|Allsky1 System Log|
2021dec06 - after moving starlightcascade.ca website to a new host, we have removed allsky1 archives. Only the most current days worth of data will be here online for now.
2021oct08 - removed allskypi software v0.6 and replaced it with v0.8, reconfigured it, then ran dark frames most of the night. Much Better results!
Need more dark frames at cooler temps.
Credit to this presentation to alert us about the new software version:
software project available here: https://github.com/thomasjacquin/allsky
2021sept21 - a new raspberry pi 3b was purchased, shipped, arrived and installed. working well!
2021sept20 - the Raspberry pi 3B has failed, after being outdoors in the elements for almost 2 years. The allsky1 camera system will be down until we can figure out where to go next.
2021August - the 8GB microsd card is just not big enough.. once there is not enough tmp space to create nightly runs, it fails. So.. cloned it over to a 32GB microsd card and installed. runs great!
2021june11 - the allsky1 camera system failed in the last two days and was removed today from the field for testing.
2021May12- received new 1.55mm f2 lens and replaced the existing 2.1mm f2 lens that came with the zwo asi120mc camera. Much better! *almost* full 360 degrees but the camera sensor is probably just a little too small.
2021April02 - did a quick analysis on limiting magnitude of this system and at first glance on a clear night was a star near 83 UMa at mag 4.75, 60 second exposure.
Allsky1pi has uptime of 111 days. It uses an 8GB microSD card for storage, wifi network to send files. It keeps 7 days of images onboard and deletes the oldest one day at a time.
2020Nov 10 - disconnected air blower as the dome was freezing, condenseing and frosting over on the cold nights. The airflow through the main body was taking away the heat generated by the resistive heater in the dome itself
2020Sept27 - solar panels to power air pusher in main body of the allsky1 has been replaced with a 12vdc adapter. Hopefully this will cool the camera sensor during the hot daytime temperatures.
2020Sept22 - not uploading data for 5 days now. reboot after 42 days uptime and rerun data upload scripts, all ok again. do an update and upgrade and reboot again into a new pi kernel. still running ok.
2020sept21 - due to many power outages, some lasting hours, the mechanical timer powering the dome heater was out of sync with reality and frost was forming on the dome by morning. The timer has been completely removed as we are entering into a much cooler season.
So now the heater is on 24/7. The solar powered forced air circulation was disconnected as we look to add a power adapter to allow it to run 24/7 as well, keeping the camera and sensor cooler than otherwise.
2020April08 - awesome uptime of 31 days without intervention! Occasional failure to generate an .mp4 video summary of the night. That's about it.
2020March22 - we were finally able to create a dark frame for the allskypi camera system. The trick was to take a dark frame in the total darkness with a dark dark cover over the dome. The previous attempt was during the light evening with a cover that was not great... the dark frame processed out almost all of the signal of the real image!
2020Feb26 - 3rd night in a row that the overnight video file has been created with errors... troubleshooting now
2020Feb24 - installed new power supply for dome heater... a variable 3-12vdc 30 watt system. Plugged in this evening, set to 4.5V = 2Watts to start.
From amazon.ca for $21 [Upgraded Version] SoulBay 30W Universal AC/DC Adapter Switching Power Supply with 8 Selectable Adapter Tips, Including Micro USB Plug, for 3V to 12V Household Electronics and LED Strip - 2000mA Max
2020Feb17 - rotated camera 180 deg so that north is up. With a lens, up on the camera is at the USB connection.
Dark frames are not yet implemented and the heater seems to be not working either.. we got frost last night.
2020Feb16 - scripts to upload latest daily images and video now operational
2020Feb13 - Had a good run overnight, albeit cloudy and snowy. ToDo: increase heater voltage from 6vdc to 9vdc. That should clear away the snow on the dome.
In addition, look at taking a dark frame in the next few days.
In addition, rotate the camera so that North is Up!
2020Feb12 - Done! rebuilt the allskypi housing, removed the 110vac and now have a heater in the dome.
Reinstalled just before sunset and powered up. Success! Software starts, wifi network autoconnected, imaging software autostarted and images were copied over to the server every 30 minutes or so. Looking good!
2020Jan13 - looking to add heating element inside the dome. condensation and frost and snow a common occurence now. Looking at adding another cable, either 5vdc or 12vdc.. not sure of how much wattage to add yet.
