Questeions about ACP functionality

[Roy Salisbury]

I'm re-evaluation my options for automated imaging at my observatory (ACP, CCDCommander, CCDAP, ect) and was wondering if you might be able to answer some of these questions about ACP.

1. Is there a way to have ACP cancel taking light images at dawn so flats can be taken properly (or even cancel dusk flats at astronomical twilight)? I underestimated my imaging session this last weekend and ACP was still happily "trying" to take images at 6am with the sun already up. So my flats were skipped. Note: I have only been using the web interface for setting up plans, so perhaps the standalone planner has this option that I am not aware of.

2. I really like the System Status display. It’s something that I started messing with as a remote display for my web site awhile back, but the ACP one has given me other ideas. Is it possible to somehow interface with the necessary ACP object from "outside" the normal ASP pages so I could get that information? I would probably need to write a web service that my web site could then communicate with as a proxy to ACP. I’m sure I could get the necessary information from MaxIm's objects, but might let ACP do the heavy lifting. I'm only looking at the "System Status" info. Basically I want to be able to show on my website what activity is going on at the observatory. Heck, even a "read only" version of the system status that could be linked to from an IFrame would be good.

3. G2V star selection and calculation. I'm pretty sure this one is not there, but I'll bring it up for a future enhancement. Would be nice if at the beginning of a session the system could select a G2V star, calculate the necessary RGB values and either store them in the log (for later use in stacking) or even possibly shorten/extend the exposure times based on the values.

4. Is the weather server object documented somewhere? I could have just overlooked it. I have my own full weather station setup at the observatory and would like to write a COM object that interfaces with my web service.

5. This is sort of related to #2. Any chance that you may be working on a version of ACP that exposes ACP via SOAP? If not, do you think it’s possible for a 3rd party (such as myself) to do it using the ACP COM objects, or are they too tied to a web (ASP) model.

I have some other questions, but I'll keep searching other posts to see if they have been asked already.

Regards,
Roy (www.hualapaivalleyobservatory.org)

[Bob Denny]

1. Is there a way to have ACP cancel taking light images at dawn so flats can be taken properly (or even cancel dusk flats at astronomical twilight)? I underestimated my imaging session this last weekend and ACP was still happily "trying" to take images at 6am with the sun already up. So my flats were skipped. Note: I have only been using the web interface for setting up plans, so perhaps the standalone planner has this option that I am not aware of.

Philosophically, ACP does not do planning, just execution. There are some exceptions (the #quitat directive) but in general, it's up to you to plan your night out ahead of time. This was a definite design decision. I can discuss the decision with you on the phone if you wish, in order to give you a clear idea of the reasoning behind that design decision. Feel free to call +1 480 396 9700 business hours M-F MST (UTC-7).

2. I really like the System Status display. It’s something that I started messing with as a remote display for my web site awhile back, but the ACP one has given me other ideas. Is it possible to somehow interface with the necessary ACP object from "outside" the normal ASP pages so I could get that information? I would probably need to write a web service that my web site could then communicate with as a proxy to ACP. I’m sure I could get the necessary information from MaxIm's objects, but might let ACP do the heavy lifting. I'm only looking at the "System Status" info. Basically I want to be able to show on my website what activity is going on at the observatory. Heck, even a "read only" version of the system status that could be linked to from an IFrame would be good.

Anything is possible in software! Just joking sort of. Yes I can point you in the direction of the logic that provides the system status display. It is composed of a large HTML/Javascript/AJAX based chunk of code running at the browser end, combined with a rather large Active Server Page (ACP) Classic back end which is all JavaScript, and which interfaces with intrinsic objects served by ACP as well as objects made visible by the observing logic script (AcquireImages.js) as well as its support library (AcquireSupport.wsc). You seem to know COM so this shouldn't be foreign stuff...

3. G2V star selection and calculation. I'm pretty sure this one is not there, but I'll bring it up for a future enhancement. Would be nice if at the beginning of a session the system could select a G2V star, calculate the necessary RGB values and either store them in the log (for later use in stacking) or even possibly shorten/extend the exposure times based on the values.

I'm not sure what you mean by "necessary RGB values". It sounds like the first step to automated processing of [L]RGB images into art images ("astro-imaging"). I'm open to discussion on this one, though my experience tells me that virtually no one wants the processing to be done at the observatory, scientists and artists alike. Only a tiny fraction of my customers, for example, even use the auto calibration feature of ACP, even though it would save them tons of image downloading (of cal frames). Anyway, tell me more! Perhaps there is a way to do what you want within the current capabilities of ACP.

