I cant find DBE controls...

[gpohls]

I am very eager to learn PixInsight....I have heard many good things...Although I am a MaximDL user I must confess PixInsight might be very powerful...but it is very user unfriendly....

I have past 2 hours and I cant find the DBE controls....I am severely frustrated..... :evil:

The tutorial doesnot resemble either downloadable versions in the website....Can someone help?

[Juan Conejero]

Hello,

I'll answer your question about DBE at the end of this reply. But let me talk about PixInsight's user interface first.

What you are experiencing is quite frequent among new PixInsight users when they face the application for the first time. This is because PixInsight implements a user interface paradigm that is quite different from most interfaces used by imaging applications. It's just that PixInsight has a learning curve that new users can't avoid. But the benefits are well worth the effort  

The traditional interface that you are accustomed to does not separate the objects you are manipulating (images in this case) from the actions that you are applying to them (processing routines and procedures). For example, when you apply a levels adjustment, a dialog box opens and you have to either complete or cancel the levels adjustment to continue with your workflow. This is known as a document-oriented user interface.

With PixInsight, there is a perfect isolation between images and processes by design. You can define virtually any process in PixInsight without having an image open, because processes are self-contained and self-defined in PixInsight (we often refer to this as process encapsulation). An example: you can define a histogram transform (the counterpart to levels in other applications), while you are modifying a mask to protect certain image areas, watching how the mask is acting in real time, inspecting image statistics and how they change as you apply the histogram transform, and even applying other processes to see how they interact all together. You can store, edit, reorder, modify and reuse any processing sequence without limits across different processing projects (with ProcessContainer objects). This is what we call an object-oriented user interface.

In my opinion, PixInsight's object-oriented user interface is much more powerful and versatile than common document-oriented interfaces. Most users report this when they break the initial effort required to feel comfortable with the application. Our implementation is also a fresh, innovative approach to the advanced image processing workflow. Process icons and process containers are only two obvious elements of this new paradigm. However, being so different and flexible, our interface has a learning curve, as I said before.

The Dynamic Background Extraction (DBE) process still has not been implemented in the new PixInsight Standard platform. We'll publish a new (and tremendously improved) version very soon, but for now, you can find DBE in the latest PixInsight Limited Edition (LE) version. You can download LE here:

http://pleiades-astrophoto.com/download/LE/index.html

Please download LE version 1.0.1.141, and follow the instructions to upgrade to 1.0.1.142. Let me know if you have any problem.

Thank you

[gpohls]

Juan,

I agree...PixInsight  is one of the most (if not the most) powerful tool available to us amateurs.(Judging by the results)

That is why I am spending days trying to learn it. I hope I can overcome all these little problems and be a regular user soon.

I am very excited and I congratulate you for doing such a high quality product. I imagine that there has to be a bit more compromise in the interphase appraoch to make it a bit more friendlier so the program becomes comercially a sucess.

"Juan, felicidades.....un abrazo y gracias por este gran programa que estoy seguro que con mucha paciencia y tu ayuda aprendere y sera mi herramienta preferida"

Saludos

[gpohls]

EXCELLENT RESULTS!!!!!!!!!!!!!!!!

DBE is really powerful...I am in awe

Great program...I need to learn more....

I have to ask maybe another question though...I cant seem to get a grip on the histogram tutorial...for some reason it is far better to worke on a already streched image than to try to do it from scratch from an image with no scretching right after stacking.

Why is this?

[Tim Kerr]

Hello,


The Dynamic Background Extraction (DBE) process still has not been implemented in the new Pix Insight Standard platform. We'll publish a new (and tremendously improved) version very soon, but for now, you can find DBE in the latest Pix Insight Limited Edition (LE) version. You can download LE here:

http://pleiades-astrophoto.com/download/LE/index.html

Please download LE version 1.0.1.141, and follow the instructions to upgrade to 1.0.1.142. Let me know if you have any problem.

Thank you

Ok,  I had the newest version (1.0.2.143) installed on my computer. Then I heard about DBE. So I came here to find out more about it. Reading your above post leaves me with the assumption that the newest version does not support DBE as of yet.  I uninstalled the version which i had installed on my PC and cleaned up the registry then downloaded the older Version (1.0.1.141). I Installed that version and I have found no instructions on upgrading it to 1.0.1.142. Where might they be located? and where might version 1.0.1.142 be located and available for download? There are only two versions available for download at the link provided above.  Those are versions 1.0.1.141 and  1.0.2.143 which do not have DBE.

Thank you

[Tim Kerr]

Never mind.  I un-installed everthing and re instaled it from scratch. I have it now.

Have A Nice ________

[gpohls]

Let me ask another question...in the DBE tutorial it suggest using the SUB operand (inside pixel math) to remove the background noise...


I find using DIFF (absolute value) is a better...any reasons or I am not getting here something?

[Carlos Milovic]

Well, there is a good reason to use the Subtraction operand instead of Absolute Difference. If you subtract, you get a flat background, keeping at the same time the pixel distribution in the neighborhood. It means, that darker features than the background will remain darker (for example, noise, or other real features, like a siluette, etc). You'll see that the final histogram of the image will show a sharper gaussian like curve, but the  background mean value will be "far" from the zero value, if you propperly rescaled the result, or normalizated the subtraction (by adding a constant value, ussually the background's median).
If you use the Absolute Difference, by the other hand, is like mirroing the pixel values arround the neighborhood's background value. That means, that darker features will become brighter features. If a pixel value was near the background value (the value of the calibration image) it will become an almost zero valued pixel. From an strictly point of view, you just destroyed the original data, but since you may not be interested in scientific purposes, you just may discover that working with the image histogram will be harder, and the overall brightness of the background will be too dark. If you look at the histogran, you should discover that a great fraction of it's pixels has a "0" value, total black, wich is not our desire.

Hope this helps...