Picture reframing and automated best focus routines!

[Matthew]

Just a thread to ask what feature folks would most like to see in the future for Guylain to ponder, what feature you would most like to have.

 

For me its centre a picture where I'd like (without using AstroTortilla or other 3rd party products).  I would love to have BYE (connected to the ASCOM scope control) snap a shot - then for me to be able to select any star and click another point of the screen and be able to say - move this picture so this star appears at a user selected point on the screen.  So say if I clicked on M20 and its wasn't framed well, select a star and click some place on the screen and command BYE to shift the scope until the star appears close enough to the point I selected (say within a maximum of 5 moves - take a new shot and see if the star is in the correct position.  If BYE knew the RA DEC directions on the screen and image scale - it could hopefully calculate to move a star from (X, Y) to a new coordinate (A,  B ) it would have to slew a certain distance in RA and DEC - issue that movement and use fuzzy logic with a set number of attempts to get the scope to the desired point.

 

I have tried this with Astro Tortilla - but it simply can't issue a sync to the ASCOM Vixen driver when the Sky6 PE is controlling the mount... (a Vendor issue they barely accept let alone would consider fixing)!

 

Desire two would be focus aid assistance - click on any bright star and have an automated imaging routine where BYE tries to find best focus for said star (rather the stop imaging - add Bhatinov masks - and manually adjust focus for 2 - 10 minutes) - then go back to imaging - and having to do this several times a night.  I would like a dummies option to simple say - Check for best focus - and know after a minute or two it had found it again!

 

I would love to be able to insert into my say 2-6 hour imaging sequence a command that said say every 30 minutes - 2 hours - automatically check for best focus!

 

What would other folk most desire?

 

Many thanks,

 

   Matthew

[Matthew]

JUst an update for those interested.  The picture reframing now is completely achieved since upgrading my software suite to The SkyX (TSX) + Dome + Camera add on + Tpoint.  TSX has a normal slew function and a closed loop slew (CLS).  CLS goes to the requested coordinates - takes and image and plate solves it and if the plate solve doesn't agree to the target coordinates it works a correction and moves the scope accordingly - which works beautifully!

 

Focusing I have really tried yet with TSX - which also support FocusMax as another add on - I still prefer BYE taking 4 seconds shots and my using a Bhatinov mask.  But a great BYE focusing routine would be really appreciated!

 

One thing TSX can do which is quite nice is create a 2*2 binned shot in software - storing it as a FITS file.  Another thing I quite like is ability to auto subtract darks from the images it captures, be nice if BYE could do that - but I will save that for a new thread...

[astroman133]

Matthew,

 

I know about the features of TSX and it's plate solving capability to give spot on pointing. I use Astro Tortilla to do the same thing. It is integrated with BYE to capture the image and it is freeware.

 

As far as FocusMax being an addin of TSX, that is incorrect. In fact, neither product is an addin of the other. They simply communicate with each other in a client-server arrangement. When FocusMax needs a new image to analyze for star size, it asks TSX to provide that image. If FocusMax were enhanced to use the BYE external API, then TSX would not be necessary for this functionality.

 

Why would you want to create a binned image of your raw data? The only reason that I could think of is if you were planning to combine that data with unbinned monochrome luminance, but since you already have the full unbinned RGB image, why go to the trouble of binning the data after the fact when it will need to be unbinned to be combined with the luminance?

 

Some people have asked about creating a FITS image in BYE, but this is basically redundant capability, since the conversion of CR2 to FITS is commonly available in image stacking software. There seems little benefit to adding that capability into BYE.

[s3igell]

Matthew,

 

Please realize that TSX is doing several things whilst providing you with Binned FITS:

1. Converting the RAW CR2 File to a TIFF / FITS Binary File
2. Debayering the Binary File
3. Binning the result 2x2
4. Writing the result to a FITS File (hopefully transferring all of the CR2 File's EXIF Data in the process)

Given that Application of Darks / Flats / Bias should really be done prior to Debayering (so that a Hot/Stuck/Noisy Red Pixel is applied only to the corresponding Red Pixel of the Light Exposure), is this feature really aiding your AP Imaging Workflow??

