Lodestar Live v12 Mac OS X - any issues?

[arkosg]

Hello all,

I was able to get out on a gusty night and finally try out the latest edition of Lodestar Live, v 12 (thanks again Paul!).

First, v12 solves the auto-filename issue I was having in v 11; on Mavericks, I'm again getting automatic generation of filenames based on target name and date, which is great!  

I did however run across a couple of new & strange occurrences, and thought I'd check to see if any other Mac OS X users of LL have experienced anything like it.  I did my usual dark frames prior to starting viewing, and then was using LL as per usual.  I had turned on the separate output window at first (with all info displayed in it) and placed it in a separate desktop; I also used LL with the secondary window disabled.

The two main issues I encountered:

1) LL locked up on me (at first last night, and then I was able to recreate it today during the day).  With dark frames taken, and with the secondary output window placed on another desktop so I could easily swipe between them, I ended viewing an object, and then went to the Focus, Alignment and Framing window to prepare to look for my next target when I got the "spinning beach ball"... everything slowed way down, including LL response to clicks, etc., and the Finder reported LL as "not responding".  After a few minutes of fruitless clicking, moving the secondary window back to the same desktop as the primary, and trying to disable secondary output, I finally had to force quit LL to fix the issue.  I was able to recreate this scenario today in testing, though I waited longer, and it appeared that LL did eventually "unfreeze" this time, but there was about 10 minutes of no response.  I'm not sure exactly what triggered this, though, since I had viewed objects fine for some period using similar steps last night before hitting this issue.

2) After viewing several objects, and adjusting colour balance with no problem on the histogram, I was suddenly presented with a histogram that had the blue and green peaks showing up normally, but the red peak was a "spike".  Otherwise, things worked fine, but it made it impossible to white balance the images.  I tried all kinds of changes to settings, unplugging the camera, etc. but the problem was only solved by quitting and restarting LL (at which point everything went back to normal, and remained that way for the remaining few hours I was observing.)

Images showing the appearance of the object in the affected/normal state below (I didn't get a shot of the window/histogram - sorry!)

Thanks in advance,

Greg A

[HiloDon]

Greg,

I did not experience your number one, but I did have some strange thing happen with the color balance. It was only with one object, M8. I suspect it has something to do with the object.colors that cause LL to have trouble with recognizing and aligning the peaks. I had something that looked like your M51.

I think this is good feedback for Paul.

Paul, let us know if there is something we can do to help.

Don

[Martin Meredith]

Re issue 1, I haven't used the secondary window much but when I have on occasion it seems to have slowed down the responsiveness of the image manipulation controls (perhaps not surprisingly). What exposure time were you using in the focus/alignment panel? I normally use 1s for searching/centering and 250ms for focusing. At these rates I guess the refresh burden on the processor with the secondary window open might be higher?

Martin 

[arkosg]

Hi guys,

Thanks for the feedback. 

Don - the strange thing was, everything was fine colour wise, and then suddenly for whatever reason, it wasn't, regardless of the object - I must have shot 5 different galaxies, a globular, a nebula, etc. all with the same issue.  As I said, changing from linear to nonlinear modes, resetting the colour balance, etc. did nothing to fix it, only a software restart, and then it worked fine (even on the same objects I'd just viewed where it hadn't).  I don't know if it was just a glitch, or something more systematic/reproducible...

Martin - your settings are very much like what I use as well.  I noticed at first a major slowdown, and then intermittent "OS X spinning beach ball", and then finally the lockup.  I *had* just gone from viewing something to using the focus/framing window (with it's rapid 1s update), so perhaps you are on to something - perhaps my 7 year old Macbook Pro just isn't up to the task, processor wise?

Thanks again!

- Greg A

[Paul81]

Thanks for the info Greg.

Hmm... does sound like a few bugs. The idea on the secondary window using more processor time is a good one, having the second window open will load the processor more. Were your windows maximised - if so then the scaling algorithm has more pixels to process so this will increase load.