2019Dec24 - 1st successful run of the night!
| width | 0 | 0 means max width. Look up your camera specifications to know what
values are supported |
| height | 0 | 0 means max height. Look up your camera specifications to know w
hat values are supported |
| exposure | 10000 | **Night** time exposure in milliseconds. During the day, au
to-exposure is used. |
| maxexposure | 20000 | This is the maximum exposure for **night** images when u
sing auto-exposure. During the day, auto-exposure is always used. |
| autoexposure | 1 | Set to 0 to disable auto-exposure at **night**. Auto-exposu
re delivers properly exposed images throughout the night even if the overall bri
ghtness of the sky changes (cloud cover, moon, aurora, etc). When set to 1, *max
exposure* value will be used as the delay between timelapse frames. |
| gain | 50 | Gain for **Night** images. Varies from 0 to 600. During the day, g
ain is always set to 0 |
| maxgain | 200 | Maximum gain for **night** images when using auto-gain.|
| autogain | 0 | Set to 1 to allow auto-gain at **night**. This mode will adjust
the gain of night images when the overall brightness of the sky changes (cloud
cover, moon, aurora, etc). **Avoid using autoexposure and autogain together** as
it produces unpredicatble results (dark frames, but not always).|
| gamma | 50 | Varies between 0 and 100. This setting increases or decreases con
trast between dark and bright areas. |
| brightness | 50 | Varies between 0 and 100. This setting changes the amount of
light in the image. |
| wbr | 53 | Varies between 0 and 100. This is the intensity of the red componen
t of the image. |
| wbb | 90 | Varies between 0 and 100. This is the intensity of the blue compone
nt of the image. |
| bin | 1 | bin 2 collects the light from 2x2 photosites to form 1 pixel on the
image. bin 3 uses 3x3 photosites, etc. Increasing the bin results in smaller ima
ges and reduces the need for long exposure. Look up your camera specifications t
o know what values are supported |
| delay | 10 | Time in milliseconds to wait between 2 frames at night. |
| daytimeDelay | 5000 | Time in milliseconds to wait between 2 frames during the
| type | 1 | Image format. 0=RAW 8 bits, 1=RGB 24 bits, 2=RAW 16 bits |
| quality | 95 | Compression of the image. 0(low quality) to 100(high quality) f
or JPG images, 0 to 9 for PNG |
| usb | 40 | This is the USB bandwidth. Varies from 40 to 100. |
| filename | image.jpg | this is the name used across the app. Supported extensi
ons are JPG and PNG. |
| flip | 0 | 0=Original, 1=Horizontal, 2=Vertical, 3=Both |
| text | text | Text overlay. **Note**: It is replaced by timestamp if time=1 |
| textx | 15 | Horizontal text placement from the left |
| texty | 35 | Vertical text placement from the top |
| fontname | 0 | Font type for the overlay. 0=Simplex, 1=Plain, 2=Duplex, 3=Comp
lex, 4=Triplex, 5=Complex small, 6=Script simplex, 7=Script complex |
| fontcolor | 255 255 255 | Font color in BGR |
| smallfontcolor | 0 0 255 | Small Font color in BGR |
| fontsize | 0.7 | Font size |
| fonttype | 0 | Controls the smoothness of the fonts. 0=Antialiased, 1=8 Connec
ted, 2=4 Connected. |
| fontline | 1 | font line thickness |
| latitude | 60.7N | Latitude of the camera. N for North and S for South
| longitude | 135.05W | longitude of the camera. E for East and W for West |
| angle | -6 | Altitude of the sun above or below the horizon at which capture s
hould start/stop. Can be negative (sun below horizon) or positive (sun above hor
izon). 0=Sunset, -6=Civil twilight, -12=Nautical twilight, -18=Astronomical twil
| time | 1 | Replaces the text overlay |
| darkframe | 0 | Set to 1 to enable dark frame capture. In this mode, overlays
are hidden and the image is saved as dark.png by default |
| showDetails | 1 | Displays the exposure, gain and temperature in the overlay |
The second file called **config.sh** lets you configure the overall behavior of
the camera. Options include functionalities such as upload, timelapse, dark fram
e location, keogram.
