bokser01
Posts: 19
Joined: Sat May 09, 2020 8:59 pm

HQ camera timelapse startrails

Wed May 27, 2020 6:04 pm

Hello, is there any order to take 60s of HQ camera exposures immediately after each other, as was the case with V2 using raspivid? Any tips? Thanks

User avatar
HermannSW
Posts: 3822
Joined: Fri Jul 22, 2016 9:09 pm
Location: Eberbach, Germany
Contact: Website Twitter YouTube

Re: HQ camera timelapse startrails

Wed May 27, 2020 7:44 pm

I used the 8x8 dot matrix display controlled by Arduino described here:
viewtopic.php?f=43&t=273358&start=25#p1664566
At each point in time exactly one dot is lighted, and for 4s this time.

I modified the command for capturing by adding "-tl" for timelaps and adjusting values:

Code: Select all

pi@raspberrypi2B:~ $ time ( raspistill -t 310000 -md 3 -bm -ex off -ag 1 --shutter 60000000 -st -tl 60100 -o long_exposure_%02d.jpg )
mmal: Frame 1 is 637 ms late
mmal: Frame 2 is 1028 ms late
mmal: Frame 3 is 1379 ms late
mmal: Frame 4 is 1728 ms late
mmal: Frame 5 is 2095 ms late
mmal: Frame 6 is 2421 ms late

real	7m5.329s
user	0m0.068s
sys	0m0.751s
pi@raspberrypi2B:~ $
I asked for 5min10s, got 6 frames. Took a minute more time, I would say that is OK.
"Frame late" messages should be avoidable with -tl time being more that 100ms longer than --shutter time.

Next I did cut out the interesting 1044x1044 part of the frames, scaled them to 20% size:

Code: Select all

pi@raspberrypi2B:~ $ for((i=1; i<=6; ++i)) do jpegtopnm long_exposure_0$i.jpg | pnmcut -left 1444 -top 828 -right 2488 -bottom 1872 | pnmscale 0.2 | pnmtopng > $i.png; done
jpegtopnm: WRITING PPM FILE
jpegtopnm: WRITING PPM FILE
jpegtopnm: WRITING PPM FILE
jpegtopnm: WRITING PPM FILE
jpegtopnm: WRITING PPM FILE
jpegtopnm: WRITING PPM FILE
pi@raspberrypi2B:~ $ 

Finally I created animated .gif from that

Code: Select all

pi@raspberrypi2B:~ $ ffmpeg -r 1 -i %1d.png anim.gif
ffmpeg version 4.1.4-1+rpt7~deb10u1 Copyright (c) 2000-2019 the FFmpeg developers
,,,
frame=    6 fps=0.0 q=-0.0 Lsize=      95kB time=00:00:05.01 bitrate= 154.5kbits/s speed=27.3x    
video:94kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.885057%
pi@raspberrypi2B:~ $

I think this shows nicely how to capture 60s exposure frames with timelapse (60s shows 15 or 16 dots lighted)!
anim.gif
anim.gif
anim.gif (94.51 KiB) Viewed 481 times
https://stamm-wilbrandt.de/2wheel_balancing_robot
https://stamm-wilbrandt.de/en#raspcatbot
https://github.com/Hermann-SW/Raspberry_v1_camera_global_external_shutter
https://github.com/Hermann-SW/raspiraw
https://stamm-wilbrandt.de/en/Raspberry_camera.html

billgs
Posts: 7
Joined: Fri May 22, 2020 8:57 pm

Re: HQ camera timelapse startrails

Thu May 28, 2020 3:02 pm

Greetings,

I have used cameras V1 and V2 for astrophotography time-lapse exercises
with settings such as
:- raspistill -t 30000 -ex night -ISO 800 -bm -tl 0 -ss 10000000-o timelapse1%d01.jpg
for V2.

With my HQ camera on a 80mm refractor telescope, I applied similar settings but varied
-ss up to 30000000. I found that, with -ss=20000000, after a wait of about 1min 10secs,
two 10.3 second images appeared followed by a series of 20 second images one after the
other up to the limit of -t. There was no increase of exposure length with -ss greater
than 20000000.

After the latest rpi-update I repeated the execise. I input:-
raspistill -t 600000 -bm -ex night -ISO 6 -tl 0 -ss 60000000 -o HQCam%d01.jpg
After a wait of about 1min 10sec two 10.3 sec images followed by a series of 7 60sec
images as indicated by EXIF.
Further tests with increased exposure length resulted in a maximum value of 240 secs.
(239.54 according to EXIF)

I set up an analogue clock test to approximate dark sky conditions.

Contact Sheet2.jpg shows the first 6 images from:-
raspistill -t 600000 -bm -br 10 -ISO 2 -tl 0 -ss 60000000 -o 23mayclockop%d01.jpg
[img] C:\Astronomy\raspberrypi\2020may21tests\clock2\crop2\crop3\Contact Sheet-1.jpg
C:\Astronomy\raspberrypi\2020may21tests\clock2\crop2\crop3\Contact Sheet2.jpg
[/img]
Contact Sheet-1.jpg shows the result of:-
raspistill -t 600000 -bm -br 5 -ISO 2 -tl 0 -ss 240000000 -o 23mayclockq%d01.jpg
[img]C:\Astronomy\raspberrypi\2020may21tests\clock2\crop2\crop3\Contact Sheet-1.jpg
[/img]
The values of -br and -ISO were needed to avoid over exposed images.

In both cases images 001 and 101 are the first two 10.3 second images to appear
after the initial 1min 10sec wait.

I use WinSCP to monitor the RPi output and pass the images to XnView for EXIF
readings etc..

Can anyone confirm that the above settings do indeed produce 4 minute exposures?

I know that a minute plus wait for the first image is a pain but, if a series of
relatively ong exposures follows, it is worth it.
Attachments
Contact Sheet2.jpg
Contact Sheet2.jpg (56.23 KiB) Viewed 418 times
Contact Sheet-1.jpg
Contact Sheet-1.jpg (60.29 KiB) Viewed 418 times

User avatar
HermannSW
Posts: 3822
Joined: Fri Jul 22, 2016 9:09 pm
Location: Eberbach, Germany
Contact: Website Twitter YouTube

Re: HQ camera timelapse startrails

Thu May 28, 2020 3:21 pm

You did not specify mode 3, you need to do so.
And your command misses some other options stated required by RPF people in long exposure thread.
Find the options needed in the raspistill command in this posting, it took 3m21.808s real time to capture 200s:
viewtopic.php?f=43&t=273358&start=25#p1664566
This (similar) posting command successfully captured 239s exposure (each 8x8 matrix dot lights for exatly 4s, one at a time):
viewtopic.php?f=43&t=109137&p=1667307#p1666929
Image
Previous posting in this thread showed that long exposures can be captured seamless with raspistill timelaps option .
Look at top row left point of captured 16 lit leds, position moves one position to left and two rows down for the next capture.
That is 2*8-1=15 positions, caused by one led lit for 4s.
https://stamm-wilbrandt.de/2wheel_balancing_robot
https://stamm-wilbrandt.de/en#raspcatbot
https://github.com/Hermann-SW/Raspberry_v1_camera_global_external_shutter
https://github.com/Hermann-SW/raspiraw
https://stamm-wilbrandt.de/en/Raspberry_camera.html

Return to “Camera board”