4. Is the weather server object documented somewhere? I could have just overlooked it. I have my own full weather station setup at the observatory and would like to write a COM object that interfaces with my web service.

Start in ACP Help, Weather Safety, the bottom two sections Weather Servers, and Developing a Weather Server. In the first of these is a link to the interface documenation, which is in ACP Help, Scripting Guide, ACP Programmer's Reference, Weather Object. Also, in the WeatherComponents folder (under ProgFiles/ACP) there is a template WSC that implements the server aide of ACP's weather interface. There are also three WSCs that are complete weather servers. Under that is a weather simulator written in the old VB6 language, complete with sources.

5. This is sort of related to #2. Any chance that you may be working on a version of ACP that exposes ACP via SOAP? If not, do you think it’s possible for a 3rd party (such as myself) to do it using the ACP COM objects, or are they too tied to a web (ASP) model.

No SOAP :-) It's dead, like CORBA. I put in low-level support back in 2000, but I haven't used it for anything. I don't know if others have, so I have to leave it in. You can do virtually anything you want with ACP COM objects. Everything you see is done using COM objects, Classic ASP (built into ACP), AJAX, JSON, script components, etc. It's all in the open for you to use, and documentsd in the ACP Scripting Guide. I don't think you will find anything else out there that's anywhere near as rich and well documented. :-)

[Roy Salisbury]

Philosophically, ACP does not do planning, just execution.

I completly understand. I thought perhaps that some of the different areas/sequences might be broken up so that canceling the image run could be stopped on certain conditions (sort of the way the cloud senser can).

Anything is possible in software! Just joking sort of. Yes I can point you in the direction of the logic that provides the system status display. It is composed of a large HTML/Javascript/AJAX based chunk of code running at the browser end, combined with a rather large Active Server Page (ACP) Classic back end which is all JavaScript, and which interfaces with intrinsic objects served by ACP as well as objects made visible by the observing logic script (AcquireImages.js) as well as its support library (AcquireSupport.wsc). You seem to know COM so this shouldn't be foreign stuff...

Was sort of hoping that it would not be a lot of conversion involved (but was pretty sure it would be).

I'll continue looking at the ASP/HTML pages and track it all down. Might be worth it, might not. But its always fun to try .

I'm not sure what you mean by "necessary RGB values". It sounds like the first step to automated processing of [L]RGB images into art images ("astro-imaging"). I'm open to discussion on this one, though my experience tells me that virtually no one wants the processing to be done at the observatory, scientists and artists alike. Only a tiny fraction of my customers, for example, even use the auto calibration feature of ACP, even though it would save them tons of image downloading (of cal frames). Anyway, tell me more! Perhaps there is a way to do what you want within the current capabilities of ACP.

Basically what I need to do is calculate the spectral signature of a sun like star to get the color combine ratios. These ratios are later used when combining the [L]RGB images into a final colar image. This involves taking RGB images of the same object through all 3 filters using the same exposure setting. Its best to do this a few times to get a good average, and should be done before during or after the session becuase hight thin clouds can laed to different results for the blue filter. In a nutshell.. your calculating your white balance based on the session conditions so the final images looks better. Last time I manually did the calculations I had ratios of 1.06/1.0/1.29 (R/G/B) so during the stack the different channels were given a different weight.

I'm sure adding it to ACP would be easy enough, the question is does the necessary star clasification exist in the catalogs you have access too. It would sort of be like FocusMax's AcquireStar but based off clasification instead of magnitude.

No SOAP :-) It's dead, like CORBA.

I agree that CORBA is dead (heck, anything that Borland seemed to latch onto is sure to die). But SOAP? Well, that a different discussion for a different time.

You can do virtually anything you want with ACP COM objects. Everything you see is done using COM objects, Classic ASP (built into ACP), AJAX, JSON, script components, etc. It's all in the open for you to use, and documentsd in the ACP Scripting Guide. I don't think you will find anything else out there that's anywhere near as rich and well documented. :-)

I'll look deeper into the ACP COM objects and see what I can come up with. I've already automated portions of MaxIm so its not going to be that different (hopefully).

Thanks for taking the time to answer my questions and hopefully I will find everyting I need in ACP.

Roy

[Bob Denny]

I completly understand. I thought perhaps that some of the different areas/sequences might be broken up so that canceling the image run could be stopped on certain conditions (sort of the way the cloud senser can).

This can be done (along with all sorts of other things) for your specific needs via the UserActions hooks. These allow you to insert logic into the acquisition process at various points. See ACP Help, Customizing ACP, Adding to ACP's Logic.

Was sort of hoping that it would not be a lot of conversion involved (but was pretty sure it would be).

