71 replies to this topic

[SeymoreStars]

Like the other fellow above, I can't recommend PixInsight for astrometry. Right tools for the job, etc.  The #1 piece of software used to be Charon, then it was superseded by Astrometrica. These days, it's very much Tycho Tracker.  

I purchased Tycho Tracker a couple hours ago, cut me some slack.

Edited by SeymoreStars, 27 December 2024 - 11:36 PM.

[SeymoreStars]

Fortunately a new laptop with NVIDIA GeForce RTX 4060 card is available. Now if the clouds would go away.

[SeymoreStars]

Seeking advice, which instrument is better suited for MPC asteroid work, Edge C14 (F1.8) with hyperstar or a planewave CDK17(F6.8)?

 

Both instruments would have a full frame B&W CMOS camera.

 

Thanks!

[Xilman]

Personally I would go for aperture. The more you can see the more you can measure, within reason.  This does assume, of course, a plate scale which is not too small.

 

No need for a wide field of view in these days of Gaia.  There are always well-measured stars within a minute of arc or so.

 

 

Doubtless others will chime in with counter-arguments, which is A Good Thing.

 

Paul

[pvdv]

They are both great tools. The Hyperstar C14 would be better for surveys (but plate scale will limit astrometry precision), the CDK better for photometry/astrometry. So it boils down to what you want to do and, imho, you already own a dream setup.

Aperture and sky quality are essential for a lot of what you could do. Now, before someone gets angry, smaller instruments also have a role to play, for example for lightcurves of objects too bright for bigger scopes. 

 

There is one other aspect to consider as well: the processing pipeline and IT aspects. I have acquired 100+ GB of data over the last two nights. I was, among other things, going for relatively faint and fast (moving between 25" and 60" per minute)  NEOs confirmation (I apparently failed, but have to look at the data more carefully). 
Recovering a slow moving mag 20.5 is much easier than recovering a fast one, even if one discounts the uncertainty on the position. It boils down to how quickly you can reach your limiting mag and I definitely would have liked to have more aperture. My 3080 field laptop was barely able to keep up with the data flow, and only because I was limiting PA and speed range. If you are lucky enough to have many clear night and collect a ton of data doing stuff manually quickly becomes a chore.  I am plagued with very poor skies, at most a couple of clear nights per month, 3 consecutive clear nights only once in the last four years, but if I had more the amount of data to process would be really insane.

https://www.minorpla...rm_tabular.html

But, fortunately, there are many other things you can do.

[SeymoreStars]

Thanks Xilman and pvdv for the advice, I'll stick with the CDK17. 

 

With Tycho Tracker fully installed I processed my data, it turns out none of the asteroids in the FOV were visible.

I will run through the DS1 example provided  with TT today.

 

The TT Calibration module implies that Master Darks, Flats and BIAS have been previously stacked. I couldn't find where within TT that is done.

 

Does TT stack and create Master Darks, Flats and BIAS?

 

Thanks again very much!

Steve

[pvdv]

I use a master dark and master bias from my standard calibration library.
I take flats on the night I do the capture and create a master flat. 

That being said, Tycho does a very good job generating a pseudo flat and removing hot pixels. The constraints are a bit different than for pretty pictures.
I usually run Tycho live on my captures, in 30 minutes batches, without flats so if I see something interesting I can return to the field a bit later.

Very strange that you didn't find anything: if you want to share your captures either publicly or privately, I can run Tycho on them if you want. 

I would indeed suggest running through the examples step by step then, once you have done that, move to Express Mode (and then later batch mode).

[SeymoreStars]

I use a master dark and master bias from my standard calibration library.
I take flats on the night I do the capture and create a master flat. 

That being said, Tycho does a very good job generating a pseudo flat and removing hot pixels. The constraints are a bit different than for pretty pictures.
I usually run Tycho live on my captures, in 30 minutes batches, without flats so if I see something interesting I can return to the field a bit later.

Very strange that you didn't find anything: if you want to share your captures either publicly or privately, I can run Tycho on them if you want. 

I would indeed suggest running through the examples step by step then, once you have done that, move to Express Mode (and then later batch mode).

I will post the images to the google drive soon. Thank you!

[SeymoreStars]

I took a sequence of 10-60 second images, one after the other. Images 1,4,7,10 have been uploaded to google.
I can provide all 10 if it's worthwhile.

They were plain-vanilla calibrated with pixInsight.

 

I just processed the "stacked" master of all 10 images and will upload that, it shows more detail.

Here's the link - https://drive.google...SO4?usp=sharing

 

Thanks

Steve

[SeymoreStars]

For the DS1 example, no known objects are showing either.

 

1) Proper observatory is set.

 

 

Attached Thumbnails

• 
• 

Edited by SeymoreStars, 29 December 2024 - 09:57 AM.

[pvdv]

Tycho doesn't work the way you expect it to.

It will take a series of images (a minimum of 11) and will apply synthetic tracking to them. Basically it will try 10s of thousands of stacks for all movement orientations and speed specified.

You can think about it as comet stacking in pixinsight, except that it tries _all paths_ (essentially stacking shifts) the comet could take.

Some paths will will match the path the asteroids took and will therefore increase the SNR on the asteroid which would otherwise not be visible because standard stacking would essentially reject it.

Try with older images where you have a series of 11+ images. Even if they are in color (Tycho can debayer). A standard series of 12*5mins exposures will do fine (don't try more at first because it will increase computing time exponentially). An try to take a field near the ecliptic so you have plenty of asteroids 

The synthetic tracker will yield a series of potential track, compute their confidence then use the match known objects menu.