Might be worth making the window smaller, or setting the display scaling to fast (will be more blocky though).

I do want to optimise some of the code more as and when time allows.

Ideally, if you can re-create the problem repeatedly and let me know the steps I can then look into debugging to find the cause.

The histogram issue is weird. Did you set LL to save the raw exposures to FITS files?

[arkosg]

Thanks for the info Greg.

Hmm... does sound like a few bugs. The idea on the secondary window using more processor time is a good one, having the second window open will load the processor more. Were your windows maximised - if so then the scaling algorithm has more pixels to process so this will increase load.

Might be worth making the window smaller, or setting the display scaling to fast (will be more blocky though).

I do want to optimise some of the code more as and when time allows.

Ideally, if you can re-create the problem repeatedly and let me know the steps I can then look into debugging to find the cause.

The histogram issue is weird. Did you set LL to save the raw exposures to FITS files?

Hi Paul,

Yes, I had the secondary window maximized, so perhaps it was just a case of old Macbook Pro & too much going on!  I also had scaling set to better mode...

I did not have FITS output option on.  It was strange - worked fine for a while, then suddenly didn't... and then after a quit and restart LL, worked fine again, 

even on the same objects where it was acting up only minutes earlier.   As I said, I tried changing between linear/nonlinear modes, tried turning stacking on/off, 

anything that I thought might reset things, but nothing seemed to work except quitting and starting up again (and taking new darks! ; - )

But Paul, really, these are minor quibbles - I still love using your program - it's just really easy and does such a good job!  Fun!  : - )

If I run across the issues again next time, I'll take more careful notes and see if I can narrow down the conditions causing them if I can....

Cheers,

- Greg A

[Paul81]

[sorry for taking so long to reply..]

Thanks Greg 

Hopefully it carries on behaving itself!

[arkosg]

Hi Paul,

I was out for another run with v.12 last night, and I had similar problems reappear as the first time.

MacBook Pro from 6+ years back, running Mavericks 10.9.4.

Last night, I had successfully taken dark frames, was all set up to go, was seeing things in the focus/framing window just fine, and then was ready to start imaging.  So I checked "enable" for the secondary output window, and then returned to the focus/framing window, set up a 1s integration time, and then things locked up (spinning beach ball).  Some events seemed to proceed VERY slowly (eg. task bar for integration), but I couldn't get user interface items like buttons, menus, sliders, etc. to respond at all.  Again, I had to force quit to solve the issue.  I proceeded the rest of the night with the secondary output window disabled, and things were fine.

I also had the strange colour quirk reappear.  What I noticed this time was that it occurred after a period of using the non-linear modes, jumping between the two & adjusting the brightness, contrast, histogram, etc.  And then I attempted to return after several objects to linear mode, and suddenly again the red peak showed up as a spike and refused to operate properly.  No amount of twiddling with settings, returning to nonlinear, pressing reset for the histogram/image display settings worked.  Luckily this was near the end of my planned observing time anyway, so I just called it a night, figuring it was a sign.  ;- )

Cheers,

- Greg A

[Paul81]

Hi Greg,

Thanks for the info. Do you know what CPU your mac is running (Apple logo - About this Mac)? I am going to move the image repeat to a separate thread in this next release to ease processing, as currently a single CPU core will process the main display and the repeat. If there are multiple cores may as well make use of them! (Stacking runs in its own thread).

The colour quirk is very strange. Did you happen to be exporting the raw images as FITS? If not can you export them next time, and also try disconnecting / re-connecting the camera (just pull the USB out / in on the laptop).

Paul

[arkosg]

Hi Paul,

I have a 2.4 GHz Intel Core 2 Duo, with 4 GB of RAM.  It's an older model - about 6 years old.  Graphics are  NVIDIA GeForce 320M 256 MB.

For the colour issue, I was NOT doing any sort of FITS exporting, and I didn't try to unplug/replug this time, but I did last time to no avail.

I can try the FITS export next time it arises....

Cheers,

- Greg