SGP Autofocus Questions

[kirkster501]

Hey all,

Got my autofocus going at last after weeks of cloud.  Upped my step size and all seems good.

Got a coupe of questions please?  I notice that sometimes the system goes through the 9 steps, does a perfect V curve on those 9 steps and jobs' a good 'un. Other times though, it does the nine steps, but flat at the bottom of the V (more like a mirrored "J")  then decides to restart and takes about 15 steps to eventually deliver the same [excellent] focus result.  Trying to understand why it does that sometimes?

Secondly, with a steady temperature (albeit at 14C last night), I did the above five times for each of my filters - LRGBHa and averaged them out per filter to give a focus point for each.  So if I set the "focus with filter" against each of the filters it is that parameter that determines if the system is going to use LUM as the focus filter (and then apply the offset) OR to use that actual filter (RED say) and then run the autofocus with Red?   Is that how it works?  it is that parameter "focus with filter" that determines what filter gets used?

So for example lets say I had:

LUM  Focus with filter LUM

RED Focus with filter RED

GREEN Focus with filter GREEN

BLUE Focus with filter LUM

HA Fcus with filter LUM

In the last two, it tells the system to use LUM as the focus filter and then use the offset to apply to BLUE and Ha *after* focusing with LUM?  An for Red and Green, the system does the nine step routine with Red and Green???

And finally, I established the relationship between the filters with the focus offsets.  But that was at 14C.......  So let's say the temperature is now 2C - the focus point will be a lower number. But is the *relationship* or delta consistent between the filters?  For example, My Lum is 15000 and my Red is 15015 at 14C.  At 2C, will there still be a 15 step difference?  Surely it must since otherwise the offset principle could never work.....? It's setting the *first* focus with say Lum that is critical and then the system, applies the offset no matter what the temperature.

Thanks for reading my War and Peace question!

Steve

 

 

[steppenwolf]

48 minutes ago, kirkster501 said:

But is the *relationship* or delta consistent between the filters? 

The relationship is constant so the offsets that you have 'discovered' will work fine for our range of temperatures.

[gnomus]

Set focus with filter.  Set it to Lum.  All focussing will then be done through the Lum filter and the appropriate filter offset will then be applied.  You need to supply these offsets in the set filters window.  It makes little difference for LRGB but may save time on NB.  I would check each filter offset several times going back and forth between Lum and the chosen filter.  This is because your temperature is changing all the time.  Use an average of the difference.  The offsets are relative - it doesn't really matter what you have under Lum just add or subtract your offsets from this figure.

15 points.  Mine does this.  It happens because you finished late at night when it was cold.  Now you start up again when its warm. The focuser is in the wrong place.  It achieves proper focus before it gets the three points on the right of the curve that it needs to calculate a slope.  So it starts again with a much wider initial offset.

[swag72]

I use both offsets and focusing in all filters. Offsets were easy to do with the Tak.... With the TMB, not a chance! So I just bite the bullet and accept that narrow band focus is a long old process.... Focus is bang on though, so no real complaints.

[steppenwolf]

49 minutes ago, swag72 said:

With the TMB, not a chance!

That's an interesting observation, Sara, why is that? I routinely use offsets from Luminance with my Ha and OIII filters on the Esprit 150 and they work fine - I wonder what the big difference is?

[swag72]

9 hours ago, steppenwolf said:

That's an interesting observation, Sara, why is that? I routinely use offsets from Luminance with my Ha and OIII filters on the Esprit 150 and they work fine - I wonder what the big difference is?

I do wonder if it's because I tried to do them early on before the scope had really settled......... But for the sake of a few minutes I'm not bothered really

[kirkster501]

11 hours ago, swag72 said:

I use both offsets and focusing in all filters. Offsets were easy to do with the Tak.... With the TMB, not a chance! So I just bite the bullet and accept that narrow band focus is a long old process.... Focus is bang on though, so no real complaints.

