Plate solving for dummies??

[Lee]

After a fruitless search for a faint fuzzie, I realise I'm in need of a plate solving application....

From a lot of searching online, there seems to be a bunch of options, all seeming quite confusing as to exactly what is needed, and how it works....

What I'd like is to be able to take an image of sky, and have the application tell me the centre of the image (which from what I can tell is the minimum of a plate solve?)....
Of course a bonus would be a link to my G-11 Gemini (L1).....

Can anyone suggest a solution to my needs?
I don't mind paying, but don't feel I need the all singing all dancing MaximDL?? I've asked to join the Elbrus Yahoo Group, but haven't heard back yet....

Is MaxPoint what I'm after??
I use nebulosity for imaging at present, with a Meade DSI Pro II (an Atik 420 on the way..), from what I've read, the Atik Capture software can link to MaximDL +/- MaxPoint??

[allan gould]

The application that you are after is Elbrus which is free and used within sequence generator pro which you can get a 45 day free trial. It does exactly what you want and much much more. See this site for more information and videos. http://www.mainsequencesoftware.com/
Hope this helps

[multiweb]

Another freebee is UNIMAP (http://larryo.org/astronomy/software/unimap/download.html).

[Octane]

MaximDL with PinPoint LE has been a great success for me.

H

[Terry B]

I use elbrus as well. It talks to ascom very easily. Not sura about G11 but I'm sure someone will know.
The database files are here
http://atlante.org.es/catalogos/elbrus/
You need to download all of it and extract it into a folder..

more info here
http://astrosurf.com/pulgar/elbrus/elbrusin.htm

I'm not sure why the elbrus group hasn't contacted you yet but there hasn't been any activity for over a week.
I can send you the latest exe files if you like. The zip file is only 512k but this makes it slightly too big for me to attach here. Send me your email and I can send it to you.
Cheers

Terry

[swannies1983]

Lee, I was in the exact same position as you. I have been imaging with a RC8 on an EQ6. The problem is that my mount isn't GoTo compatible. I currently use settings circles but I wanted a more accurate way to find faint fuzzies. Plate solving was the way to go.

After doing some reading, I downloaded Astrotortilla http://sourceforge.net/p/astrotortilla/home/Home/ and it does exactly what I need.

You need to download star catalogues, which is what AstroTortilla will look through to match star patterns when it comes to plate solving. The amount of catalogues you need will depend on your image scale. The smaller the scale, the more catalogues (larger download required). See the manual on the home page to guide you.

When you install AstroTortilla, it will ask you which index files (star maps) to download. It's generally pretty slow but you can download index files directly from here (http://broiler.astrometry.net/%7Edstn/4000/). For my image scale, I had to download 4004, 4005, 4006, 4007, 4008 and 4009 series. Each series covers the entire sky. However, 4008-4019 has only one file to download and each covers the full sky. 4005-4007 are made up of 12 tiles (see the hp.png in this (http://broiler.astrometry.net/%7Edstn/4000/) directory to show you what regions of the sky each tile covers). Series 4000-4004 are made up of 48 tiles (see the hp2.png in this (http://broiler.astrometry.net/%7Edstn/4000/) directory to show you what regions of the sky each tile covers). So, you can reduce your download by only getting the tiles that cover what is visible in your sky. Let me know if this is confusing!

So once it's all installed, this is my workflow:



Polar align
"Sync" my setting circles to a star
Look up object of interest
Move scope and match setting circles RA and DEC based on RA and DEC of object
Connect camera
Take a 5 sec shot at ISO1600 (only need a 5 sec shot). I use jpeg files.
Go into AstroTortilla and connect to my mount via ASCOM*
Click Tools and then GoTo Image. This allows me to load my 5 sec image.
It should start to plate solve. It generally takes 30-60secs but can take up to 2 mins depending on how many index files it needs to solve. Once solved, I move the mount in RA and DEC and plate solve again until I'm happy the object of interest is centred.

A few things that will affect solving time:




Scale Minimum and Scale Maximum. I generally set this to +/- 10% of my maximum image scale (based on the longest dimension of my sensor)
Sigma. This is how many objects it detects. When it starts plate solving, it should give a number of how many objects it detects. Aim for about 20-200. Anymore, and it will take a long time to plate solve. Any less, and it may not solve at all. Increasing the sigma, reduces the number of objects identified.

