[Robert N]

From the Pentax operating instructions for the K-3iii astrotracer type 3, it says avoid ground scenery in images. On this basis, it seems like the usage is limited to star/constellation images only. I presume the only way to include foreground interest is to take a set of conventional, relatively short exposure (say 10s) images and combine with software such as Sequator to reduce image noise and allow foreground inclusion. Unless anyone has managed to use a technique with astrotracer type 3 for longer exposures with say a 35mm lens and successfully included foreground interest. Interested in seeing your replies.

[kypfer]

The Type 3 Astrotracer is limited, to a degree, because it relies on analysing subject movement to calculate it's compensations, using what is effectively a test exposure.
As ground scenery doesn't move (usually), including it in the image causes the analysis to fail.
For scenery inclusive 'astrotracing' you'll need the auxiliary O-GPS2 module … this will allow an image of stationary stars and blurred local scenery … stationary stars and stationary scenery will only ever be a compromise (Type 2 Astrotracer?) or using more sophisticated third-party applications.
There's a lot of useful information here PENTAX: The choice for astrophotography / explore | RICOH IMAGING and here GPS UNIT O-GPS2 | RICOH IMAGING

[MossyRocks]

Having used it a fair amount some ground can be in the image but generally try to avoid it if possible. Also from using it it really needs you to have a good focus, proper focal length entered, a non fogged lens, and don't have a gigantic object in the frame. I have had all of those problems some were deliberate in trying to figure it out and trying to figure out what causes the "data processing error", others were just happenstance things. The bad focus and fogged lens were things I discovered by accident, apparently it really struggles with really bloated stars that result in the test shot or shots (still don't know). The gigantic object was when I tried it with a full moon in the frame at 800mm and was part of my figure it out testing. Same thing with inaccurate focal length. Surprisingly it will pick out stars even in very heavy light pollution and can figure out how to track them. I tested this when doing my test shots in my bright bortle 8 backyard with my 400mm lens with the full moon just out of the frame and it found stars no problem and tracked them.

Usually when I shoot landscapes or night scenes with a foreground I am running a wide to ultrawide lens and there I just ride the rule of 200 for my exposure time (200 / focal length=exposure time in seconds). From there I take a pile of images and will stack them in software like Sequator which can freeze the ground and the sky.

[Robert N]

The Type 3 Astrotracer is limited, to a degree, because it relies on analysing subject movement to calculate it's compensations, using what is effectively a test exposure.
As ground scenery doesn't move (usually), including it in the image causes the analysis to fail.
For scenery inclusive 'astrotracing' you'll need the auxiliary O-GPS2 module … this will allow an image of stationary stars and blurred local scenery … stationary stars and stationary scenery will only ever be a compromise (Type 2 Astrotracer?) or using more sophisticated third-party applications.
There's a lot of useful information here PENTAX: The choice for astrophotography / explore | RICOH IMAGING and here GPS UNIT O-GPS2 | RICOH IMAGING

Kypfer,
Many thanks for the for the link to the very comprehensive article - perhaps this will be updated in the future to include the Type 3 Astrotracer. MossyRocks has consolidated my original thoughts of using the conventional technique (with typically Sequator software) to include foreground interest. Another option would be to use a composite of a Type 3 generated sky and overlay a separate image of foreground interest.
Unfortunately here in the UK, ideal opportunities for astrophotography are limited due to frequent cloud cover and light pollution. My own limited trials with a 35mm lens at f2.8 required a good black sky for the Type 3 to function. Perhaps the MossyRocks example of a long lens enlarged the stars sufficiently to allow the Type 3 system to pick them up more easily. However, this does not detract from the excellent innovation and ease of use of the Type 3 system.

[MossyRocks]

My own limited trials with a 35mm lens at f2.8 required a good black sky for the Type 3 to function. Perhaps the MossyRocks example of a long lens enlarged the stars sufficiently to allow the Type 3 system to pick them up more easily. However, this does not detract from the excellent innovation and ease of use of the Type 3 system.

I have tried it with lenses as short as 28mm and type 3 works fine even in my very badly light polluted backyard which is a bortle 8 location for light pollution with a 9 about a mile away. If you want to find out how bad your location is you can check it here. I even tried shooting with the full moon just out of the frame and it was able to find enough stars to calculate tracking. From what I have experienced, type 3 will fail to track if the stars get too big like what happens when you are out and the lens gets fogged, or your focus is off. My guess is that unless you manually focused with a bahtinov mask or really fiddled around in magnified live view looking to turn on the dim stars that will flicker if focus isn't perfect your problem was that the focus was off by a bit. With astrotracer it works better for longer lenses as they have a narrower field of view so the movement throught the frame is more similar. Once one gets much shorter than 28mm the corners start suffering as the difference in movement from the center of the frame to the corners is just too different and you get streaking. That isn't to say you can't use it to extend your exposure time beyond the good rule of 200 (200/focal length) but I haven't done that. I should try that out and see if I can get a good 40s exposure with my 12mm as that would double the exposure time.