 

[pvdv]

You are on the wrong orbit . 
I'll try to post a small step by step example. But the videos are great.

[SeymoreStars]

Okay I will watch all the videos. Thanks for the assist.

 

Steve

[pvdv]

Very short Tycho tutorial (very incomplete, all errors mine)

Load files - at least 11

Configure for automatic reload of processed files - the processed files will be immediately reloaded in file manager for further processing (change default) 



Express process - launch the process with your options. Uncalibrated files here tycho will simply kill hot pixels and apply a synthetic flat

Image Evaluation - I like to have that so I know what I can expect

Configure Synthetic Tracker - configure to taste/need. Defaults OK to get started.

Launch Synthetic Tracker - launch processing on GPU



Get Tracks - potential tracks are returned - this is blind - Tycho does not know what is what

Compute Confidence - assess the quality of the data

Examine results - load known objects  (note: some easy objects have low confidence because of the nature of the Pleiads field and the huge spikes of the Newtonian)
 

 

[555aaa]

The more aperture the better. I like Astrometrica better than Tycho but the solver in Astrometrica is very picky. Tycho has a steeper learning curve. I use both at the moment, Tycho for fast moving objects and Astrometrica for slow. Also Astrometrica has problems with very dense images such as from the RASA 11.
The best non custom scope is probably the RASA36 but anything large aperture is good because the whole metric is how fast you can get to the faintest magnitude. Focal ratio only gives you field of view, it has virtually no effect on SNR. With the CDK17 you are limited to objects with low error if you are doing NEO confirmation or recovery.

Also, almost always you are doing either synthetic tracking or a “track and stack” using a known position angle and motion rate. All the slow moving asteroids are gonna get swept up by the all sky surveys so there is no scientific usefulness for them (except light curves). Fast moving asteroids need coverage because there are only a couple dozen observatories doing follow up and there’s like 100 objects needing follow up every night.

Edited by 555aaa, 29 December 2024 - 11:26 AM.

[555aaa]

Even better than one RASA36 is four RASA36es

[pvdv]

Even better than one RASA36 is four RASA36es

This is why I am ultimately grateful to live under **** skies: I get to keep my money!

Seriously, your systems are insane in the good sense of the term.

Edited by pvdv, 29 December 2024 - 02:48 PM.

[555aaa]

Thanks for the kind words. Just installed a motorized tilting system on the four RASA36 rig so that each scope can be moved a few degrees by itself.

Edited by 555aaa, 29 December 2024 - 11:59 AM.

[SeymoreStars]

Small success.

 

As mentioned previously 10-60 second images were taken FOV (0.7° x 0.47°) Binned 2X2 Resolution = 0.53" per pixel

 

Question - Binning 4X4 will reduce the file size, but will it brighten the image
and improve SNR?
Binning 4X4 Resolution will be = 1.06" per pixel

 

I must have skipped the Tycho Tracker "Align" step because everything worked as expected.

Thanks for the support getting over the hump.

 

The objects found by Tycho Tracker coincide with the PixInsight objects.

Attached Thumbnails

• 

[555aaa]

What FWHM are you getting now when you look at the star profile using something like AstroImageJ (that is for an UNbinned image).

[SeymoreStars]

What FWHM are you getting now when you look at the star profile using something like AstroImageJ (that is for an UNbinned image).

Unbinned the QHY600M file size is 120meg. I haven't used that mode in ages. I'll have to get back to you.

 

Thanks

Steve

[pvdv]

The objects found by Tycho Tracker coincide with the PixInsight objects.

Yes - they both use the same data set to annotate the image. (although Tycho's data set is constantly updated.

Important note: at this point, neither Tycho nor Pixinsight have found anything beyond a plate solution. They just have added a mark where the asteroid should be.

[SeymoreStars]

 

That being said, Tycho does a very good job generating a pseudo flat and removing hot pixels. The constraints are a bit different than for pretty pictures.

 

I ran Tycho on the images straight from the camera and it did an excellent job of calibrating them without flats,

Edited by SeymoreStars, 30 December 2024 - 10:47 AM.

[pvdv]

Pixinsight, in its default configuration, is shockingly bad at preserving the interesting signal.

Same test field, same data set.

On the left, the result after a default wbpp calibration by pixinsight (correct bias, correct flat, correct dark). The shot_name_c is used. Tycho calibration is disabled.

 

On the right, uncalibrated frames with hot pixels removed and pseudo Tycho flat.

You have the names of the objects present in the field (by catalogue/orbit) and the confidence level of the Tycho detection.

People who discourage the use of Pixinsight do have a point.

Edited by pvdv, 30 December 2024 - 12:01 PM.

[italic]

Pixinsight, in its default configuration, is shockingly bad at preserving signal.

Same test field, same data set.

On the left, the result after a default wbpp calibration by pixinsight (correct bias, correct flat, correct dark). The shot_name_c is used. Tycho calibration is disabled.

On the right, uncalibrated frames with hot pixels removed and pseudo Tycho flat.

You have the names of the objects present in the field (by catalogue/orbit) and the confidence level of the Tycho detection.


People who discourage the use of Pixinsight do have a point.

I've seen the recommendation but I've never seen the result that demonstrates why. Now I wonder what steps in wbpp need to be changed to get the same results. I imagine rejection and stacking method are the big things, but there are lots of settings to play with.