*You can use AstroTortilla to move you mount after plate solving to the correct RA and DEC. My ASCOM driver is old (no updates since 2006) and it doesn't really work so I just manually move the mount, plate solve, move the mount, plate solve etc until I'm happy.


I hope this helps!


Dan

[Lee]

Thanks everyone for your detailed replies.... Will give these options a try....
as a test last night I slewed to M83, and took a frame, uploaded to http://nova.astrometry.net/ .... to my delight, it reported M83 was actually near frame centre.... the old Gemini is pointing better than I thought!

M83 -- RA 13h 37m 00.9s; DEC -29deg 51' 57"

centre of my frame reported as - 13h 48.711s; DEC -29deg 56' 23"

only missed by several minutes.....

[swannies1983]

Astrotortilla uses the Astrometry.net engine. But you don't have to wait for someone to solve it for you ;). How long did it take for your image to get solved?

[Lee]

Took about 30s to solve I'd say.... that was just uploading a fit file, anf giving it no further info....

Seems faster if you give it some limits, like a RA/Dec radius, or approx image scale....
I experimented by cropping little squares of random star fields from some jpg images I'd done, and just tell the engine the field is say 0.01 deg to 1 deg, seems to take 10s or so for a solution....

I'm still going to get a self-contained solution going, but astrometry.net is very cool.... I'm continually amazed by what we have free access to, and the high prices we need to pay for other stuff!

If anything - it seems an easy way to get a precise arcsec/pixel measurement, just upload an image you've taken with whatever OTA/CCD, and it will tell you....

[alistairsam]

Maxim and pinpoint works great from me, the useful bit is when your object is slightly off centre, you right click and choose centre here it slews and centers the object.
Else Elbrus should work on its own, or its great with SGP.

I think it'll be more useful if you use it with a scope control software as then it can auto correct and centre after the plate solve. Saves so much of effort with manual centering if you just had plate solve.

[Octane]

The other advantage of MaximDL with Pinpoint is that it solves the image after you take it, and stores the RA/DEC coordinates in the FITS header.

Why is this important? Because when it come to stacking, you can use the astrometric method of stacking which is the most precise of all the methods.

H

[Terry B]

Elbrus also does this.

[Octane]

Good to know. Thanks, Terry!

H

[Lee]

Have had a look, asked a few questions - SG Pro/Elbrus looks like what I'm after for sure, thanks..... $100 seems cheap at twice the price! :)

[allan gould]

Lee, you will not regret it. The main thing to get right is the arc sec per pixel for your camera and scope. Ccdcalc gives an accurate guide to this. Once an image is solved just use 2x2 any angle and the Arcseconds per pixel that you have determined and it will always work as long as your RA and dec are reasonably close to your target.
The yahoo group will respond almost immediately to any queries. There is so much more to the program but I'm sure you realize this as there are quite a few of us now using this program exclusively.
Allan

[Lee]

Even easier I think - I just take a frame, and get nova.astrometry.net to plate solve it, they tell you your exact arcsec/pixel and angle..... I won't be able to play with it yet, am waiting for my Atik 420 to arrive, SG Pro won't drive the old Meade DSI I'm using.....

[Octane]

Can SequenceGenerator Pro handle the STL with its built-in guide camera?

H

[allan gould]

H to be honest I don't fully know but if it's ASCOM controlled then I assume both can. Send a question to yahoo groups at http://tech.groups.yahoo.com/group/mainsequencesoftware/
In the home page it says it supports sbig cameras. There is a 45 day free trial with no restrictions, so have a go.
Allan

[Lee]

I'm impressed - fired up sequence generator Pro/elbrus last night....

Turned on the Gemini, cold started, no model.
Typed in the RA/Dec for M83, automatically slewed, plate solved, correction slew, solved - bang, on target.... did this for several other fuzzies just for laughs....
No more craning my head into eyepieces dangling off my GEM trying to build a reasonable pointing model! Now I just need more aperture.... :)

[Terry B]

Well done.:D

[Octane]

Plate solving is a wonderful, wonderful thing.