 

The reason that many Astro-Imagers with Mono CCD  Cameras resort to Binning is that they can thus reduce the time spent on RGB or Narrowband Exposures (relying on the Binning to produce the equivalent of Larger Pixels).  On any One-Shot Color Camera (CCD or DSLR), that benefit of Binning cannot be realized because the Bayer Color Mask means that adjacent Pixels cannot readily be combined and provide useful Photon Intensity data.

 

Perhaps you are looking for Software Binning to produce some other Benefit??

[Matthew]

To answer the above thoughtful replies - I was only using binning at 2*2 in regards to plate solving - as I was over-sampling drastically on my current F10 C9.25 -> a Canon 40D - it gives me 0.44 arc seconds a pixel - which is way out of sync with my quality of night skies.  So for Closed Loop Slews in TSX I use 2*2 binning to also hopefully make the plate solve faster by taking less time to create an image where sufficient stars stand out.

 

I do like Astro tortilla - but I could never get it to work with The Sky6 Professional Edition talking to teh ASCOm Vixen SkySensor200-PC driver - something in the chain (at the Software Bisque end) retricted syncs being sent from AT to the telescope driver - so AT could never get the CLS working when The Sky6 was controlling the scope.  Remove The Sky6 and it was a different story!

 

Just saying its a feature that I love!  Overall I am not sure how much I like TSX's general camera control function... still learning it but in general I prefer BYE!  The same way that at the moment I prefer PHD and PHD2 to MaximDL or TSX guiding my mount at present!

[astroman133]

Matthew,

 

Thanks for re-posting to this forum area. This is where Guylain looks for user recommendations for new features in BYE/BYN.

 

I am not a SB user, but you are not the only person I have heard the sad story from regarding how difficult it is to get SB to recognize what you are saying and to take ownership of a problem. In this regard I have a couple of suggestions. First, if you don't want to abandon The Sky, you could upgrade to The Sky X which supports ASCOM for scope control. That way you can use TSX to control your Vixen mount with ASCOM AND use AstroTortilla to point the mount, and sync, through ASCOM.  Alternatively I would suggest Sky Tools 3 as an observing session planning tool and planetarium program. It uses ASCOM to communicate with your mount would also allow AT to work.

 

A few years ago Guylain made a comment about adding automatic focusing into BYE. As he saw the investment that he would have to make to get it to work he retracted that comment and I do not believe that it is now in his development plan for the product. However, I would suggest that at least he could take a step in that direction and add focus tracking to BYE. My idea here is that after the first image in a capture run, the app would pause the capture plan and allow the user to select a star. The size of that star is measured in each subsequent image and a graph or table is presented. That way you can see easily see how the size of the star has grown and can suspend the capture plan to adjust focus when the size has grown too much.  It is not automatic focusing but should be much simpler to implement.

[s3igell]

I'm not sure that I understand how BYE is going to be able to Alleviate the AT-to-ASCOM-to-Mount vs Sky-to-ASCOM-to-Mount Conflict.  Or are you suggesting that BYE should also become a Planetarium as well as Plate-Solving App so as to totally replace the need for SKY6.

 

I am also afraid that this might be an Underestimation of the Complexity necessary to manage the "Simple" Task of directing the ASCOM Mount Interface in an exact combination of directions necessary to shift a Star to the desired Position in the FOV.  Consider all that PHD2's Calibration Routine needs to do for the similar Task.  ASCOM does not provide any details about the Correlation of Celestial Directions to the Mount Motions, nor any measure of the Absolute Motion in ArcSeconds that any one Impulse in any one Direction (nor are they the same in RA vs DEC), nor compensation for Backlash.  This is why Plate-Solving is pursued by AstroTortilla and SGP and most all the other Apps which produce similar Positioning Controls.

[s3igell]

Speed during Frame&Focus and Snapshots is handled by defaulting to JPG Downloads and Shorter Exposures.  The Former are created In-Camera regardless of your Output Selection - so No Cost.  The Latter keeps the Star Count limited to the Higher-Magnitude Stars.  AstroTortilla defaults to JPG Images, in any case, because of the "Universal Format" and avoiding any need for Camera-specific Decode Routines.

