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 - LarryC

Pages: [1] 2
I reported this a couple of years ago and Juan agreed it was a problem and that it was going to be fixed, but apparently it has not been.   I have saved icons for each step of my process with all the settings so I don't have to figure them out each time.  However, with at least Cos Cor (possibly others) if the master dark file location from the last time I used the saved process icon is no longer valid, I get an error pop-up and the entire CC module opens with all settings reset to defaults!  This makes no sense.  Why not just default the single setting (a file location) and leave the rest of my settings alone? Or better yet, let me know, but then let the process fail when I try to run it? 

I'm on, Win 10. 


General / PI on Hyper-V Ubuntu way slower than Win 10
« on: 2018 June 26 10:11:00 »
I finally figured all the installation and file issues in installing PI in an Ubuntu VM on my Win 10 PC, but when I run benchmark I get about 50% higher swap speed (great) but only 25% of the Windows CPU speed (~6K to ~1.5K).  I must have set up something wrong, but I have no idea what to tweak.  Any advice would be appreciated.

The PC is a 4770K OC'd to 4.4gHz with 32Gb Ram and ~7TB of SSDs and HDDs.  I use a 16gb Ramdrive with 4 swap files for PI.

The VM was set up with 8 logical cores, 16Gb ram allocated, and a 32GB swap disk and 30Gb OS partition on a Samsung 850 SSD shared with various files (mostly photos) but not my Windows boot drive).  I've not set up anything else in Hyper-V or in Ubuntu.   

Any suggestion on what I either did wrong or need to tweak?


Bug Reports / 1.8.5 trivial
« on: 2017 August 04 06:50:17 »
Very trivial item, but ImageIntegration Cache Preferences pop-up box's button's "Purge persistent cache now" text  doesn't fit on the button and is partially cut off.

Bug Reports / Bug or what did I do?
« on: 2016 December 30 11:23:52 »
Working with a project and CurvesTransformation, I had a real-time preview open and pressed something, but I don't know what, and the preview went full size in the workspace with no way to close or resize the image.  The real-time preview controls on the left side are present, but the top (or bottom) "close preview" button is gone - the preview goes from the main PI tool bar to the bottom PI tool bar, and there is no way to resize it.  I can select another image from the lower PI toolbar, but it is hidden behind the real-time preview. 

I don't know what I did to cause this or if it is a "feature", but I have tried every method I can think of to close or resize the preview, but nothing has any affect on it.


Bug Reports / DBE fail on re-opened projects
« on: 2016 November 23 15:45:10 »
This has happened a number of times now over the past few months.  When I open a saved project that contains several integrated images and then open a saved DBE Icon with pre-established points, the points appear on the image like one would expect but they cannot be moved or deleted.  The only way to correct the problem is to close image then re-open the image from where it is saved.  The points generated by the saved DBE icon then can be moved or deleted.


General / PIxinsight in Linux - file access
« on: 2016 September 19 12:58:58 »
I have Linux Mint runing on my Windows 10 PC.  All of my several TBs of imaging data, process icons and output files are in Windows NTFS folders/drives.  I can browse and open files my entire PC from the Linux desktop, but in PI I can not browse any of the Windows folders and therefore can't open any data or icon files.  Is this a PI limitation or do I have a setting wrong somewhere?


General / Linux no faster than WIndows !
« on: 2016 September 17 15:12:57 »
I've been hearing how much faster PI is on Linux, so I spent the last 3 days getting Linux Mint 18 installed on my Windows PC as a separate bootable drive.  This was my first foray into Linux and it heck of a challenge.   In any case, it now runs great and I pointed to the same swap files on an SSD that I had set up for my Windows PI - Linux/PI is on a HDD and my WIndows/PI is on a SSD.   I get 5900-6200 total score on the PI benchmark test with Windows, and I got only 5800-5900 with Linux.  I am a little disappointed.  I was expecting more of a bump in speed. 

We're my expectations unfounded? 


