Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - blueridgeDSIA

Pages: [1]
1
Wish List / Read Write speeds
« on: 2018 July 19 10:39:01 »
Im not sure if there is a bottleneck somewhere in my system but I would really like Pi to push my system at least to 75% when it is performing a process.  It does a good job of this during image registration and integration but when I do a batch file conversion for instance or an image container/process container batch process on lets say 200 dslr images, I find that Pixinisght is not pushing my system very hard and it seems to take longer than necessary to complete an operation. 

For example today I converted a big stack of RGB images to Lum and my 960evo m.2 ssd has very high read wrote speeds, but Pi was utlizing less than 25% of the write speed, the cpu and ram were not maxed out.  I dont see a bottle neck anywhere, is there something in the software that limits read/writespeeds?  Could this be user adjusted in Windows?


2
General / Help! Can no longer open .CR2 in Pi
« on: 2018 May 26 19:28:07 »
Im not sure what has happened but I cannot get Pi to see .Cr2 files any longer.  It is not an option in the files list when I go to open them either.
It was working fine up until about an hour ago, Im not sure what is wrong.

I recently installed the Raw file format support module version 1.5.0.397

Package type : Module
File name    : 20180205-windows-x64-module.zip
Release date : 2018-02-05 00:00:00
Install date : 2018-05-23 17:56:15
This update installs a development version of the new RAW file format support module for PixInsight 1.8.5. RAW implements digital camera raw format support based on the LibRaw open-source library. The new RAW module replaces the old DSLR_RAW module, which was based on the (now discontinued) dcraw program.
Copyright (c) 2003-2018 Pleiades Astrophoto S.L.

3
Bug Reports / Image integration
« on: 2018 February 05 13:11:32 »
Hello, I am trying to integrate a stack of data and it keeps failing.  I am not sure what the proper way is to share this bug, if it is a bug.  2 days ago I registered the data with frame adaptation selected in image registration, there is a slight overlap in these two data sets.  Today I re registered all of the data without using frame adaptation to see how the integrated image would look by comparison.  For some reason the data will no longer integrate.  It does not matter if I try to use an ROI or the entire image, image integration crashes every time.  I am getting this message below. Let me know if I need to pull more info from anywhere else. 



A:/M81 and M82 master/t2i plus t5/IMG_1502_c_d_r.xisf
Loading image: w=5695 h=4038 n=3 RGB Float32
Computing image statistics:

* Waiting for running tasks to terminate...
*** Error:
*** Error: Unknown exception
<* failed *>

Preferences: Global context
1.018 ms

4
General / RAM Bottleneck Win10 Image integration
« on: 2018 January 16 20:51:19 »
I am working a large data set and I am troubled because I have a custom page file setup on one of my SSD's for when I run out of RAM, which is not working for some reason.   
I am using image integration right now and it appears that the ram is bottle necking, Pi has been integrating 450 dslr light frame subs now for 3 hours.   
I need to figure out what the optimal settings are for my machine in a win10 environment.  maybe I have something configured incorrectly?
My imaging equipment consists of multiple dslr's and I typically end up with large data sets, calibrated outputs approach 500gb.

Whenever this gets finished I would like to make sure I have everything optimized to handle workflows of this volume.
Can anyone offer me a suggestion on how to improve performance?  I did some some screen shots of the current bottlenecking status if they might help.

 

5
General / Understanding Local Normalization Jargon
« on: 2017 November 28 19:44:04 »
Hello I am trying to utilize local normalization in my process flow.  The "no scale component" check box has a hint which says that it is to limit local normalization to correction of additive gradients exclusively.  Does this refer to clouds passing over by chance, which might add a gradient?   

6
Wish List / FWHM Matching star shrink and deconvolution
« on: 2017 November 28 11:07:56 »
I had an idea last night that may be plausible by someone capable of writing scripts for pi, and maybe this already exists but I am too novice to realize it. 
When you image for multiple nights your FWHM can change and your image as a result softens or gets much sharper and cleaner. 

If there was a tool that allowed you to compare and analyze images from different nights and apply a star shrink and deconvolution in an effort to match data by its fwhm from other nights where the same target is imaged, lets say these images are fully calibrated stacked light frames, you could then create a mosaic from this data with very similar star shape/size profiles and dso clarity.

 Now into murky water, if you could batch apply this fwhm based shrink and convolution matching to calibrated light frames with poor fwhm and integrate them with calibrated data which had optimal fwhm, this might offer better stacking results.  I am not sure if this could be designed to keep the data linear or not which is why I think it is murky waters.  None the less I think it would be great to take my 2.3 FWHM data on say Andromeda and combine it with 1.3 FWHM data from the year before and have two sets of subs which could integrate a much deeper image statistically without sacrificing as much star quality.  I assume that the larger data set of the two nights and registration frame selected would have some sort of effect of how the final integration would accept or reject the tighter stars which is why i began to ponder the possibility of such a tool.  Anyways, thanks for listening to the idea, I hope I don't sound like a crazy person.   

7
Image Processing Challenges / Registration without cropping
« on: 2017 November 26 13:25:40 »
Hello, Im trying to find out if anyone has methods for registering data without cropping.  LEts take an example scenario where you use two cameras and 2 scopes side by side and image the same target two nights in a row.  The images do not perfect;y align and registration wants to crop this perfectly good data at the resolution of a single frame. 

Right now the way I am getting around this is by calibrating all of my data and then making a mosaic from several images from camera 1 on night one, and a second mosaic from teh images from camera one on night two.  I then merge these two mosaics.   I then do the same thing with the data in my second camera.  After that I merge the data from both cameras and create a master mosaic from all of the data. I should say that these mosaics are created using the Register/Union Mosaic tool in Star Alignment. I simply keep selecting my main image as a refrence and adding to that to build the master mosaic larger and larger.
 Finally when I have built a master mosaic from 2 cameras and 2 nights of imaging, I register all of my data to this master mosaic.  I then take this data into image integration and use linear fits to stack everything.  This has proved to work extremely well, I have done it 4 times and there have been no seams between the images, I guess because everything is integrating at once and rejecting outliers on a massive scale.  I should also mention that I am typically working with 200-400 light frames when I do this.

Now obviously this is a seriously time intensive process and alot of errors can cause you to go back and meticulously track down flaws.  It works but I am wondering if I am going about this the wrong way?  Are there any other suggestions for this type of work flow?  I know back when I used Deep Sky Stacker, Merge Mosiac was a stacking option instead of selecting a reference image.  I am knee deep in this process right now waiting on images to stack so I figured I might as well get on here and ask around.  thanks

8
Hello I would really appreciate if anyone could make an attempt at helping me restore the stars in this Luminance stack from over the weekend. 
I used a monochrome converted DSLR to collect 360 light frames, I applied about 60 darks, 80 flats, and a large master bias.  The image calibrated well and is fully processable.
The problem is that I used an Astronomik ClS-CCD filter which I had in the camera already before I got to the dark site.  I did not check the filter and it had become partially unseated unfortunately.
This caused a strange thing to happen with my stars, the bright stars are fine but the faint stars have a bright central dot, it is hard to explain but if you look at the attached image file you will see what I mean, look to the left of the center of the image in the star field.

  I know that there has to be a good way to repair these stars via morphological transformation or some other technique.  I am really hoping some of your more advanced pixinsight users can share some creative approaches to dealing with this issue.  Please feel free to download and post your results as well as your repair process.  I am really hoping to be able to fix this image.  Thanks!

https://drive.google.com/open?id=0B3h479N_kZr6bXFuWUREXzBGQzQ

Pages: [1]