I don't think it remains clear for ten minutes in UK to do autofocus with narrowband Sara......!  In a worse case it could take that long with 15 sec exposures.  On a night where you'd want to autofocus due to dropping temperature every 1C (or even 0.5C) I don't think it is feasible in UK.  Offsets are the way with NB I think.....  Be interested what the other guys think?  I am thinking of autofocus for LRGB and then offsets for NB (of which I only do Ha at the moment)

[swag72]

1 hour ago, kirkster501 said:

.... I am thinking of autofocus for LRGB and then offsets for NB (of which I only do Ha at the moment)

As far as I know in SGP you either use offsets or you don't..... I don't think you can specify in which filters you want to fully focus and in which you want to use offsets.... or can you? 

[gnomus]

32 minutes ago, swag72 said:

As far as I know in SGP you either use offsets or you don't..... I don't think you can specify in which filters you want to fully focus and in which you want to use offsets.... or can you? 

I think you are correct.  It's all or none.  

[kirkster501]

So you guys using all then?  You run autofocus routine on every filter?  So the "offset" sets the start position for that filter?

[gnomus]

I focus every filter through the Lum filter then SGP applies the offset. 

[steppenwolf]

I used to focus using the filter in question and this worked out very well for me BUT I then got 3nm NB filters and this changed the whole thing!! Now I have gone back to offsets from Luminance and this works well too although I must echo the need for very careful measurement of the offset and carrying this out in the shortest period of time possible to obviate the effects of temperature change during your original calibration. I don't use SGP but cannot think of any reason why SGP won't do what is required!

[gnomus]

5 minutes ago, steppenwolf said:

I used to focus using the filter in question and this worked out very well for me BUT I then got 3nm NB filters and this changed the whole thing!! Now I have gone back to offsets from Luminance and this works well too although I must echo the need for very careful measurement of the offset and carrying this out in the shortest period of time possible to obviate the effects of temperature change during your original calibration. I don't use SGP but cannot think of any reason why SGP won't do what is required!

What I found was that I would focus in Lum (get value X); focus in (say) Ha (get value Y); focus in Lum again (get value Z).  Sometimes Z did not equal X!!!  So you do need to be careful.   However, the control is so fine with the stepper I doubt that the few points difference between X and Z was significant.  I did 5 'pairs' of Lum/Ha before deciding what my Ha offset was.  And so on through all the filters.... It may take a couple of nights.  But hey - we are 'testers' not imagers aren't we?    

[steppenwolf]

4 minutes ago, gnomus said:

However, the control is so fine with the stepper I doubt that the few points difference between X and Z was significant.

I think this is a 'modern' issue, Steve, we see digital resolution and assume that it translates to real life resolution! I would expect some variance in the numbers but for the focus itself to be comfortably within tolerance but taking too long over the task does open you up to false figures caused by temperature change. Each of my step sizes equates to 0.81 micron!

[kirkster501]

You don't have to work out the offset though right?  You calculate the absolute focus position for each filter?  So, say Lum = 20000 and that Ha is 20200.  The offset is 200 and that is applied as the delta to the focus calculated on Lum?  So say the temp is colder that night and Lum comes into ficus at 19500.  Then, Ha will be 19700 - 19500 PLUS the offset of 200?  I did not have to calculate that offset, SGP calculated it by subtraction.  My concern is that the delta of 200 itself changes as the temperature drops, i.e. the relationship of the offset and temperature is not linear?

I did all mine the other night.  I did 5 focus runs on each filter and took the average.

Are you guys sure you can't have offsets AND autofocus through the different filters?  In the filter window you can select on each filter what filter to focus with.....?  So against RGB you say focus with RGB respectively but against Ha I say focus with Lum (and then apply the offset)??????