| Configuration | Default | Additional Info |
| ----------- | ----------- | ----------------|
| UPLOAD_IMG | false | Set to true to upload (ftp) the current image to a server
(website, blog, host, etc) |
| UPLOAD_VIDEO | false | Set to true to upload the timelapse to a server |
| POST_END_OF_NIGHT_DATA | false | Set to true to send some data to your server
at the end of each night |
| TIMELAPSE | true | Build a timelapse at the end of the night |
| KEOGRAM | true | Builds a keogram at the end of the night |
| UPLOAD_KEOGRAM | false | Set to true to upload the keogram to your server |
| STARTRAILS | true | Stacks images to create a startrail at the end of the nigh
| BRIGHTNESS_THRESHOLD | 0.1 | Brightness level above which images are discarded
(moon, head lights, aurora, etc) |
| UPLOAD_STARTRAILS | false | Set to true to uplad the startrails to your server
| AUTO_DELETE | true | Enables automatic deletion of old images and videos |
| NIGHTS_TO_KEEP | 14 | Number of nights to keep before starting deleting. Needs
AUTO_DELETE=true to work. |
| DARK_FRAME | dark.png | Path to the dark frame use for hot pixels subtraction.
| DAYTIME | 1 | Set to 0 to disable daytime liveview. |
| CAMERA_SETTINGS | /home/pi/allsky/settings.json | Path to the camera settings
file. **Note**: If using the GUI, this path will change to /var/www/html/setting
In order to upload images and videos to your website, you'll need to fill your F
TP connection details in **ftp-settings.sh**
## Dark frame subtraction
The dark frame subtraction feature was implemented to remove hot pixels from nig
ht sky images. The concept is the following: Take an image with a cover on your
camera lens and subtract that image later to all images taken throughout the nig
You only need to follow these instructions once.
* make sure config.sh has a DARK_FRAME configuration. Default is "dark.png"
* Place a cover on your camera lens/dome
* Set darkframe to 1 in settings.json
* Restart the allsky service: ```sudo service allsky restart```
* A new file has been created at the root of the project: dark.png by default
* Set darkframe to 0 in settings.json
* Restart the allsky service: ```sudo service allsky restart```
* Remove the cover from the lens/dome
A **Keogram** is an image giving a quick view of the night activity. It was orig
inally invented to study the aurora borealis.
For each image taken during the night, a central vertical column 1 pixel wide is
extracted. All these columns are then stitched together from left to right. Thi
s results in a timeline that reads from dusk to dawn.
To get the best results, you will need to rotate your camera to have north at th
e top. That way, using a fisheye lens, you end up with the bottom of the keogram
being the southern horizon and the top being the northern horizon.
Note that it will only show what happens at the meridian during the night and wi
ll not display events on the east or west.
The program takes 3 arguments:
- Source directory
- File extension
- Output file
**Startrails** can be generated by stacking all the images from a night on top o
f each other.
The program takes 4 arguments:
- Source directory
- File extension
- Brightness treshold to avoid over-exposure: 0 (black) to 1 (white).
- Output file
Example when running the program manually:
./startrails ./images/20180223/ jpg 0.15 ./images/20180223/startrails.jpg
2019Dec06 - still doing systems integration testing and software configuration.
2019Dec01 - developing the replacement Allsky1 system, based on an existing ZWO ASI120MC camera, and a Raspberry Pi 3B. So far so good!
2018 July 02 - SBIG ST237A camera system has failed and the system has been taken offline
2018June22 - allsky dome cleaned inside and outside and replaced. It was removed;
repaired midnights script and regenerated
2018April04 - The roof of the SCGObservatory blew off in a windstorm during the daytime. Max gusts reached 66kph on our SCG Weather Station. The Allsky1 computer and control system was exposed to a few inches of snow after that, and may or may not be online until the situation is resolved.
2017Dec10 - upgraded the 5vdc power source on the dome heater to 12vdc, resulting in a change of wattage delivered from 0.5 watts to 2.88 watts.
was: v=ir; i=v/r=5/50=0.1 amp; W=IV=0.1*5=0.5 watts
is: i=v/r=12/50=0.24 amps; W=iV=0.24*12=2.88 watts; which is about a x6 increase in heat delivered.
2017Nov29 - The 0.5 watts of heat are not nearly enough to combat snow and ice, although it has been doing not a bad job on condensation. Will ramp that up to a 12vdc supply in the near future. The setup photo at the top of the page has been updated. The black plastic around the dome is a screen to block local lighting.
2017Nov03- computer hardware as of today: Microsoft Windows XP Service Pack 3, on a Dell Precision workstation 370, Pentium 4 CPU @ 2.8GHz with 2GB DDR2 RAM, 250GB SATA drive. The camera controller uses the LPT port and is turned off and on each day via a mechanical timer on its power.