It is essential when building mosaics! :)

H

[alistairsam]

Hi,

I don't mean to detract, I know SGPro can automate flats, but can you program it to take flats at a particular time, say 5am or 5:30am for the appropriate ADU range and then park the mount to avoid accidental pointing at the sun when it rises?
I've tried running the trial, but always had windows errors.

I can currently auto park using eqascom, but it would be good if the capture software can do that after the flats sequence.
and do you really need a T-shirt if you're taking flats at dawn as the sky would be pretty evenly lit if there are no clouds.
Thanks

[Lee]

I cant see why you couldn't program it to point at a particular RA/Dec and take short flats exposures, then park.... I'm pretty sure it will do things at various times also....

[Lee]

Another thing I found very useful was focusing, I just pulled out Starry Night, looked for a little asterism of moderately bright stars, it slewed there/solved/corrected in about 60s.... love it!

[allan gould]

H
Apparently this is going to come in the pipeline. Have a look at the yahoo site for the latest as I think they are seriously thinking of implementing this as a new step.
Another thing is it now uses astronomy.net in combination with Elbrus to plate solve. If Elbrus fails to plate solve it will then do it blindly over the Internet, sync the mount and then center the mount to where you want it.
Allan

[allan gould]

Yes it can if you download the program you will see definite start, stop and park times are available?
It has a routine to automate the acquisition of flats at different binning settings and populate the sequence. All done automatically for a user defined ADU level.
Allan

[batema]

I think what i understand from this id plate solving programs allow you to take a photo of a section of sky and the programs recognise where you are and then you can ask the program to goto an object and when you take a photo it will point the object central. Or if you have already taken an image and want to continue the next night this plate solving will move you scope to exactly where the previous image was taken with relation to the stars. Am i sort of right ?????

Mark

[Lee]

The 'plate solve' is software that analyses the stars in an image, and works out where you are - ie the RA/Dec of centre of frame.
The software can use this to tell the mount where to go....

With SGPro - I manually typed in the RA/Dec of M83, and told it to goto it - it slewed the mount to this area, took an image, plate solved that image, then issued a correction slew to the mount (as the initial slew was off) - then it takes another image, plate solves that, and issues further corrections until you are happy with where you are pointing.
You can also give the program your image from a previous session, it can plate solve that one, then line up your scope exactly on that spot again....
It's VERY cool.... :)

have a play by uploading images (or crops from images) to nova.astrometry.net - you can see the info you get....

[batema]

Thankyou Lee.

[gbeal]

I'm about to re-plunge here, and was to embarrassed to seek help.
A while back I used Maxim and the USNOA catalog to solve and slew, when I used a "real" CCD camera.
Since using just DSLR's of late I had considered I was unable to solve. Maybe, maybe not, so I ask.
I use Cartes de Ceil, and connect the NEQ6 to it with EQASCOM. I am NEVER connect to the Internet.
Given I don't use any other listed supported/compatible cameras, I assume I can't use SGP?
My cameras are either a Sony or Fuji DSLR.
If I get serious again it will possibly extend to an old and cheap QHY8 so using Maxim is an option, but right now I'd like to know if I can solve with the gear I have. Both cameras can present an image in the PC so the question is can I work with that as is?
Gary

[allan gould]

Gary
SGP will work with an ASCOM mount ie eqmod and DSLR or a qhy8 camera. The plate solving is via Elbrus which is free and doesn't need an Internet connection. There is a 45 day free trial so you can use it with your system and make sure it all works. There is a yahoo site if you have any questions or you can pm me if you want
Allan

[swannies1983]

Gary, see my earlier post. I image with a modded Canon 30D and plate solve using Astrotortilla, which is free and doesn't require an internet connection (once you have downloaded the star catalogues).

[gbeal]

Allan/Dan,
Appreciate the replies guys, really.
I've had a bit of a glance at both and if anything I reckon the sticking point is the fact that my choice of DSLR can't be controlled by the required software. Neither SGL nor AstroTortilla List Sony or Fuji as controllable. I have looked at the PDF from AstroTortilla and it appeared that you can simply point the file selection to a point on the hard drive, so maybe I could get around it by that.
It would be nice if someone else out there was using either/both and without a actually using the camera control software, Maxim/Nebulosity for example.
Gary