Jared
With a timed (or otherwise) sequence, SGP follows the following order of events:
Auto Connect equipment (if not connected already)
Waits for sequence start (optional)
AutoSlew
AutoCenter
Auto Resume Autoguider (if required)
Auto Calibrate Guider (if required)
Autofocus with Lum (assuming the start of sequence AF is enabled and Lum is AF filter position)
Guider resumes (if paused )
Selects filter for event (and adjusts focus offset, if enabled)
I think this all-automatic sequence has a drawback. I have an off-axis guider and share a focus module between 3 Feathertouch systems. My power-on module position is at a fixed reference so that I can switch it between scopes and not lose my calibration. I’m losing some of the power of SGP since I have to manually connect and put the focuser in the right position for the autoguider and centering to work at the beginning of the sequence.
The following (for me) would overcome these issues.
Auto Connect equipment (if not connected already)
Wait for sequence start (optional)
AutoSlew
Automatically select the filter used for autofocus
Move focuser to the absolute focus position for that filter (rather than do this manually, if this is a selection in the focuser settings).
AutoCenter (now works as the image is in approx focus)
Resume Autoguider
Calibrate Guider (now works as the image is in approx focus)
Autofocus with Lum (assuming the start of sequence AF is enabled and Lum is AF filter position)
Guider resumes (if paused)
Moves filter for event setting (and adjusts focus offset, if this option is enabled)
I know we all have different setups and needs but off the top of my head, I couldn’t think of a downside of this suggestion. Hope it is useful.
I agree that AF probably needs to be moved. I had meant to say something in the other thread as you mentioned it in your video.
We’ll see about addressing this soon. The one “gotcha” is that not everyone uses slew prior to center on targets. Which would mean you could be doing the initial focus run at your park position. Where as with the current method you would at least be on your target. But this could be easily resolved by using the “Slew” option as well.
I agree - I was not proposing to focus at the park position - merely to move the focuser to the stored focus position for the filter identified as the one used for AF in the autofocus settings. The AF would occur after the guider calibration, since it is quite tolerant of small focus issues. To clarify my words better:
Auto Sequence:
Auto Connect equipment (if not connected already)
Wait for sequence start (optional)
AutoSlew (optional)
Automatically select the filter that is used for autofocus (normally LUM or clear)
Move focuser to the absolute focus position listed in the filter table for that filter (rather than do this manually, if this is a selection in the focuser settings).
AutoCenter (Optional, now works as the image is in approx focus)
Resume Autoguider
Calibrate Guider (now works as the image is in approx focus)
Autofocus (assuming the start of sequence AF option is enabled)
Guider resumes (if paused)
Selects filter for first event (and adjusts focuser position using difference between stored filter focus positions, if this option is enabled)
I think,… this would allow someone to literally run PHD2, connect all, run SGP, load sequence and run sequence without any further manual intervention. All the information needed for this is already in the autofocus settings, the filter tables and stored in the sequence. If any of the slew, center, AF or guider options are disabled, I’m assuming the program logic just skips that step…
I have been looking at the implementation of this and it seems like we can simply state (and implement) this:
There are many possibilities that can differentiate the first filter used in a sequence… it might be the filter for AF, it might be for Plate Solving, it might be the first filter in the sequence. SGPro does not really care what it is or how you got there. For the first focuser movement in any sequence, SGPro will move the focuser position as defined for that filter (in an absolute manner). From this point forward, all focus points (per filter) will be handled in a relative manner.
Does this sound resonable? @buzz It seems like it would handle your issue. Auto center happens first for you. Whatever filter you use for plate solve will be set and the focuser racked to that absolute position. Then, assuming your focus is solid enough that the solver can derive a solution, the rest of the sequence will start falling into place.
This should still be watched. If your temperature is distinctly different from the temperature at which the focus pos for your filters was set, you could still be out of focus enough for AF or plate solving to succeed.
Maybe I’m missing the benefit of the focus position moving to the absolute location? I might be misunderstanding how this works, but that could be disasterous for me if I understand correctly. The reason why is I move the focuser to control different scopes (each has its own motor but only 1 controller ). The position may vary greatly however it is the offset that is important for filters so this was of no concern. I rely on the fact the filter positions are offsets. Also, I will sometimes manually rack the focuser on setup or shutdown, which may throw the focus position off – I do this if I don’t remember to rack it all the way in for transport to fit in the case and everything else is disconnected.
IMHO, the focuser positions should always be relative… But like I said, perhaps I am misunderstanding…
Maybe you are misunderstanding?.. I can’t really tell. All focuser movements due to filter changes will remain relative except for just one… the first one. I realize from your comments that this is not something that suits you perfectly, but it is possible that your profiles carry actual meaningful numbers. We are trying to move toward automation… this is a step we would like to move toward (if it turns out terribly, we’ll can it or tweak it, but this is where we test things). I have the same issues as you… one controller, multiple setups… Each profile carries it own set of numbers.
Noted, but lots of other folks have humble opinions that asked for this… Again, we are only talking about one movement here.
Not sure if you read the release notes… maybe this will clear it up:
To try and distill this feature, it will work like this (assuming you are using “Adjust focus position per filter”):
Prior to running a sequence, if you fiddle around and plate solve or auto focus or even just change the CFW position manually, you will find that, during your manual operations, the first change of filter wheel will set the focuser to the absolute position for the first CFW change requested (as defined in the filter table). This is OK, you are at your rig and can make sure your operations complete successfully. When you start the sequence, the first filter wheel change will cause a relative change in focuser position (like you are used to now).
Optionally, you can also, just open the sequence and run it. This assumes that your initial focuser position for the first filter (might be Plate Solve, might be AF, might be the event filter, etc…) is precise enough to satisfy the first sequence operation (AF, Plate Solving, etc). This will set the focuser position to the absolute position specified for that filter in the filter table. From then on, only relative focuser movements will be made.
