Thoughts from Starlight Cascade

The allskypi camera upgrade from 0.6 to 0.8 took about 10 minutes and the new software is *amazeballs*

A new web GUI interface for settings which has more info about the settings, the much better dark frames at temps from 23-30C. Unfortunately, I covered the dome with a dark hoodie jacket and that actually insulated enough and blocked the vents enough that the sensor temp actually rose through the night instead of dropping.

Will have to do it again with a smaller black cover that does not cover vents. even so the resulting dark frame at the actual temperature is doing a much better job at removing hot pixels than before.

I will now look at adding outdoor temp and humidity as well to each image.
Two quick sample shots from this morning.

The first at 05:33:06, before dark frames, 60sec exposure, hot pixels EVERYWHERE

The second, at 05:40:27, after using a dark frame taken at 29C, a much better image.

Orion is just on the bottom, Capella near the centre. Cassiopeia near the upper right.

Can’t tell if that is M31 Andromeda Galaxy or not.. there was a lot of cloud starting to come up from the south.

Regulus and Leo are over to the left side. The Big Dipper/Plough upper left.

Great first light results!

The woman on the presentation, Linda Thomas-Fowler, spent about $650USD on the project, mainly for buying a higher end ASI camera, and a new raspberry pi 4, etc. If you had spare bits lying about it would be more on the order of $100-200Can

software project available here: https://github.com/thomasjacquin/allsky

Will be trying to get more dark frames tonight once the sensor temp goes below what we already have… 23C

This was the last images of the Jupiter imaging test from Friday night, after 2000 EDT and into the next day UTC.
So.. the details… The ZWO ASI290MC camera has been giving unsatisfactory images of late so I borrowed an ASI120MM to try out and try to narrow down what may be happening. So with 0 barlows on the native 1800mm focal length VC200L OTA on a Skywatcher AZ-EQ6GT mount, I started with 15 minutes of trying to find Jupiter… before taking the lens cap off.
Note to self.. take off the lens cap.
OK.. found Jupiter, exposed using firecapture so the histograms come up at 80-85% with default gain and gamma. This turned out to be a 4ms exposure. I normally take a 30 sec firecapture unaligned followed by a 30 second firecapture aligned run to have a baseline for comparison of the seeing that evening.
Last night was average seeing and average transparency… until the clouds came along.
After that a 180sec exposure (old rule of thumb that I used on an older telescope.. the max exposure before Jupiters rotation moved a feature from one pixel to the next.. you don’t want that), I then repeated it all with a x1.5 barlow (there was some dust on it!) at 10ms and then with a x2 barlow (dust on it as well) for a 20ms exposure.

During the very last run (the image above), terrestrial clouds were moving across the face of Jupiter, causing it to totally disappear at times from the display.
I kept a small Region of Interest of about 700×700 pixels, to increase the framerate because of the smaller USB bandwidth needed.
The files were moved from the imaging laptop to the desktop inside the house and today I processed them.
Again, first step is to use PIPP to crop and center at 600×600 pixels and also convert from .SER to .AVI format video.
Then Autostakkert! v3 to analyze and stack using the best 5, 10 and 25% of the frames
I had 11 imaging runs in total and processed all of the best 25% of those. Then I did it again using the best 5%.
The image above shows on the left, the best 5% version and on the right the best 25%.
Possibly because of the clouds, the best 5% is visibly better than the best 25%. There were approx 7500 frames used from the 180sec run, so the best 5% had approx 375 frames and the best 25% had about 1900 frames.
Focus was still difficult but all in all I was satisfied with them. Jupiter was still low in the much, over the Kingston SkyDome (a light pollution dome over the nearest city). Azimuth 147 degrees and altitude 24 degrees.
Jupiter would be at its highest altitude around 22:06, past by bedtime 🙁 That is the end goal however.. picking a night where I can stay up that late.

This is the original best 5% along with the filename which contains many of the processing details.

This is the original best 25%.

I think I will need to use both cameras during the same run to compare them… seeing varies so much from night to night or even from start to end of night that looking at them from different nights yields no new insights.

I too spent a few hours last evening/night.

Mark had kindly repaired the JMI remote motofocus USB interface and it worked great! Having numbers to record for focusing with no barlow, x2, x2.5, x23 etc will be awesome!

The new cable that Mark built also works! Yay! My windows 10 laptop can now communicate with the Skywatcher AZ-EQ6 GT mount! Yay! It requires the ascom driver for “skywatcher telescope” however, not the “EQmodHEQ5/6″ that we had tried out earlier on the zoom session.. go figure!

So. I downloaded the latest PHD2 and installed it. No clear cut process to follow so I just turned it on, told it to use the primary camera (ASI290MC) and the scope with no barlows (1800mm) and it did some dark frame stuff.

I’ve been told that PHD2 will let me polar align the mount easily.. so that it what I was aiming for last night.

But first I had to park the scope at home, restart, do a 2 star alignment. On the first star (Arcturus) I had to use a 40mm eyepiece to find it as it was a few degrees off. Then I had to achieve focus. OK so far.

star #2 Dubhe. hmm don’t remember which one that was so cycled through the other 10-15 choices and did not even recognize most of them. Had to fire up stellarium to positively identify the star… and it had problems… it stalled on startup as it was fighting for telescope control with some other program … arrg.

Finally 10 minutes later I identify Dubhe as the last star in the Big Dipper dipper/bowl.

Align on it. ok. good align it says.

