ICX825 based camera test results

[jimthompson]

Greetings all,

Some of you may be aware that I have been picking away at a big project involving the systematic testing and comparison of many EAA cameras.  The scope of my project seems to be growing without bound as new camera models are released.  In an effort to keep up I have taken a divide and conquer approach, focusing on a subset of cameras at one time.  In this case I have been looking specifically at EAA cameras that use the Sony ICX825 sensor:  Atik Infinity, StarlightXpress Ultrastar and Mallincam StarVision.  I was very fortunate to have friends loan me the cameras needed to do the testing, to them I give a big thank you.

Below are links to my summary test report, and a comparison table.  Please feel free to have a look and let me know if you have any questions.

Best Regards,

Jim T.

http://karmalimbo.com/aro/reports/Test Report - 825 Sensor Camera Comparison_07Oct2016.pdf

http://karmalimbo.com/aro/reports/825 camera comparison table_07Oct2016.pdf

[Chris]

Thanks Jim  Very interesting results there, they all seem to have pros and cons! The infinity could be the best for beginners then if they can tweak the software to be more stable. The SX Ultrastar looks very good on most fronts other than noise, and the Mallincam has the bonus of planetary and lunar but doesn't appear to handle DSO's as well (my quick take on your results  ) 

[RobertI]

Very useful comparison and analysis Jim, thank you for taking the time to do this and share with us. The halos around the bright stars with the Ultrastar are interesting. I wonder if @Paul81, the author or Starlight Live software, has anything in the pipeline to fix this, and whether it might further improve other aspects of the Ultrastar's images?

[HiloDon]

Hi Rob,

I'm sure Paul is aware of this issue.  He did make some changes in v3.1 to address this, but apparently was not enough.  Note the third change in Paul's change log for v3.1.

Don

[Dom543]

Hi Jim,

Very interesting and useful study. You are doing a great service to the EAA community.

I just have one minor comment, or rather question. Does StarlightLive indeed run on the scaled-down iOS? Or does it require the full-fledged MacOS?  I am not an Apple user but I thought that imaging applications would not run on iPhones and other iOS devices.

Based on the images included in your study, and the ones posted by Martin, it seems clear that the Infinity applies a software smoothing. This makes its unstacked images look smoother and this is why the quality of their captures don't seem to improve proportionally with stacking. Whether to apply or not to apply smoothing is a design decision. And, in a sense, a design compromise. In contrary to stacking, smoothing in software doesn't improve the data on which the images are based. It alters the data by putting it through a low-pass filter. This cuts out changes on the fine scale that may be noise but may just as well be be fine detail. Smoothing, sometimes also called "noise reduction", produces a smoother looking image at the cost of losing fine detail. This is the reason why the double stars are not split on the Infinity image in Martin's post and this is probably also part of the reason, why Atik made sure that the Infinity software cannot be used with their astrophotograpy cameras. They wanted to make sure that people are not using the Infinity software to collect data meant for post-processing. Astrophotographers need to be in full control of what details they want to retain and what they are willing to discard.

Thanks again for the interesting and useful comparison study.

Clear Skies!  --Dom

[Paul81]

Great test Jim, very thorough and well presented.

I suspect the artefact on the Ultrastar-C images is the bayer to RGB conversion. The algorithm is one I have used for daytime video processing so might need some tweaks to co-efficients for astronomical usage, and is likely to get effected by noise in the image.

SL doesn't run on iOS, it is desktop OSX. I could make it run on iOS but access to a USB camera would be a problem.

[jimthompson]

Thanks Paul, the iOS entry in my table is a typo.

Regards,

Jim T.

[Carl Reade]

Interesting comparison. As an Infinity owner there is a bit of softness to the image I would agree. I have to say never had an issue with the stability of the software it has worked fine for me. The software has just been updated which I have yet to try weather pending.

Have to say it's a great time for the hobby.?