Walking noise and EAA

[elpajare]

In the latest version there is an option to reduce walking noise. At EAA we generally use the software of each camera to capture the images and it only allows very basic modifications and dithering is not one of them.

As an alternative, Ivo proposes the walking noise option that appears in the latest version of ST. I would greatly appreciate a few words of guidance on the subject, on how this new tool works and what we can expect EAAs from it.

Thanks for the answers

[admin]

Would you be able to share a dataset with me that shows walking noise, so I better can demonstrate?

It is still experimental, however I have been getting good results on non-EAA datasets where no (or poor) dithering was performed.

The idea is to model and take into account the distinct noise signature (streaks) that walking noise leaves, by modifying the expected grain distribution the
be elongated accordingly.

E.g. this is a comparison between noise reduction (with identical settings) without walking noise modelling (left), and with walking noise modelling (right);

light-BINNING_M51_wnsbs.jpg (208.24 KiB) Viewed 4149 times

The difference is subtle but noticeable, particularly at 6 o'clock in the galaxy's outer arms.

[elpajare]

Thanks for the reply, Ivo. You are always around when there are doubts or problems, this is one of the best things about Starttools.

I am sending you a link to a 12x25 second photo without guidance with the IMX294 uncooled. This is the maximum exposure I do serve.

If you want to make an in-depth comment on this data in order to improve it, I will greatly appreciate it.



https://drive.google.com/file/d/1sEx4aK ... sp=sharing

[admin]

No problem! The dataset doesn't seem public. Can you kindly give me access?

Thank you!

[elpajare]

https://drive.google.com/file/d/1sEx4aK ... sp=sharing

Is this OK?

[admin]

Thank you, however this appears to be a JPEG. Could you kindly provide the unprocessed, linear TIFF or FITS file?

[elpajare]

I always use jpg. The tiff or fits files are too heavy for my simple computer's computing power (EAA!)

See what you can do with this jpg and if not don't worry

[admin]

Ok. There should not really be a difference between processing a JPEG or FITS/TIFF, except for final file size though?

Anyway, this is an example of the walking noise reduction in action;

NGC 6907 GX CAPRICORNUS 12X25 A25 M0 W6W RC8 + IMX294 + UVIR_ncomparison.jpg (478.43 KiB) Viewed 4089 times

Again, all other the settings were left to be identical just to show how Denoise handles walking noise when the feature is enabled. It can obviously be tweaked further.

The difference in the background noise signature is subtle, but it makes the difference between me being able to say "yep, that's walking noise - he didn't dither" and "oh, that's just noise". It is much less distracting, as humans are exceptionally good at seeing patterns.

Settings used Grain Size 10 pixels.
And for Walking Noise SIze 20 pixels, Walking Noise Angle 101 degrees.

[elpajare]

The difference is important! From now on I will use this tool, thank you very much again, Ivo.

The type of file I do not think influences at all with this type of simple photographs if we except the weight of the file and the difficulty of handling it on many computers. At least I honestly cannot find any difference when I have tried it.

In all my photos with this chip (IMX294C) there were tenuous bands, almost horizontal, that with this tool (Walking noise) were greatly attenuated and almost disappeared.

One last question, should I always use the same parameters for Walking Noise in each photo as long as I use the same chip?

Muchas gracias!

[admin]

Glad you find it useful! The orientation (angle) will change and the amount of noise will change as well, depending on how many exposures you stack. So it is best to test and use unique values for every image.

Clear skies!