I'll continue looking at the ASP/HTML pages and track it all down. Might be worth it, might not. But its always fun to try .

To see the browser side of this, enter the authoring mode and open the System Status item. Then Edit it and you'll see all of the stuff there. The browser side logic runs on top of a framework called TiddlyWiki, and I have developed a bunch of plugins for it. They are self-documented and visible in the Authoring environment. You will need to work with the Authoring environment some in order to "get it". Follow the directions you see when first opening the Authoring environment.

Basically what I need to do is calculate the spectral signature of a sun like star to get the color combine ratios. These ratios are later used when combining the [L]RGB images into a final colar image. This involves taking RGB images of the same object through all 3 filters using the same exposure setting. Its best to do this a few times to get a good average, and should be done before during or after the session becuase hight thin clouds can laed to different results for the blue filter. In a nutshell.. your calculating your white balance based on the session conditions so the final images looks better. Last time I manually did the calculations I had ratios of 1.06/1.0/1.29 (R/G/B) so during the stack the different channels were given a different weight.

I'm sure adding it to ACP would be easy enough, the question is does the necessary star clasification exist in the catalogs you have access too. It would sort of be like FocusMax's AcquireStar but based off clasification instead of magnitude.

I would do this via observing plan. Acquiring this G2V data is no different than acquiring any other type of data. Make up a plan using a star that will be usable for a few weeks to a month and just run it. You could chain to it from another plan.

If you want to get more sophisticated like having this run multiple times a night, interspersed with other things, you're getting into the realm of ACP Scheduler. Look up the AAVSO Bright Star Monitor and Photometric All Sky Survey projects to see what's being done. Also the Sonoita Research Observatory is another all-out use of Scheduler.

I'll look deeper into the ACP COM objects and see what I can come up with. I've already automated portions of MaxIm so its not going to be that different (hopefully).

There is a LOT there, and the overall system design may not be evident. If yo have a specific thing you want to do:

1. Look to see if it is already done
2. If not, ask me about it. Maybe I know how to do it with the existing logic
3. If not ask me how one might go about it
4. Implement it yourself

Thanks for taking the time to answer my questions and hopefully I will find everyting I need in ACP.

I think you will, and I think you will find that many of the things you think you have to write yourself are already done.

-- Bob

PS: Yeah, the G2V thing isn't done explicitly, but it's a planning thing not a doing thing -- the G2V data can be acquired with a simple ACP observing plan, no script programming needed!!! Well maybe a separate script to make the plan in the first place, but that would not have anyhting to do with ACP, it would just produce a text file.

[Roy Salisbury]

This can be done (along with all sorts of other things) for your specific needs via the UserActions hooks.

Cool.. I had not ran into that yet.. I'll take a look..

To see the browser side of this, enter the authoring mode and open the System Status item. Then Edit it and you'll see all of the stuff there. The browser side logic runs on top of a framework called TiddlyWiki, and I have developed a bunch of plugins for it. They are self-documented and visible in the Authoring environment. You will need to work with the Authoring environment some in order to "get it". Follow the directions you see when first opening the Authoring environment.

OK... I'll look at that. Might be just what I need. If it comes down to it I guess I can create a COM object to insert into that sequence so it allows me to share the data with my web service on the same system. I'll know more once I look at the pages.

I would do this via observing plan. Acquiring this G2V data is no different than acquiring any other type of data. Make up a plan using a star that will be usable for a few weeks to a month and just run it. You could chain to it from another plan.

OK.. I'll try it that way. I usually use the same G2V star, but I have been reading that its better to choose one closer to your intended target. But again, its just a matter of pluggin in the right star for the session. The only thing that might be tricky is getting the exposure times set properly. Usually this will be based on the ADU (sort of like flats), but perhaps I can find a common ground. If anything I could perhaps write a script based off the acquire flats logic and just run it manually at the start of the session.

I'll let you know what I come up with and if its anything cool I'll be sure to pass it along.
Thanks again for the help.
Roy

[Bob Denny]

I'd still like to know what you're trying to do so I can have a chance to offer "in-architecture" solutions.

[Roy Salisbury]

I'd still like to know what you're trying to do so I can have a chance to offer "in-architecture" solutions.

All I really want to do is the following:

1. Select a sutable G2V star above 60 deg and close to the evenings target. Slew the scope and center/sync.
2. Starting with the green filter, take a series of images of the star to adjust the exposure based on the ADU.
3. Onece the exposure time is determined, take a series of exposures of that star through the red, green and blue filters.
4. Determine the ADU values of the star for each filter and calculate the color combine rations for later use when creating the final image.