 

On the other hand, both the 2x2 Software Binning and the RAW-to-FITS Conversions are relatively Time Consuming operations.

 

We recognize that you are having Mount Vendor and Software Developer Support Issues (none of then O'Telescope).

It's too bad that you are receiving so little support from them...

[astroman133]

Matthew,

 

It seems like you are using older software and equipment. Perhaps it is time to consider upgrading. Specifically I am suggesting upgrading The Sky 6 to The Sky X.

 

The Sky X supports ASCOM for talking to devices. This would allow you to use Astro Tortilla to talk to the Vixen SS2K driver without going through TSX AND allow TSX to talk to the SS2K driver, as well. I have not tested it, but depending on the driver's capabilities, it should allow AT to sync the mount without The Sky getting in the way. The only issue may be that the SS2K driver is not written using the ASCOM Local Server template so it only supports one connection at a time. Fortunately, you can use POTH (part of the ASCOM Platform distribution) to overcome this restriction. Simply connect POTH to the mount via the SS2K driver then connect TSX and AT to POTH (POTH Hub will be in the Telescope Chooser dropdown list).

 

Also, with AT you can downsample the image before extracting the stars. This would eliminate your need to bin the image in your capture software.

[Matthew]

I'm not sure that I understand how BYE is going to be able to Alleviate the AT-to-ASCOM-to-Mount vs Sky-to-ASCOM-to-Mount Conflict.  Or are you suggesting that BYE should also become a Planetarium as well as Plate-Solving App so as to totally replace the need for SKY6.

 

Well for a start - BYE could possible:

 

1. understand the image scale and RA +/- and Dec +/- directions

2. At a button click project a X on the screen showing which way +/- RA is and which way +/- DEC is

3. Allow you to click on an object and point to where you wish it to move

4. Calculate with the simplest of Mathematics How much RA or DEC this move would involve and where it was plus or minus in each

5. Offer you a job or nudge mount amount in RA and DEC

 

So the first point - knowing imaging scale and RA and DEC.  One could have a routine to say calculate RA and DEC (say the way PEMPRO does):

1. Select the brightest star on the screen - move it a set distance in +RA - say 10 arc minutes and re click on the star - that gives BYE both the image scale and RA + and - compass directions

2. Do the same and move it in + DEC 10 arc minutes and once mark the star and end of a star - this confirms the image scale and +/- DEC directions.

 

By having the user click on a target star and then click on where this star moves to after a set size move means you don't need this intelligence built into BYE.  That makes the rest of the solution very, very simple - no plate solving required.  BYE doesn't need to know where the star field is or should be - it just needs to know how much you want it to move in RA and DEC and issue this command until you are happy with the framing.

 

Given the mount is connected to a guide program - that support relative or absolute moves - which has know position provided - calculating a nudge scope in any direction should be simplicity itself.  Would be helpful when the framing is wrong by (for me) - typically < 20 arc minutes but the target is very dim and you might need a 20 - 30 second shot to resolve stars and you are trying to guess whether its slightly + or - in RA and DEC but the target is too dim to see in live view.

[astroman133]

I think that the OP is postulating that BYE could determine the scale and calibration on the fly by making a move and calculating whether the target point moved closer or further from the center, and then adjust accordingly and make another move, do the same thing until the target point is close to the center. It is essentially doing a PHD type calibration on-the fly.

 

I believe that a good solution for him is to replace The Sky 6 with some other planetarium program that supports ASCOM telescopes and then use Astro Tortilla to adjust his pointing.

[Matthew]

Pointing is generally accurate to 1-2 arc minutes - sometimes better than 30 arc seconds and tracking is superb.  So rather than toss out about $500 of The Sky and Tpoint - a simple reframe was queried - but your right The Bisque boys and ASCOM aren't the best of friends!

 

So the feature suggested could I agree be made superfluous by having proper plate solving that my planetarium program uses.  But without that step (and all the costs and risks that involves) I was simply asking if it would be possible for BYE to:

 

1. Understand RA and DEC angles on the screen

2. Understand Image scale

3. Offer an assisted re-centre image functionality simply using the information in points 1 and 2 above and a desired travel vector a user wishes to specify...