General / RAM and PI speed
« on: 2016 September 12 16:55:52 »
Before I build an entirely new computer, I'm looking into ways to speed up PI on my current PC.  It's an OC'd i7-4770K running at 4.3 GHz with separate SSD's for OS, images and swap files.  I currently bench mark at ~6,100, but I'm looking for more speed.  I currently have 16GB of RAM and wonder if buying 32GB will really matter.

I routinely process 200-400 x 32MB files at a time but rarely see PI use more than ~10GB or of of my available RAM, so it's not clear to me that having 32GB will make much difference.  Has anyone bench marked before and after adding more, or faster, RAM or have any real evidence that more than 16GB is useful?  If I go to 32GB would a 16GM RAM disk make a noticeable difference?


Bug Reports / Another "Pause/Abort" annoyance
« on: 2016 September 10 14:27:56 »
This might be a Windows issue, but when I accidentally selected 150 x 32mb frames in a folder after double clicking the workspace to open an image, I could not stop the process.  As soon as I realized my mistake, after 2 images opened (it takes ~2 seconds to opens each image) I tried to "pause/abort" from the Process Console.  I could not stop the opening of images.  After about 50 of them opened, and a coupe dozen clicks on the "pause/abort" button, the pop-up dialogue opened asking if I really wanted to abort, and the opening of images paused.  "Yes" and "no" were not responsive.  I could move the pop-up around on the screen and I could close it with the "X", but it would not stop the process.  Closing the dialogue box would resume the opening of images.  I ultimately had to terminate PI through task manager. 

General / StarAlignment with differnt image scale sets fails
« on: 2016 July 31 13:50:04 »
I understood that one could register images taken at different image scales against a common reference frame.  When I try I get up to 100 out of 200 images fail to register, but with fairly extreme adjustments to StarAlignment's star detection parameters I can get that down to about 30 or so.  All of the images look fine and each set registered 100% against a reference frame from within that image scale set using near default StarAlignment parameters.

Should this be possible to accomplish or am I asking too much from StarAlignment?


Bug Reports / More an annoyance than a bug, but...
« on: 2016 July 29 12:50:00 »
When trying to open a saved cosmetic correction icon, if the master dark is not in the location specified when the icon was last saved it throws and error message, which would be fine, but when you proceed past the error notice, it resets everything related to master darks - old location, file name, hot pixels settings, etc.  I would much prefer that it not error on opening but rather error like other processes do upon application.  And I would much prefer that it allows me to find the master dark without deciding I need to reset all of my parameters. 



SFS always works very well, but tonight I got this error in all of my calibrated and corrected frames:

"Warning: No convergence in MRS noise evaluation routine: using k-sigma noise estimate"

And the values for most of the output variables were "0".  As far as I know I processed the light frames the same as I have done for some time. 

Anyone have any ideas what might be going wrong?


General / Drizzle Image Weights
« on: 2016 July 20 17:35:34 »
After I drizzle integrate my images PI gives me a "Image Weights" image.  What is this and what can it be used for?

Bug Reports / More of a glitch than a bug, but...
« on: 2016 July 09 20:35:54 »
If one accidentally opens StarAlignment and forgets to clear and load new files, so that StarAlignment tries to run with the file names from the previous run, and those files no longer exist, it is impossible to abort the process because each failed file-access pops-up an error dialogue that blocks all other functionality, such as the "Pause/Abort button, but the pop-up for the next failed file access pop-ups before any other button can be pressed.  That means the ONLY way to stop the process is to terminate PI through Task Manager, or face a long night of clearing error message pop-ups one at a time with 300+ files in the list.

Bug Reports / Can no longer adjust screen size
« on: 2016 April 09 13:11:47 »
No other program are affected, reboots do not help.  I can no longer adjust the vertical size of the overall workspace.  This is a problem in that the program always opens too large vertically for my screen so I routinely shrink it from the upper right, re-position, then enlarge from the lower right to fit properly - an old annoyance, but never a problem until recently. 

The change occurred "around" the time of the recent updates to...I think, PSF?

Pages: [1] 2