My problem with the one movement is that it could very well try to rack the focuser past the end of limits which can potentially be damaging.
My profiles do carry meaningful numbers and may work sometimes, but as time goes on they will drift due to manual set up, or more likely due to scope changes. Because of this drift it isn’t possible to have different profiles in this case – I have profiles for each scope with their relative offsets.
I certainly understand where this could be useful. Like smart focus with SCTs, I’d really like to see a way to not have this focuser movement occur. The only workaround I can see is either adjusting the filter offsets before each start, so that the Lum filter (for example) is at the current location (ugh), or manually resetting the focuser position on the focuser to the current offset.
Taking a step back, I think the larger problem is the combination of filter position and offsets, as they are two different things. Not that I have an ideal solution, but perhaps reworking the filter setup to have a single absolute position (“on start go to xxxx”, “initial position: xxxxx”, not sure), and simply an offset amount (-10, +20, etc.) for each filter. Offsets would be self explanatory, and perhaps a checkbox for absolute where disabled means “use current position”.
I did see this in the release notes but hunted for the thread to find what the genesis of this was – because the title isn’t that obvious, perhaps another thread dealing with just this issue would be best to see what others are doing? Perhaps it’s my workflow that needs to change.
Interesting feature but with my current equipment (GSO newtonian stock focuser motorized with rigel systems stepper and controller) it will do more harm than good.
My system has a horrible backlash and after a couple of nights the absolute focuser position is meaningless, with every focus run the true focused position slowly drift from the absolute’s. Usually the last position of the previous imaging run is closer to focus than the absolute position.
I like Bhwolf idea of adding an extra option to the filter definition something like: on start go --> drop down with 3 options (absolute, current, custom).
There were two separate suggestions - one, which I think is less contentious, is re-ordering the autoguider and AF in the start-up sequence. This makes sense for those of us with OAG’s and as far as I can imagine, not an issue for those with separate focusing systems for guider and imager.
The second was the initial focuser position. I have several OTA’s but one focuser module, which I return to a fixed value at the end of a session. It would have been a way to get the focuser into the ballpark - but realise, upon reflection that not all OTA’s are as well-behaved and consistent as simple refractors with a good focuser on the end. A simple option, along side the auto-adjust for each filter, to set the initial focuser position may be all that is required.
I delayed my response to this to give everyone a bit of time to get their thoughts out. I wish there weren’t 4 different discussions on this right now. I am just going to reply here. Here are a few things to note:
OAGs make automation hard
We use betas for more than testing functionality. We test response to ideas here too. I appreciate all the feedback on this issue.
We try to improve the usefulness of SGPro, not make more work for you. The problem here is that what is good for one group of people is a chore for others. This inevitably equals yet one more option (in a growing sea of options… we try very hard to avoid this).
So… let’s talk about this feature specifically (here are some decisions I have made):
Most people don’t like the absolute positioning of the focuser on first filter use. Either they do not have gear that can achieve repeatable focus results or they don’t care about absolute positions… only the difference between them. Either way, both are valid reasons to not want this.
Some people have high end gear and have moderately consistent temperatures. This means they can ask the focuser to make a single absolute movement at the beginning of the sequence (or first filter use)
Because this seems like a Most vs Some, it is the Some that, in this case, will need to find the option to turn this behavior on (it is now tucked away in the “Other” sections of focuser props with backlash and stuff like that).
If you do nothing SGPro will continue to operate as normal
If you check this box, you need to be aware that this will happen on your first use of a filter (in or out of a sequence).
@buzz Pushing AF between slew and center (and thus before guider calibration) is a little more complex than I am willing to accept for 2.4. I like the idea of it… just pretty invasive the way we have target changes set up right now.
OK, let me see if I get this. I have been off the forum for some time due to a shop remodel and a knee surgery.
This filter/focus thing mostly started as a result of wanting to be able to go “all automatic”.
Checking the option implemented in 2.4.3.7 would work fine assuming:
a) One is set up to focus with one filter (typically luminance) regardless of the imaging filter.
b) The focus position in the setup for that (typically luminance) filter is close to actual focus at the time of sequence startup.
OTOH, if the focus position for that (typically luminance) filter was way off and NOT close to the actual focus at the time of sequence start, then it could result in failed initial focus and the other problems that would flow from that.
Do I have that right?
My system is certainly accurate enough to do that, but it might require some adjustment of the default position for that focus filter if the starting temperature changes a lot from session to session.
Actually one correction here… the first focuser movement will be absolute. You are assuming here that the first action requiring a filter is AF. This may be true, may not be true. The first action requiring a filter (and thus absolute positioning of the focuser), could be auto centering, solve and sync, the event filter or whatever. It’s the first use of a filter with a valid focus position that matters.
Sorry - I only saw the email on my phone and had not realized you had already implemented the proposal. Boy, you’re quick!
To answer your question, I have one focus module and 5 potential system configurations. Each has a defined extension with a nice round number, say 5000. At the end of an imaging session, I return the module to 5000 (it remembers its position when you turn it off). When I power up the same or different configuration, I use the appropriate equipment profile to get the focuser into position so that the startup will function. The AF routine takes care of the rest. I think this option will be very handy. All I was doing before was doing the same thing manually. Even if my one Crayford focuser slips, I can set it with digital calipers so it is back where it should be. I think this will work well with the refractors. The RCT will probably be more of a challenge, since any collimation adjustment also adjusts focus too. I have established accurate focus offsets using the gold focus masks (posh Bahtinov grabber)
I just need some clear skies…but the good news is that I should have a permanent setup in the near future to make the most of what I have.