Initially I had thought that this could be a part of the overall session (just like flats), but perhaps it might be useful as a utitlity script like the filter offsets or limiting magnitude ones.

Roy

[Bob Denny]

OK, now I have some stuff to sink my teeth into :-)

Flats are actually a separate script which reads a separate sort of "flat plan". It looks seamless due to the #xxxflats directive in regular observing plans, but the main observing logic sees these and chains to the separate script (AutoFlat.vbs). There is a system for chaining from one script to another using a "Chain Token" object which has several properties that can be used by the chained-to script and then used on chain-back to return to the original/chaining script with continuation info.

You want a custom "closed loop" operation to measure the G2V and calculation the combine ratios. From my perspective, I think the best way to start is with a special script (and I strongly suggest JScript not VBScript). It would function analogous to AutoFlat.vbs (not converted to JScript... it's old), except it would not take any input. It would look up a G2V star by RA (given LST) then go there and center it (see below), take the pic, measure the ADU (using PinPoint, it's EASY), adjust the exposure, iterate till the ADU is in your range, then with that exposure time, take images in the other colors, measure their ADUs and then calculate the combine ratios. Write these to a text file (use the built-in Scripting.FileSystemObject).

There are many support functions in AcquireSupport that will make this simple. Do not read the AutoFlat code!!!!!! StartSlewJ2000(), WaitFOrSlews(), TakePicture(), UpdatePointing(), these are the ones you need to use. Also note the capabilites of Scripting.FIleSystemObject in the Windows Scripting Help (included in the ACP Scriptikng Guide).

This is a very advanced usage of ACP. I would recommend that you first get your observatory into shape to run automated with ACP and it's off the shelf capabilities. Only when you are highly confident in its stability and reliability (the whole system) should you start attacking the advanced customizations. Assuming you already know the ACP API as well as those of the common Windows Script components, this is a day or two's work to write it, then probably a few nights learning what you did wrong and fixing thost things. Add the learning time for the various APIs.

[Roy Salisbury]

There are many support functions in AcquireSupport that will make this simple. Do not read the AutoFlat code!!!!!! StartSlewJ2000(), WaitFOrSlews(), TakePicture(), UpdatePointing(), these are the ones you need to use. Also note the capabilites of Scripting.FIleSystemObject in the Windows Scripting Help (included in the ACP Scriptikng Guide).

I'm thinking that this script could be kicked off form the "Target Start" custom action. That would allow me to determine the target and find the correct G2V to slew too (have to figure out how to find one still). I could also put logic in it to only do it once a night if there were multiple plans to run.

This is a very advanced usage of ACP. I would recommend that you first get your observatory into shape to run automated with ACP and it's off the shelf capabilities.

Totally agree. I've still got a long way to go (but a short time to get there). I've got the observatory working with all the hardware and its ran all night without issues. Unfortuantly I only have maybe once or twice a month that I can get out to my observatory, so a lot of work will need to be done with simulators and fake FITS files.

I thnk the first thing I need to do is try to get the ability to abort taking images as the end of astronomical twilight/dawn. Looking at the custom logic ability I am not sure that will work becuase, from the way I understand it, you can only abort the whole plan and not just a portion of it. Perhaps there is a way to adjust (not abort) so it thinks the light frames are done and not actaully an error. Still reading up on everything.

Roy

[Dick Berg]

Roy,
This is a great project and potentially useful for any auxiliary catalog of interest, like the Henden standards or Landolt standards, particularly. Tweaking such a script for various catalogs would be necessary based on the information available in the catalog.

Here's a reference to a (the?) G2V catalog compilation out of the Hipparcos catalog:

http://web.njit.edu/~gary/322/G2V_stars.html

and attached is a spreadsheet I put together once upon a time (in 2005) as input to build a plug-in for Starry Night that would display the G2V stars. (Note that SNP required the stars' RAs to be in decimal degrees.)

Here also is a pointer to the most recent update of the Hipparcos catalog from which the G2V stars can be selected. Click on the FTP tab to get the embedded pointer links to the .dat file. That is the catalog.

http://cdsarc.u-strasbg.fr/viz-bin/Cat?cat=I%2F311%2F

I don't remember now, six years later, whether I built the spreadsheet from the njit table or whether I filtered the primary source file stored at:

http://cdsarc.u-strasbg.fr/viz-bin/Cat?cat=I%2F259%2F

- probably the latter. The 2007 version of the Hipparcos catalog wasn't available in 2005! I hope you'll continue to think about this and pursue a solution.

PostScript: Here's yet another website with spectral standards that I found interesting. It turns out that the "G2V Catalog" is available lots of places.
http://www.gemini.edu/?q=node/10165