image, even though it might look “nicer”. Both images are sharpened. As you suggest, I will check the profile. To be frank, I have no illusions as to my competence at present. But that will come and I do not want my captures to be inherently flawed. Thanks for your help.

[David Arditti]

Yes I think it would generally be regarded as best practice to do any data summation (i.e. stacking) before any data chopping, as it is the summation stage that clarifies what is data and what is noise, and therefore doing it the other way round risks throwing away data.

In planetary imaging in twilight or daylight (with filters) we often chop off the sky background level to increase overall contrast, but I do this after sharpening.

But another (odd-sounding) technique which I think does have some legitimacy here is to attempt to guide the way Austostakkert! stacks the images by using as a reference frame an image that has been manipulated, by masking, to be the shape the planet is expected to be. I have seen another other observer get good results on Mercury by this method.

[Grant Privett]

So you give it an externally sourced template/reference to aim for? Isnt that dangerous?

Can understand running the process and getting an improved result and using what that processing generated as your template/reference for an iterative process.

But using imagery from some other source means the result is not led by the data itself.  

[Martin Lewis]

Hi Ron,

This is a step I often do to darken the background on planetary images including images of the Ice Giants and particularly images taken in daylight or bright twilight, often Mercury or Venus. I would always do this after wavelet processing as the application of the wavelets will change the histogram distribution quite markedly. You won’t lose planetary data unless it is something very dark like the Crepe ring on Saturn. Removal of the background like this is a simple way of boosting the contrast.

Martin

[Mr Ron Palgrave]

I took profiles across each image. The in-chopped image went smoothly to a small positive value whereas the chopped went to zero. No negatives, But then I noticed you stipulated the unsharpened image! I will have to do it again.

[Mr Ron Palgrave]

Thanks David, that looks interesting. But  I cannot [yet] see how I can force Autostakkert , against its will, to take an imposed image as the reference frame. That does not seem an option. Would I need to merge it into the PIPP stack first?

[Mr Ron Palgrave]

Thanks, you have mentioned the subjects that I have used the technique on. I have not applied it to anything else [yet]. Glad to see it is an acceptable step.

[Author]

Posts