Clarification on Guide Speed and Aggr.

[Alvise Dorigo]

Hello, I am new in this forum, but not in PHD2.

I still am not sure to understand the relation (if any) between Guide Speed and Aggressiveness.

My understanding about guide speed is "at what speed wrt the tracking one, the pulse will tell to the mount to move in RA or DEC". For me it means that if PHD2 sends a correction pulse of 100ms, for the duration of 100ms the RA will be corrected by moving the the guide speed. So if the guide speed is 1.0x, the entire pulse is applied. If the guide speed is 0.25x, a quarter of the pulse is applied (the same as if it was a pulse of 25ms).
Aggressiveness means, e.g. 50%, that if PHD2 finds an error E, it will apply a correction 0.5 x E.
Is my understanding correct ?
If yes, I "feel" a kind of redundancy, because it seems to me that I can control the same thing from both guide speed and Aggr. (well with Aggr. I can fine control, from 0% to 100%, instead of only 4 discrete values 0.25x, 0.5x, 0.75x, 1.0x).
I would like some "rule" that, based on some well known setup parameters, tells me how should be the "sweet" range of Aggr. for a certain guide speed.
Why people chose to guide at 0.5x (the most popular one), while other people want to guide at 0.75x ?

Thank you

    Alvise

[bw_msgboard]

Hi Alvise.  I think you’ve got some misconceptions on this.  As you’ve said, RA guiding is normally done in the mount firmware by either slowing or speeding up the RA motor or by turning the motor on and off for Dec.  If the guide speed has been set a 1x sidereal, the RA motor will run at 2x sidereal to move the scope west, and it will briefly stop running to move the mount east.  If the mount guide speed is 0.5x sidereal, the RA motor will run at 1.5x sidereal to move west, 0.5x sidereal to move east.  The Dec motor will always run at the specified multiple of the sidereal rate.  The PHD2 aggressiveness parameter has nothing to do with any of this and is completely independent.

 

The aggressiveness factor just says how much of the computed correction should actually be applied. So if the aggressiveness is 0.80 (80%), a computed correction of 100ms would be reduced to 80ms – the RA motor will still speed up or slow down as before but for less time.   This is done to keep the system from going into oscillation and as a hedge against seeing effects and small errors that may have occurred during calibration.  It is essentially the same as ‘gain’ in a closed-loop control system.  This is described more completely in the PHD2 manual – look in the contents for ‘Guiding Algorithms’.  In any case, there is rarely any need for you to change the aggressiveness and it’s nearly impossible to do correctly by squinting at the real-time display.  Changing it incorrectly can really screw up your guiding.

 

The choice of mount guide speed depends on the mount mechanics.  A higher guide speed makes the mount more responsive, helps it push through small resistance, and reduces the guiding impact of Declination backlash.  With some mounts, higher speeds may be a problem because of weaknesses in the mount or driver software – for example, higher guide speeds result in shorter correction pulses which in turn puts more pressure on the timer logic in the software.  I think many of the default guide speed settings are just historical, perhaps even going back to the time when people guided manually.  In other cases, the choice may be made to avoid known and undocumented shortcomings in the software.  We don’t recommend anything lower than 0.5x sidereal and a higher rate is usually better unless there are good reasons for not doing that.

 

Hope this clears things up,

Bruce

 

---

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/4efcd1e8-0ede-4873-ba2f-7fa90d38f850n%40googlegroups.com.

[Alvise Dorigo]

Hei Bruce, thank you very much for the long and detailed explanation.

I understand that it could be better to set the guiding speed to 0.75x (the middle between the minimun reccomended - 0.5x - and the maximum allowed by PHD2 - 1.0x).

While I am not sure to understand a sentence: "there is rarely any need for you to change the aggressiveness and it’s nearly impossible to do correctly by squinting at the real-time display".

I've read in many place in the web that several people need to change the Aggr. if the guide graph is quite "jumpy". Are they wrong ? Should I keep the default RA/DEC Aggr. values set at the first setup instead ?