It is connected with a Gigabit network ethernet. The video display resolution is 1024x768. The CCDOps software is run by a mouse match script, so positioning of the CCDOps windows is critical.
2017Oct 22 - added 0.5 watts of heat inside the 4" dome in the way of a 5vdc powered two 100ohm resistors in parallel (50ohm). No noticable difference so far, after one night. Also removed a lot of bad silicon and replaced with sticky velcro
2017March28 - cleaned dome surface with concentrated rainx so repel water better. Dome has adhered to camera but there are still gaps. No heater has yet been installed.
2017Mar20 - system restarted with new small 4" dome overtop of SBIGST237A camera system. Seal not complete, open to air but not rain. Main reason is to protect from rain. Reset start and stop times and noticed frost in the morning. May need heater again as well.
2017Feb27 - water in the lens. The system will be down for the new few days while the lens is removed and dried out.
2017Jan01 - celebrating 10 complete years of this system in operation (since 2006 December). There have been a few downtimes, mainly when the old computers with an LPT: port dies in the outdoor cold (has gotten down to -35C) and the time needed to find a replacement. The camera itself has been solid.
2016Oct20 - looks like the mechanical timer that starts and stops power to the camera control box is failing... the camera shut down at 01:28 EDT this morning instead of 08:00 or so. will remove the timer for now and leave the camera controller on 24/7
2016march31 - fixed code generating daily .avi summary of images.. we including older images as well
2016Jan22 - update code references from 2015 to 2016, especially in the processing of grabbing and displaying midnight images.
2015Oct19 - system failed last night, first time in many months. will manually watch over it this evening. Since the camera was removed from the housing, the CCD temperature and peltier cooler are working much better now, on the order of 25deg C
2015April12 - completely removed camera system from existing housing and 14" dome and mounted in a similar orientation on the same post just the camera system with integrated lens. DOcumentation suggested the system was waterproof and it would remove the huge heating requirements of a 14" dome airspace, a 2nd optical surface and fewer bugs sitting on the inside of the dome.
2015Feb18 - .avis have been small but poor quality.. change starting tomorrow using modified conversion parameters /web/allsky1/ffmpeg -r 10 -i %05d.jpg -vcodec mpeg4 -b 800k today.avi
will increase file size from approx 1.5MB to 4.6MB
2014Dec11 - added new FFMPEG to software resources and from today onward will generate much smaller .AVI files (<3MB) replacing the larger .GIF files (!40MB)
2014Dec10 - removed radiojove software from this computer and put onto standalone laptop
2014Jan25 - added radiojove software radio skypipe to this computer as well as the RJ webcamera. 24vdc power supply for heaters still working. Extreme cold temperature mean the heaters cannot keep up. Exposure time still 90sec. Considering a move to 60 sec. would have greater percentage of dead time as images download, from 10% to 15%. Would get better exposures when moon is up. Moving object detection/exposure should not change.
2013Oct - 12vdc power supply for circulation fans failed.
2013Aug10 - Power system reconfigured, now all parts of system now on UPS.
2013Aug08 - some part of the system is not on the UPS and whenever there is a short power outage, the recording of images for the night ends and does not restart. Flies and bees still getting inside dome. TIme to think about redesigning and rebuilding the camera housing.
2013Feb25 - replaced laptop with desktop again.
2013Feb24 - replaced temp desktop with 10-13 year old dell inspiron laptop (pentium mobile @2.0hz, 512mb RAM, 40gb IDE drive). Totally underpowered and failed to run the ccdsoft software.
2013Jan23 - computer crashed again. turned off around 22:00 today for repair.
2013Jan22 - discovered allsky desktop computer crashed around 23:00 EST Jan 21 due to extreme cold. Restarted at 20:00. Crashed again around 21:08 Jan22.
missed a good meteorite over NY state.
2012Nov21 - the batch script to start the ccdops program under windows 7 now works much better and is automated again.
2012Nov20 - autostart scripts failing the last two weeks.. needed manually startup every single night. Finally think this is now fixed for tonights run.
2012Nov16 - added in American Meteor Society maps for the events we captured.
2012Oct04 - everything seems to be working fine. annotations are good. Built winter insulating cover for desktop allsky1
2012Sep03 - replaced laptop with old desktop for now. mostly operational. needed a lot of configuration tweaking. Also seem to have an older version of CCDSoft then we used to. strange. internal FITS headers are different and that broke some of the scripts. The temperature and exposure data annotations are still failing on the midnight images.
