Is PEC of value when using a guide cam?

[glowingturnip]

totally unscientific experimentation from me, but I had a (rare) sweet guiding run one night, barely a glitch showing, so turned off dithering for a few frames and let PHD train the PPEC on my AZEQ6GT.  I've been tracking using sidereal+PPEC since.

Is it better ?  Dunno.  Is it worse - nope.

[Demonperformer]

People seem to keep talking about "faffing" about.

At the time I originally posted I was unaware of Autopec (and, had I been, would probably not have posted at all). Now, however, I am forced to ask the question: just how much "faffing" is it to press a couple of buttons and then let the computer get on and do all the rest of the task (including starting to use the result when ready) and then just carrying on doing what I was going to be doing anyway while it is doing it?

I appreciate that many very competent APists have managed to reach the level they have without using PEC, and as they are getting results that satisfy them (and many others) they are not going to go back to add something which would probably add nothing to their undoubted already-great abilities. Those who have reached this level will, I suspect, echo Martin's comment that it "doesn't come without considerable effort".

I have to say that, having now examined the possibilities of Autopec, the question for me has become not so much "is it worth adding it to already brilliant guiding", but "is it something that will enable me to produce better results while I am learning to guide brilliantly". Steve Richards states (MEPC p79) that (within its limitations) "PEC can be very effective". And from the bottom of this steep hill, I now see PEC as a useful rung up the ladder [if you will excuse the mixed metaphor].

Will I be satisfied with the results that PEC alone gives me? I sincerely hope not.

Will starting to get better results spur me on to see what I can achieve when (after "considerable effort") I have learned enough about this guiding lark (in practice, not theory) so that I can achieve 30 min subs? I suspect they will - certainly more than struggling on with no appreciable sign of improvement would.

So will I be running Autopec (when I finally manage to get this PA lark sorted out!)? You bet.

[MartinB]

Lovely post DP. We all have to find what works for us.

[chrisshillito]

Having gone through your (excellent) youtube channel I could only find tutorials on PEC and PECPrep ... I finally found one video called peccap3 that explains AutoPec here

Not sure why it does not show on your channel

Now, on using autopec: Is this a matter of choosing between 2 methods (PEC+PECPrep / AutoPEC) ? Is there an added value of using both methods + guiding with PHD2 ? I do believe that we should keep things simple but I would like to have your thoughts on this very interesting topic

Basically, is AutoPec meant to replace PEC+PECPrep in the future ?

I simply never got round to making a video on AutoPEC - I did document it in the EQMOD VS-PEC document though. The video you found was one I put out for the EQMOD yahoo group just to get those testing the early development releases going - the user interface has changed a little since then.

AutoPEC serves two functions - it records your mounts PE (irrespective of whether you want to use PEC) and it can generate your PEC playback curve from that raw PE data. For the PEC generation it essentially incorporates as reduced set of PECPrep functionality. The previously used method of pulling an unguided PHD log into PECPrep remains valid but has many manual steps in order to synchronise the data with the mount, convert the data from pixel movement to arcsec movement and filter the data to remove non harmonic signals etc. The concept with AutoPEC was to simplify/automate this process as much as possible. With AutoPEC the data it collects is automatically synchronised with the stepper position, movement is measured directly in arcsecs and default filters are applied to create a "steady" PEC curve.

PECPrep remains a key tool and you can pull the raw PE data recorded by AutoPEC into PECPrep for analysis (worth doing even if you aren't interested in running PEC) or if you feel you need more control over the PEC generation process itself. 

Chris.

[Vox45]

Thank you Chris for this detailled explanation

Cheers and kudos to you for this fantastic tool !

[chrisshillito]

Hmmm... I think a lot depends on the likely productivity of the faffing. Even with the number of clear nights I get here I would rather be imaging than faffing. I know you've advocated PEC plus guiding before and I respect that, but I've never tried it because I'm of the, 'If it ain't broke' school. I also try, wherever possible, to subtract rather than add complexities. 

I was brought up in the "if it ain't broke fix it till it is, then take one step back" school. Probably not good if we're all like that but it maybe beneficial to have the odd person around that works that way.

[Vox45]

Depending on the kind of mount you own (HEQ5 or HEQ6) there is another way to improve guiding. It is to mode your mount using Rowan's Belt Modification Kit

This would improve tracking a lot as it would significantly reduce the gear errors (and make the mount quieter!) My guess would be that the belt mode +  guiding alone would be enough for long subs. I have no proof of this yet as I haven't mod my mount, but I plan to do it later this year.

There is a thread on the forum about this mod that is quite positive

[Demonperformer]

Yes, I understand that belt-driven mounts are much more accurate and quieter, and for someone with that degree of mechanical competence it could be a way forward.

I have a formula to calculate the probability of me personally succeeding in any particular take-apart-put-back-together task. Where the number of instructions involved is n then the probability of my success is 1/((n/2)^2), so with 44 instructions that works out at a fairly small number .

Were I more successful with mechanical alterations, I would probably give it a go, but in the circumstances, I think PEC would probably work out better for me (and my mount!).

Thanks.