[Barry-Wilson]

1 hour ago, kirkster501 said:

You don't have to work out the offset though right?  You calculate the absolute focus position for each filter?  So, say Lum = 20000 and that Ha is 20200.  The offset is 200 and that is applied as the delta to the focus calculated on Lum?  So say the temp is colder that night and Lum comes into ficus at 19500.  Then, Ha will be 19700 - 19500 PLUS the offset of 200?  I did not have to calculate that offset, SGP calculated it by subtraction.  My concern is that the delta of 200 itself changes as the temperature drops, i.e. the relationship of the offset and temperature is not linear?

I did all mine the other night.  I did 5 focus runs on each filter and took the average.

Are you guys sure you can't have offsets AND autofocus through the different filters?  In the filter window you can select on each filter what filter to focus with.....?  So against RGB you say focus with RGB respectively but against Ha I say focus with Lum (and then apply the offset)??????

Hi Steve

You are right, once you have measured the offset by repeated AF runs using the actual filter you want an offset to apply, eg Ha, you input the focus numbers on the filter tab, as Steve had outlined in an earlier thread.  Then SGP just applies the delta offset (eg if Lum is 1850 and Ha is 1861, an offset delta of + 11 is applied) when it carries out the AF with your chosen AF filter, eg Lum.  You specify which filter to use for AF in the Filter tab.  You must also select the adjust filter option as Steve said.

1 hour ago, kirkster501 said:

You don't have to work out the offset though right?  You calculate the absolute focus position for each filter?  So, say Lum = 20000 and that Ha is 20200.  The offset is 200 and that is applied as the delta to the focus calculated on Lum?  So say the temp is colder that night and Lum comes into ficus at 19500.  Then, Ha will be 19700 - 19500 PLUS the offset of 200?  I did not have to calculate that offset, SGP calculated it by subtraction.  My concern is that the delta of 200 itself changes as the temperature drops, i.e. the relationship of the offset and temperature is not linear?

I did all mine the other night.  I did 5 focus runs on each filter and took the average.

Are you guys sure you can't have offsets AND autofocus through the different filters?  In the filter window you can select on each filter what filter to focus with.....?  So against RGB you say focus with RGB respectively but against Ha I say focus with Lum (and then apply the offset)??????

You're right: when you select which filter to use for AF, eg Ha use Lum, OIII use Lum, Red use Red, the specified filter will be chosen for AF then offset applied.  If you chose to use Red for the Red filter, the offset is zero.  In this way you are effectively only using Lum for the NB filters and the native filter for the RGB (and Lum of course, as you use Lum ).

HTH - or have I just made it more confusing ?

[kirkster501]

Hi Steve (Gnomus), Barry,

Would you post a screemshot of your SGP "filters" window and your AF settings window please so I can see how you two have done it?  I know SteveR uses Maxim.

Thanks. Steve

[sloz1664]

Hi Steve,

 

Here's my settings:-

 

Steve

 

[kirkster501]

Thanks Steve.

Interesting you have the "default" set.  I selected the actual filter itself, so RED against RED, Blue against Blue thinking that this tells SGP to use that actual filter to focus.

Would you also send your screen shot of your autoficus settings as well please?

 

[sloz1664]

Hi Steve,

 

5 minutes ago, kirkster501 said:

Thanks Steve.

Interesting you have the "default" set.  I selected the actual filter itself, so RED against RED, Blue against Blue thinking that this tells SGP to use that actual filter to focus.

Would you also send your screen shot of your autoficus settings as well please?

 

I just imported my original filter settings from MaximDL.

 

Here's my focus settings:-

 

 

Steve

 

 

 

 

 

 

 

 

[gnomus]

4 hours ago, kirkster501 said:

Hi Steve (Gnomus), Barry,

Would you post a screemshot of your SGP "filters" window and your AF settings window please so I can see how you two have done it?  I know SteveR uses Maxim.