Thanks a lot.

    Alvise

[bw_msgboard]

Hi Alvise, see below

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Alvise Dorigo
Sent: Sunday, February 21, 2021 10:15 AM
To: Open PHD Guiding
Subject: Re: [open-phd-guiding] Clarification on Guide Speed and Aggr.

 

Hei Bruce, thank you very much for the long and detailed explanation.

I understand that it could be better to set the guiding speed to 0.75x (the middle between the minimun reccomended - 0.5x - and the maximum allowed by PHD2 - 1.0x).

While I am not sure to understand a sentence: "there is rarely any need for you to change the aggressiveness and it’s nearly impossible to do correctly by squinting at the real-time display".

I've read in many place in the web that several people need to change the Aggr. if the guide graph is quite "jumpy". Are they wrong ? Should I keep the default RA/DEC Aggr. values set at the first setup instead ?

 

You should leave the settings at the default values unless you’ve done a detailed analysis using the LogViewer and completely understand what you’re seeing.  In most cases, people are confusing cause and effect when they see “jumpy” guiding graphs.  Those problems are coming from the mount or the atmosphere not from the guide commands.  In general, a lot of what you see on the web with regard to PHD2 and guiding is outdated, context-dependent, or just plain wrong.  Many of the forums are essentially un-moderated, so bad information goes unchallenged and the useful pieces of advice are buried in the noise.  On this forum, we work hard to weed out misconceptions, and you are quite often getting answers from the developers or from people who have a deep understanding of the subject matter.  You might want to read this document if you’re interested in the details:

 

https://openphdguiding.org/tutorial-analyzing-phd2-guiding-results/

 

The appendix section on “Fooled by Randomness” deals with a lot of these parameter adjustments people like to talk about.

 

Bruce

 

 

 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6bdf2a3b-960a-4796-953f-0eb50c19ff43n%40googlegroups.com.

[Brian Valente]

Hi Alvise

Just to add to what Bruce said, if you want meaningful feedback that is educated and based in experienced, this forum is really the best place to start. Bruce & Co. have analyzed thousands of guidelogs (tens of thousands?) and have seen it all: What actually works and what are dead ends. 

I find it takes quite a while (many guiding runs) and a lot of log analysis to start understaanding what is really going on. It's easy to fall into the false positive trap, where adjustment is made and it appears things are better, "so that must be what works".  

Most of the time when I read guiding advice in other forums, someone posts a screen capture and there is a flurry of recommendations that all seem legitimate. I personally would never be able to tell anything from a screen capture, so if I find advice being doled out without a lot of clarifying questions and guidelog review, I woudln't put much weight on that kind of answer.

There are times when adjusting the aggressiveness makes sense, but tweaking it here and there is not going to do anything meaningful. of the times i see adjusting aggressiveness, many times it's to undo a bad setting!

Guiding can definitely help things, but it won't compensate completely for a poor performing mount, and it won't compensate completely for poor seeing conditions, no matter how much you fiddle with it. 

Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/78D36C6192FB4D20AFF21124C531FBF6%40HomeDesktop.

--

Brian 

Brian Valente

portfolio brianvalentephotography.com 

[Alvise Dorigo]

Thanks a lot Bruce and Brian, I really appreciate your intervention (I must admit that I only read the user manual to get the basic functionality for lack of time... but I will cover it better in the next days of full moon). I will reset back to default settings and try to better understand the algorithms behind phd. Of course I know the added value to ask people that devoted time in studying, debugging and developing... (I am a system engineer and spend a lot of time in debugging, digging in syslogs, user logs, etc.). I haven't trusted 100%, of course, what I found on the web. Something sounds reasonable and sometimes inspiring (on different things than phd2), something else is simply rubbish (the majority). This is why (considering the little time I have to dedicate to astrophotography) I eventually decided to bother the official forum of PHD, even if only for a silly question.

   Alvise