 

Many thanks,

 

   Matthew

 

PS

 

Given I only image using two scopes into two Canon 40D cameras - on a permanent set up... if no software assist could be given I just realised I could make two simple transparencies with marked RA and Dec axes - set the image scale for each OTA and also capture the RA and DEC angles and then whenever I have image re-framing needs I could hold the transparency over the current image - and read off how much + or - RA and DEC I need and then simply use The Sky 6 Professional Edition's nudge scope 1-60 arc seconds / arc minutes / degrees in +/- RA and DEC.

 

And If I where really clever I may be able to make an software jpg that I could hover over the correct spot on the BYE screen rather than use clear plastic transparencies.  It's that simple functionality I am considering given I do not wish to ditch expensive software for the equivalent rather expensive latest versions that may work - or may introduce new glitches!

 

Now from memory BYE's polar alignment tab already gives me the axes and their directions - so I guess all I am requesting being considered is please add

 

1. image scale in arc minutes and

2. a an x-y grid on this image scale so I could easily overlay this with any image and red off how much RA and DEC change is required!

 

https://www.flickr.com/photos/53052745@N05/8492279106 - is an image of the BYE guiding / polar alignment feature - those two simple add ons above (grid scaled to show 1 arc minute) and +/- RA and DEC labelled on the axes would give me 90% of what I'd like!

[s3igell]

Pointing is generally accurate to 1-2 arc minutes - sometimes better than 30 arc seconds and tracking is superb.  So rather than toss out about $500 of The Sky and Tpoint - a simple reframe was queried - but your right The Bisque boys and ASCOM aren't the best of friends!

 

So the feature suggested could I agree be made superfluous by having proper plate solving that my planetarium program uses.  But without that step (and all the costs and risks that involves) I was simply asking if it would be possible for BYE to:

 

1. Understand RA and DEC angles on the screen

2. Understand Image scale

3. Offer an assisted re-centre image functionality simply using the information in points 1 and 2 above and a desired travel vector a user wishes to specify...

 

 

I still think that you are significantly over-optimistic about both your estimation of the tasks involved AND of the difficulty level of some of those tasks.

 

You have existing options:  Alternative Planetarium Apps, Plate-Solving Apps

Why not leverage those, instead of lobby for Changes in yet-another-app??

 

For myself, I believe that Guylain has a significant amount already on his Development plate.  Planned-Features and Feature-Requests which can keep him fully occupied, including a few that I've posted.

[Matthew]

Because Guylain asked for suggestions - over a year ago - raise it then and he was interested in considering it, thought I'd bring it up again to keep it in view.

 

Rather than criticise my intent, why not suggest if a slight change to the drift alignment would do what's needed - it can already capture RA and DEC directions - simply labelling these axes and showing how far and arc minute is and putting up a grid pattern would give me 80% of what is required - yet you state this is way too hard rather then perhaps seriously consider if this simplistic approach would be valuable and able to be done as a series of simple add ons.

 

I know you are trying to help - but if you complained your local road to school was too bumpy and I suggested maybe you should buy a new car - you might get how I feel

[s3igell]

I don't believe that I am criticizing your intent - rather pointing out that you have existing options (other than seeking significant updates to BYE).

 

Regarding the Drift Align Tool screen - again I believe that you significantly underestimate the amount of work necessary to deliver what you suggest.

The Drift Align Tool is a significant addition to BYE/BYN, but all of the Reticule Positioning actions are Manual User Actions (there is no Automation; no Mount Control or Communication; no concept of RA/DEC directions or motion or FOV).

 

All of my comments about the Coding Difficulty of Determining FOV and Mount Responsiveness and Image Location and other requested tasks still stand.

 

But, yes, of course you are free to propose your Feature Request.

[Matthew]

I think our thinking is aligned - you see the mount control I always envisaged running thru The Sky6 PE!

 

The Positioning for drift alignment I always presumed would be manual user actions.

 

BTW The way I often determine image scale is simply image the full Moon and say the diameter is 30 arc minutes.