2012Aug31 - allsky1 laptop hard drive fails. Ran diagnostic and many many sectors are toast and unrecoverable. Looking to rebuild a new computer. Allsky1 will be out of service until this is done.
2012Aug24 - updated annotation scripts again to add Eastern Time (-5/-4 UT). Black plastic mask around lens positioned well to block local houses direct light. Starting to shift start and end times again as sunrise and sunset are later and the nights are getting longer. Housing still needs new coat of white paint, dome needs a cleaning.
2012Aug09 - the new png scripts worked well but the uploads to this server did not. Also the noise moves around a lot.. changing to -noise 3 for tonight.
2012Aug08 - change batchfits processes to create .PNG images instead of .JPG; also change -noise 1 to -noise 2 in an attempt to get rid of the noise in the images since the ImageMagick upgrade on June 23, 2012
2012july06 - new network cable laid to observatory and new gigabit switch. changed processing for 'collection of midnights'
2012may01 - network to observatory failure again from april 30th 17:00 to 06:30 today. temporarily fixed, awaiting new network cable installation. Partial data collected. Adding enhanced annotation to images, part 1 (north,south,east,west)
2012april26 - added Aurora Collection
2012april23 - created new collection of images every midnight for an easily accessible historical record of that nights sky.
2012april23 - big wind and snowstorm overnight. Looks like the lens mask has shifted position twice through the storm.
2012march16 - network switch port burned out/dead, moved to a vacant port, all good.
2012march15 - very close lightning storm strikes, lost network to observatory
2012mar07 - data processing glitch.. will take another day to work itself through
2012Feb20 - laptop back in service, allsky partial run, complete run Tuesday Feb 21st evening.
2012Feb17 - laptop harddrive fails, allsky out of service
2012Feb08 - removed internal mask #2, installed slightly larger internal mask #3
2012Feb05 - removed internal mask #1, installed internal mask #2
2012Feb04 - removed external light mask from dome, installed internal mask #1
2012Jan19 - indications of failing network cable to observatory, especially in cold (-20) weather. Will wait for good connection on warm day, transfer over images from local
observatory storage and reprocess manually.
2012Jan11 - removed .mpg generation of large .gif as it was overwhelming the old slow processor on the web server and locking up the whole site for an hour each day.
total combined parallel heater resistance=33ohms@24vdc giving 17 watts of heat.
2011Dec16 - broke the main processing script on Dec 13 and just repaired this morning
2011Dec04 - replaced the dome and platform, replaced circulation fan with a larger one and mounted it inside the dome, remounted the two resistive heaters.
2011Nov04 - fixing some of the code for archiving images and animations.
2011Sep28 - added new "slide show" method of displaying the nights run of images, in addition to keeping the old one.
2011Sep19 - now annotating CCD Temp in Deg C, and now exposure time! CCD temperature starts out at about 17- -11=28degC below ambient. By the end of the nights run it is 12- -11=23degC. This has improved slightly since the housing vent was installed.
The first image of the night shows the camera at ambient.
2011Sep18 - removed housing door and installed vent with insect screen as camera was often 5-10 degrees C above ambient at the start of the nights run.
2011Sep16 - change startccd time from 00:30 UT to 00:01 UT
2011Aug02 - problems on the last two nights runs. should be fixed for tonight... just as the clouds come in :)
2011Jun19 - close to solstice, changed start times to 01:59UT (21:59EDT) as camera is oversaturating and process the last images at 08:00UT (04;00EDT)
2011May13 - startCCD time has been 01:19 UTC (21:19 EDT) and stop time 08:23 UTC (04:23 EDT) for some time now and even then the start and end are overexposing.
Problems with the parallel skypipe software interfering with ccdops seem to have been resolved as well.
2011Mar26 - changed StartCCD time to 00:30 UTC (20:30 EDT) and stop time to 10:00 UTC (06:00 EDT)
2011Feb22 - changed StartCCD time to 23:45 UTC (18:45 EST) and stop time to 11:15 UTC (05:15 EST)
2011Feb15 - changed processing scripts a bit to lighten the images as they had become too dark since migrating over to the laptop.
Changed startccd time to 18:32 EST changed STOP time to 06:04EST
2011Jan13 - another laptop (7 years old) installed with all software and allsky now up and running again.