[Solved] Java error when beginning a new session

Hi George @digug,

Thank you very much for sharing this issue, this was a bug in beta22 or before. Please use beta23 or later, that has this issue fixed 😉

you can find downloads in the downloads section here:

https://www.astropixelprocessor.com/downloads/

In addition, APP uses it's own version of java (is included in the installation package and is in APP's installation folder once installed), it does not matter if you have installed java or not, nor does the java version that you have installed affect APP's performance nor operation. If it would, it would not be good... 😉

Mabula

hi @mabula-admin, i also have this error with 2.0.0-beta26, but only with flats.

the only working workaround that i found is to load only one light, completely leave out any other calibration frames than flats and integrate. the calibrated light can obv. be deleted, but with that i at least get the masterflat.

from there on everything works as usual.

both below posted error messages appear twice, then APP gets stuck. in my example i loaded 176 lights for 7 different filters (so multi-filter mode), 350 flats (50 per filter), 350 dark-flats (50 per filter, flat-based exposure time), 3 master-darks (3 different light exposure times) and 1 master-bias.

Encountered error in module:                     
FrameDataCalibration

Error message: 
java.lang.ArrayIndexOutOfBoundsException: Index 1 out of bounds for length 1

Cause: 
Index 1 out of bounds for length 1
fP.a(FrameDataCalibration.java:992)
fP.doInBackground(FrameDataCalibration.java:47)
java.desktop/javax.swing.SwingWorker$1.call(Unknown Source)
java.base/java.util.concurrent.FutureTask.run(Unknown Source)

Encountered error in module:                     
LightFrameCalibration

Error message: 
java.lang.ArrayIndexOutOfBoundsException: Index 1 out of bounds for length 1

Cause: 
Index 1 out of bounds for length 1
fP.a(FrameDataCalibration.java:992)
fP.doInBackground(FrameDataCalibration.java:47)
java.desktop/javax.swing.SwingWorker$1.call(Unknown Source)
java.base/java.util.concurrent.FutureTask.run(Unknown Source)

Hi @deepsky-photo,

Thank you for reporting this.

Can you check for me the settings in 0) Raw/FITS when this error occurs? Does the error occur if you reset all setting with the CFG button?

Feel free to upload the data so I can test myself and solve this properly :

https://upload.astropixelprocessor.com/

username: uploadData

password: uploadTestData

Please make a folder like: deepskyPhoto-dataCalibration-bug

and upload it there and let me know once done, thanks !

Mabula

Mabula

Hi Mabula,

thanks for your reply, much appreciated. So what I did in the meantime is to;

1. CFG->RESET to defaults
2. CFG->RESET all menu tabs to defaults

(..while still being on 2.0.0-beta26).

Since then, no additional errors occured. I also tried to replicate the issue to give you further hints, but I'm afraid I can't.

In general, as I'm working with a mono-setup;

• 0) RAW/FITS => pattern=supported, algorithm=adaptive airy disc, force Bayer/X-Trans CFA=false
• 1) LOAD => Multi-Channel/Filter processing=true, Multi-Session processing=false, auto-detect Masters & Integrations=true
• 2) CALIBRATE => everything default
• 3) ANALYSE STARS => everything default
• 4) REGISTER => everything default
• 5) NORMALIZE => everything default
• 6) INTEGRATE => I tend, after manual selection, to set "lights to stack" between 90-95% (maybe that was the issue behind my initial post in this thread - but it's working now after resetting CFG), LNC degree=1st degree LNC, LNC iterations=3, enable MBB=true, MBB=4%, everything else default

CS,

Christian

Hi Christian @deepsky-photo,

Okay, that is reassuring. I do know what probably went wrong if I look at our code. Your error message has a bit more information when compared to the error message in the first post of the topic, because newer APP versions will show a bit more error detail now.

The error indicates that the MasterBias that was used was a CFA masterbias, so it contained a Bayer pattern and R,G,G,B statistics for it's median, average etc...while the Masterdark was not a CFA Masterdark, so no Bayer pattern was in the header. Maybe that explains it?

I think I know enough now to reproduce this error and then fix it in our code, to make sure that this error is gone 😉

Thank you very much for reporting this.

Mabula

Hi @digug and @deepsky-photo,

This bug is fixed in beta29 which will soon be released. The bug woudl manifest is the masterbias had a CFA pattern in the fits header and the masterdark not while processing monochrome or OSC data. So it would be caused by mixing up monochrome and color lights and masters...

In beta29 the error will not be thrown and the data would still be calibrated.

Thank you both for reporting this 😉

Mabula

@mabula-admin Thanks for the update.  I haven't seen the error since I posted the OP.

Great George @digug, that is good to know, thank you 😉 I guess it would only occur when color and mono processing is mixed up for some reason. Glad to have caught this bug now !

Posted by: @mabula-admin

↑

The error indicates that the MasterBias that was used was a CFA masterbias, so it contained a Bayer pattern and R,G,G,B statistics for it's median, average etc...while the Masterdark was not a CFA Masterdark, so no Bayer pattern was in the header. Maybe that explains it?

Now that you mentioned it I remember that I took a look again once I resetted the CFG and found that the calibration masters were indeed flagged as RGB. They were created within my first (mono) integration run, must be a setting that was wrong as I recently changed to a mono setup. Anyhow, glad I was able to support with the logs above and thanks for that great piece of software!

CS everyone!

I'm now getting the java error every calibration.  I can't finish any stack. I only stack color lights.

Dear George @digug,

Can you try with 2.0.0-beta29 which I released today?

The error occurs because there must be a mixup with color and mono data. If you are using older master calibration files, they probably have no bayer pattern in their header which can cause the issue.

I have added a fix in beta29, so I think that error must be gone now though.

Mabula

@mabula-admin this is great news!  I will say tho I only image in color and even when I was creating masters from fresh stack sessions I got the error but I will try beta29 later.  I also saw in another post someone reported beta29 is faster on Mac so I hope I see some performances gains on Windows.

@mabula-admin

Installed beta29, spun up a session, and it stopped at creating master flat, no error messages.

UPDATE:
Wanted to point out that I installed beta29 over beta28 then tried stacking when it hung.  I uninstalled beta29 and cleared registry entries related to app then rebooted.  Now I got through Calibration.  Seems to be working fine now.