Then I try out the remote focuser to try and achieve the best focus possible. hmmm. not quite symmetric going in and out of focus. OK then. try the 8” bahtinov mask! Wow. The realtime result looks nothing like the sample images you see on the net. So another compromise and will use the best focus possible, which was not perfect.

OK them back to PHD2. Tools Polar drift align. ok.. it says move the scope to within 6 deg of polaris. Wow.. every try to get a scope to go to polaris when it is equatorial? It just did not want to. So down on the ground looking through the newly aligned telrad to get it within 1/2 degree. ok. PHD says pick a star. I did. click on start, wait for awhile, click on stop then physically move the mount in RA and Dec to put the seleted star into the circle at the end of the line. Hmmm the line goes offscreen and there is no circle. Tried to drive the scope with the hand controller.. nope.. other issues.

try again and physically move the mount.. still no circle and no end of line. Hmmm.

read through the online help again.. still nothing.

I am wondering if it requires a much lower power setup, ie not the primary 1800mm FL of the primary scope but something much wider field of view? It talks to the camera and mount and gets info from them to calculate a field of view and to tell me where to move the mount to in the end. I don’t think it will work on the mallincam video output of the actual finder/guide scope and the frame grabbing usb adapter in the middle.

But I may have to try that i nthe future.

So.. failure on the PHD2 front. )@#$& this lets get some Jupiter images. Tell the hand controller to go to Jupiter, it misses by a few degrees. arrgg. Doesn’t matter.. its getting late and I just want some image runs.

Center Jupiter with the eyepiece, then the camera. wooho.. Jupiter.. long time no see.

hmmm.. seems to be moving awfully fast… tracking is pretty bad. Try a reduced region of Interest (about 600×600 pixels) and the best I can do is about 30 seconds.

OK.. using firecapture I invoke the “autoguiding” feature that I used to use with the Meade LX200GPS, where firecapture would track the centered object and issues commands to the mount and keep it centered.

30 minutes later, everything I tried failed… will have to spend a lot of time reading manuals again and experimenting with variables (options not stars). So I took a few more 60 seconds runs at full resolution and called it a night.

Ooops! Forgot to do the one thing I set out to this evening… try to polar align the scope with the polar axis eyepiece. OK.. nice red LED inside.. the handcontroller can change the brightness. awesome.

Looks for the big dipper (not easy when lying on the deck, glasses off, eye glued to the bottom of the mount essentially. I discovered that the pier should have been a few inches higher, just for this)

Look for Cassie, rotate the template to match approx by eye. (next time I do this again I will use the actual RA numbers and dial in the template much more accurately). ok.. loosen the clutches.. hmm which star is Polaris?

loosen the mount and physically rotate it.. I see a star? Is it polaris? who knows.. we are going to use it anyway!

OK. move the mount in RA and Dec to put the star in the circle. Done!

tighten clutches and mount. ( I think I did that). Shutdown and go inside.

Phew!

What a night. Now processing maybe 6 imaging runs of Jupiter.. will see how they look.

No mosquitos, watched the CFL game Montreal at Ottawa in the observatory.. Nice!

Next time!

a more accurate polar align with the polar eyepiece method. Learn the stars around polaris first!

Get some paper star maps! Do a 3 star align! Park it and restart it and goto anywhere. Are we even close?????

If so start looking in more detail at PHD2 mount alignment. Try it again maybe. Try it again with the mallincam finder/guider.

Start looking at using software to command and control and move the mount to targets. See how accurate it is.

Personally I would like it to find Jupiter in the full field of view of the camera without me hunting for it.

And the initial results from a lot of bad conditions, including:

Jupiter was still a few hours away from maximum altitude/zenith.

Jupiter was effectively in the Kingston SkyDome

seeing was average

transparency was below average

Quick processing run… will get better with practice.

All were initially and mistakenly process with only the first 2500 frames.

Io and Europa showed up after the registax processing. Nice!

This is the animated 6 runs over say about 20 minutes.

I predict that when I have time to reprocess this in tomorrows RAIN that it will be much better!

This is one of a few imaging runs of Saturn taken Friday evening. The new mount is still not polar aligned, so there was more drift than I was happy with. This is the first time the x2 antares barlow was used. Focus is still quite soft (I still do not have a full length extension on the electric focus controller) and that will get better.

Again, the camera would not come to focus with the x2 barlow, but it did with the 90 deg diagonal. So I am still all confused up as to orientation, flipx flipy choices. I had to go into stellarium to see what Saturn should look like and edit it after the fact.

This was a 240 second run with 60ms exposures. The best 25% of 3970 frames were used. I love the scale of 3600mm focal length… Saturn is actually larger than a pinprick!

So, altitude was still very low and Saturn was still in the Kingston SkyDome… so this again is a benchmark image, where future ones can only get better 🙂

Also attached is an image from Allsky1pi (with its new wider angle lens) camera system actually showing Saturn at the same time as the telescopic image was taken, approx 22:53 EDT Friday evening. Saturn is 19deg altitude and is located centrally in the Kingston SkyDome. If I had better prepared with a nap, I could have gotten up around 02:00 or 03:00 when it would have cleared the light pollution.

The Allsky1pi camera system is slowing coming back to normal. It stopped working a few weeks back, from a suspected overheating of the camera and the raspberry pi. The housing was cut open and an open mesh was installed at the bottom to allow for better airflow.
Configuration files were also changed and changed back and changed again.
Last nights run was reasonably successful, back up to 60 second exposures with no moon up, one image every 60 seconds, and the sensor temperature not wildly obscene.

Still to come are more and better dark frames and to figure out some way to automate dark frame processing into the system, based on the sensor temperature.
That should really clean up the hot pixels.