Thanks. Steve

As commanded (this is my Esprit 120 with a Feathertouch focuser but the principles are the same):

Under 'Set Filters':

Note that it doesn't matter what Lum is set at (although having it somewhere near a number where you once might have had focus seems to me to be sensible).  In this case there is a -44 difference between Lum and Red.  If, on the night, Lum focuses at 15400, SGP will just remove the 44 steps when using the Red filter because SGP uses relative offsets.  You will probably not be able to find an 'absolute' position for Lum during the night you run your tests since there is no such thing.  Lum will focus at different points as the temperature and other factors change.  You may as well, therefore, pick a number that makes the subsequent arithmetic easy.

This is what I have under the 'Focus' Tab in my Equipment Profile:

Nothing especially exciting here.  And finally, when I click the button with the yellow arrow (please note, if you have a yellow arrow next to your button you should contact SGP), I have:

Some of these will be equipment dependent and others will be matters of taste.  I have found that I need to refocus every 0.5 degrees with my Esprit 120, whereas 1 degree is fine with the WO Star 71s.  You might decide you want to check focus every 60 minutes irrespective of whether there has been a temperature change.  It makes most sense to me to autofocus with Lum.

Set a big number for backlash.  This is not something you want to try to calculate 'precisely'  (again probably impossible).  Rather you want to make sure that you move the focuser enough to definitely eliminate any problems with backlash.  If you forget to have a decent amount set for backlash you might find that your autofocusing seems to make no sense - I have been there and I have done that.  I refused , however, to buy the T-Shirt, since the ones they had in my size were a colour that would, in my view, have clashed with my hair (especially now that I have had my highlights refreshed).

Steve

  

[alan4908]

Hi Steve

Three additional considerations that you may want to consider when performing autofocus operations:

1. Focus Star Brightness

I don't use SGP for autofocus but rather FocusMax controlled via MaximDL/ACP.

ACP contains a script to work out filter offsets for each of the filters. It does this by focusing through each filter, working out the filter offset relative to the filter that you wish to use in your autofocus operation (in my case the L filter). This procedure is repeated fives times, it then uses the average offset in subsequent autofocus operations.

In working out the filter offsets, by default, ACP searches for a mag 6 to 9 star. Generally, I've found that this works well for my LRGB filters, allowing the star not to be saturated in the CCD (when in focus) or too faint when calculating the focus positions at the far and near focus points. However, when using 3nm narrowband filters, I've found that I needed to override the default magnitude search and instruct ACP to search for slightly brighter stars. If I don't do this, then the far focus point calculation of FocusMax sometimes get inconsistent offset results since there's insufficient light for the CCD to take a reliable measurement. 

2. Focus Star Location

In ACP you have two choices:  search for a focus star near the object your imaging or search for the focus star near the zenith. I've always chosen the latter option, since it seems logical to try to minimize the impact of the atmosphere when focusing. Apparently, some ACP users claim that this approach gives significantly better results, however, it 's not something that I've tested. Interestingly, I've never noticed any significant downside to this approach eg erratic mount issues arising from the larger slews. 

3. Filter selection for focusing (assuming the use of filter offsets)

If you believe that it's a good idea to minimize the effect of the atmosphere when focusing but you also wish to search for a focus star near the object your imaging, then I presume you'd get more consistent focus results if you selected the Red rather than the Lum filter as the default focus filter. 

I hope this helps.

Alan

[kirkster501]

Brilliant stuff guys, thanks so much!

[kirkster501]

I think what I am going to do is to create two SGP profiles.  One for the equipment with LRGB and another identical profile but with LHa.  That way I can focus through the filter I intend to capture with on the LRGB and use offsets when I want LHa.

I will test this when we get another clear night since we have had our one clear night per month for this month already......

Steve (Gnomus) all meridian flips etc work fine now with CdC but still need a clear night to test my overpole/underrpole settings with SGP.  The gin bottle is in the cupboard for you, I nearly raided it the other night!

[sloz1664]

Hi Steve,

I've created lots of profiles for my different scopes / cameras, including lrgb, HaOiiiSii, bias, darks & flats. Comes to quite a list. 

 

Steve