StarTools' pervasive "Tracking" data mining feature is responsible for the markedly improved results compared to other traditional software.
As opposed to all other astronomical image processing software packages that are currently on the market, StarTools takes a completely different approach to processing. Rather than the traditional approach of having an application consisting of many granular algorithms, steps and filters that are carried out in sequence, StarTools is akin to a living and breathing organism; everything is connected and has its place, forming a whole that is greater than the sum of its parts.
Your data remains in a super position of states, being simultaneously linear and non-linear, deconvolved and not deconvolved, colour calibrated an not colour calibrated, etc. This allows StarTools to consult the data in its most suitable, unadulterated state for the task at hand.
Meanwhile StarTools observes how you stretch your signal and meticulously keeps track of visible noise propagation, levels and processing sequences throughout your processing session. It does all of this in the background, and without bothering you.
By data mining these two facets of image processing, StarTools is able to accomplish something quite remarkable; StarTools is able to effortlessly track and calculate cause and effect when making changes to any of the different states. It's a like time travel; with StarTools you can change steps you took in the past, to affect how the image looks in the present and future.
So, what does this mean for your image processing?
Firstly, it means that you are no longer beholden to the sequence in which you processed your image. Mathematically correct deconvolution after stretching (e.g. deconvolution of non-linear data)? No problem! StarTools knows how to reverse your stretching, apply deconvolution and reapply your stretching. Correct linear colour calibration of heavily processed data? No problem! StarTools will know how to completely negate the adverse effects of luminance manipulation and recover true colours.
It's like magic, except it isn't; it all boils down to StarTools simply being smarter with your hard won data.
You don't need to worry any longer about the correct sequence of operations on your data. The notion of linear data versus non-linear data has been abstracted away completely. If you're a beginner and never even knew about the requirement of 'old' software to keep data linear for select operations, now you don't even have to worry about it. If you are an image processing veteran, you can now do things that would otherwise be impossible, all without having to bother with sub-optimal crutches like screen stretches and the like.
The Tracking data mining feature breathes new life into old tools like deconvolution, wavelet sharpening and noise reduction. By infusing such algorithms with per-pixel accurate statistics about detail, noise and historical pixel values, these tools have gained a completely new dimension when it comes to signal preservation and noise suppression.
Noise reduction in StarTools is arguably the Tracking feature's Pièce de résistance. By postponing noise reduction to the last possible moment (so that Tracking has had the longest opportunity to data mine the data's evolution and your actions), StarTools is able to provide noise reduction that is unsurpassed in its accuracy and effectiveness, all without the need for local supports or luminance masks. The latter sub-optimal crutches have simply been made obsolete due to the availability of something better; true, objective and accurate statistics on each and every pixel in the image.
Fidelity and signal preservation is helped immensely by avoiding compounding rounding errors and (user induced) "overcooking" of images. Since StarTools always selects the correct super position state of your data, depending on the requirements of the algorithms at hand, the input (source) is always the cleanest, most unadulterated version of your data possible. Results are predictable and almost always useful; no longer is the result solely dependent on what the previous algorithm or filter generated as its output. Tracking helps with promoting a feeling of 'closure' and prevents endless cycles of applying filter-upon-filter.
Finally, thanks to Tracking's data mining, all modules "talk" to each other and are aware of what all the other modules have done to your signal prior. 'Old' software effectively sees you specifying the same constraints, thresholds, regularization amounts, kernel sizes and local supports over and over again because the individual algorithms have no idea what you specified before in the previous algorithm (which invariably needs similar input), nor does such 'old' software have any idea about the characteristics of your data or what an appropriate baseline would be for the algorithm's settings. Not so in StarTools; instead asking you to 'pick a number', where possible StarTools asks you to pick a deviation from what it thinks is a good baseline (based on prior input and Tracking data mining statistics). The result is that most modules in StarTools tend to come up with usable and reasonable default results that merely need tweaking to achieve your artistic vision for your data.
In fact, this behaviour greatly reduces the number of clicks and parameters settings that you need to make. It eliminates guesswork for parameters that can already be safely, reliably and objectively derived from the data or prior input without additional human (i.e. your) intervention. Not only is Tracking data mining a great help, it's also a great time saver. The time spent processing an image in StarTools is measured in minutes, not hours.