[libraryman]

Great topic,

I would like to try autopec however I had heard that it wasn't best suited to an eq8 due to an issue with the worm period not being correctly implemented ......or something like that!!!!

My guiding is in the 0.06 RMS region (best numbers) but longer exposures I.e. 5 SEC's don't seem to smooth out the guiding. Maybe pec would help here.

Ray

[Vox45]

Depending on the kind of mount you own (HEQ5 or HEQ6) there is another way to improve guiding. It is to mode your mount using Rowan's Belt Modification Kit

This would improve tracking a lot as it would significantly reduce the gear errors (and make the mount quieter!) My guess would be that the belt mode +  guiding alone would be enough for long subs. I have no proof of this yet as I haven't mod my mount, but I plan to do it later this year.

There is a thread on the forum about this mod that is quite positive

Just adding to this topic: there is a video by Chris explaning how you can create a custom gear ratio in EQMOD when a belt mode has been made. I did not know that this was needed so I am sharing this just in case....

Surely this is important to know

[ollypenrice]

I was brought up in the "if it ain't broke fix it till it is, then take one step back" school. Probably not good if we're all like that but it maybe beneficial to have the odd person around that works that way.

Yes, that is a very fair point because those idle dossers of my persuasion are, if we did but know it, doubtless counting on you lot behind the scenes!    Low bow. Point taken.

Regarding the 30 minute sub, though, my own experience is that this relies pretty well exclusively on polar alignment. One wheel rotation has always, for me (on numerous mounts) established the level of guiding error. A second rotation has always remained in the same reference frame of error. Now this might not be so for everyone, of course. But for me the thirty minute sub depends, above all, on polar alignment.

Olly

[Demonperformer]

I know this is an old thread, but it was started by me, so (despite the fact that my progress in the subject could be described as "glacial" were it not that glaciers do move eventually) I feel justified in adding a couple of points to it, in case anyone stumbles across this thread while searching for the same topic.

(1) This document is one I have only recently stumbled upon, and pages 10-12 are well worth a read (as is the rest of the document, but those are the pages that relate to this discussion).

(2) I have started doing some "widefield" shots and, although with a short focal length lens I have not had any "trailing" issues, it does occur to me that, with a longer lens and/or longer exposures, without the hassle of fitting a guider, PEC may prove to be very useful. Of course, I will need to get my guiding sorted out before I can use the AutoPEC feature and test this theory, but my gut-feeling is I think it may prove useful once I have managed that.

Thanks.

[michael8554]

Hi demonperformer

I've got bad news for you on your progress, I'm afraid glaciers are going backwards these days........

Michael

[StuartJPP]

On 02/10/2015 at 10:48, ollypenrice said:

Hmmm... I think a lot depends on the likely productivity of the faffing. Even with the number of clear nights I get here I would rather be imaging than faffing. I know you've advocated PEC plus guiding before and I respect that, but I've never tried it because I'm of the, 'If it ain't broke' school. I also try, wherever possible, to subtract rather than add complexities. 

Like anyone else's my guiding could be better but when it's good enough my philosophy is to get on with it. This is a farily typical graph from any of my mounts. (Lodestar, Bin 2, FL400mm, graph in pixels.)

With a lot of effort I could doubtless get it down to the values I obtain on the best of nights...

... but the effect on the resulting pictures would not, in my view, be discernible. (In fact the graph above made me jump out of my skin because I thought the darned thing had hung up!)

However, if the OP does have a problem and his system is 'broke' then trying feed-forward in conjunction with feedback makes sense. If ever I feel I need it I'll certainly try it myself.

Olly

I think that the important thing about your graphs is that the scale is shown in "pixels".

I understand pixels, anything less than 1 is okay, anything less than 0.25 is good and anything less than that is great....I know this equates to arcseconds at the end of the day, depending on your pixel size and focal length, but pixels are much easier for me to understand.

I can imagine a star being captured on the sensor covering 5 pixels and it deviating by only 0.25 pixels...

[ollypenrice]

35 minutes ago, StuartJPP said:

I think that the important thing about your graphs is that the scale is shown in "pixels".

I understand pixels, anything less than 1 is okay, anything less than 0.25 is good and anything less than that is great....I know this equates to arcseconds at the end of the day, depending on your pixel size and focal length, but pixels are much easier for me to understand.

I can imagine a star being captured on the sensor covering 5 pixels and it deviating by only 0.25 pixels...

Yes, but don't forget that the pixel scale of your guide cam is probably not the same as that of your imaging scope. In my case I image with the Taks at 3.5"PP and my guider works at about twice that, so the guide graph is only half as good as it looks!

Olly

[StuartJPP]

28 minutes ago, ollypenrice said:

Yes, but don't forget that the pixel scale of your guide cam is probably not the same as that of your imaging scope. In my case I image with the Taks at 3.5"PP and my guider works at about twice that, so the guide graph is only half as good as it looks!

Olly

Correct of course Olly, my ratio is about 1:2.5  but mentally it is much easier for me and with me using only 124mm guiding focal length it is no wonder my arcseconds graph looked like a mountain range...