[sindbad1]

20s exposures with 560mm are good without a real telescope mount. I think that with such a long focal distance the camera must know exactly the real focal distance to track correctly. If the real focal length deviates too much from the specified one, errors occur earlier.

[Leumas]

......So something must be up with mine cause I wasn't able to hardly get 6 seconds on 300mm on K1

[MossyRocks]

So something must be up with mine cause I wasn't able to hardly get 6 seconds on 300mm on K1

It is finicky and requires good calibration. Going past 180 degrees of movement in each direction I find helps. Also you need to be careful about things that will throw it off, power line, transformers, iron (your car counts), fencing, etc all easily throw it off. Even soil with high iron concentrations will throw it off as I can never get good tracking up in the iron range of northern Minnesota. I also found out at one point that if your electronic level sensors are not accurate you will get garbage tracking. I had that with my K-3ii and it needed to get sent in to get that corrected.

[gaweidert]

I put my D FA 150-450 lens on my K1-II and it would not calibrate. Installed my D FA 28-105 and it calibrated right off.

[Leumas]

Do I have to use it in BULB?? Think I was just using manual, but read some other cameras required BULB for astro....

I put my D FA 150-450 lens on my K1-II and it would not calibrate. Installed my D FA 28-105 and it calibrated right off.

Now that raises an interesting question. I've never had a lens on the camera while doing the calibration routine. I figured that to be a null-effect condition, and since the camera's easier to move in space without the lens, that's what I've always done. I'd be interested to know whether anyone else has information that would answer the question, whether having a lens on the camera makes any difference during calibration. (I can't really even picture doing calibration with the 560mm+1.7xTC on the K-1.)

[slartibartfast01]

Now that raises an interesting question. I've never had a lens on the camera while doing the calibration routine. I figured that to be a null-effect condition, and since the camera's easier to move in space without the lens, that's what I've always done. I'd be interested to know whether anyone else has information that would answer the question, whether having a lens on the camera makes any difference during calibration. (I can't really even picture doing calibration with the 560mm+1.7xTC on the K-1.)

Some people even perform the calibration with the tripod attached.

Some people even perform the calibration with the tripod attached.

I've never seen any authoritative source on this issue. It would be interesting to know whether anyone has done any kind of real testing and experimentation to determine whether it does make a difference. Perhaps the metal in the lens and the tripod would throw off the measurements?

[photoptimist]

I've never seen any authoritative source on this issue. It would be interesting to know whether anyone has done any kind of real testing and experimentation to determine whether it does make a difference. Perhaps the metal in the lens and the tripod would throw off the measurements?

Yes, if either the lens or tripod contain any ferrous metals or magnets, they will distort the magnetic field around the camera. Ricoh's documentation even mentions that changing the angle on the LCD tilt screen can disturb the precision calibration.

It only takes a couple of degrees error in the magnetic compass to cause star trailing.

[gaweidert]

Well I figure that since you are not supposed to mount or change lenses with the camera on, the only correct way to do the calibration for Astrotracer is with a lens already mounted. It simply would not calibrate with the D FA 150-450 lens mounted. I wonder if I told the camera it was a 100mm lens the system would calibrate.


As I was going through the gyrations necessary to perform the calibration, all I could think of was that the front half of the lens would fall off as has happened to a few member here.

[MossyRocks]

Do I have to use it in BULB??

On my K-3 and K-3ii it requires setting it into bulb mode but when the GPS in enabled there is a sub option in bulb mode for astro tracer.

Now that raises an interesting question. I've never had a lens on the camera while doing the calibration routine. I figured that to be a null-effect condition, and since the camera's easier to move in space without the lens, that's what I've always done. I'd be interested to know whether anyone else has information that would answer the question, whether having a lens on the camera makes any difference during calibration. (I can't really even picture doing calibration with the 560mm+1.7xTC on the K-1.)

I'm not sure how things are with the 560mm+1.7x converter but when I do calibration and am planning on using a monster lens like my 16 pound SMC A* 400/2.8 I do my calibration with a smaller lighter lens. The one I most often used is the SMC A 50/1.2 as it is usually already on the camera but I will also use the DA 35/2.4 plastic fantastic. In using those I haven't noticed any discernible difference in the quality of calibration and tracking. I can reliably get great tracking results at 400mm such that I am able to get clean stacks with nice round pinpoint stars.

Here is a single shot of M42 (the great orion nebula) I took last year with the SMC A* 400/2.8 at 20 seconds using astrotracer. This was a throwaway test shot after I first got everything setup, focused and calibrated and the lens was still cooling down so it is as bad as that lens can be even when focused properly. Even there I have nice round but bloated stars (if pixel peeping the bloat is actually coma). And yes with big glass you do need it to come to temp and for everything to settle down so your 560mm would also need to come into thermal equilibrium. Here is something a bit more recent that is a stack of 20 second astrotracer shots on M31 (the Andromeda galaxy) that have been processed. In both of these examples I calibrated with either the A 50/1.2 or DA 35/2.4 on and then put the A* 400/2.8 on. With that lens I have to tell the camera the focal length I am using and with the 1.7x converter you may have to do that as well. However that combo gives 952mm so I am not sure how close you can get, if you can put in 950mm you will be good, however if you have to put in 900mm or 1000mm expect poor results. Also at almost 1000mm I wouldn't expect great results given that things are really moving across the frame fast so even 10 seconds may exceed the tracking ability of astro tracer.

With my 400mm plus the 1.4X-L converter I have to manually enter the focal length and the closest I can get is 550mm and the best I have managed with that is 10s of good tracking. I have tried for 20s every time and there are very clear star trails. I have never tried astro tracer with anything longer than 560mm eventhough I have access to the MTO 1000 f/10 but it is so slow that I only chase big bright things in the night sky with it like the Moon or Jupiter which don't require astrotracer.

Well I figure that since you are not supposed to mount or change lenses with the camera on, the only correct way to do the calibration for Astrotracer is with a lens already mounted. It simply would not calibrate with the D FA 150-450 lens mounted. I wonder if I told the camera it was a 100mm lens the system would calibrate.

It was my understanding that the camera, at least the K-1, remembers it's most recent calibration as long as the batteries are installed. So it shouldn't matter whether you turn off the camera in order to mount a lens. This is a matter that is not well described, and I'm relying mainly on Ricoh website information regarding the add-on GPS unit sold for cameras like KP. This astrotracer business is very murky, and it appears that the people who designed it either don't know how it works in practice or failed to document their endeavors.

As I was going through the gyrations necessary to perform the calibration, all I could think of was that the front half of the lens would fall off as has happened to a few member here.

Not once, but twice! (although the first occurance was due to having bought a used lens from BHPhoto which had been described as 10/10 but which had screws missing in the mount hardware, so the lens fell off the camera with the mount still intact - not a Ricoh problem, that one.)

[MossyRocks]

It was my understanding that the camera, at least the K-1, remembers it's most recent calibration as long as the batteries are installed. So it shouldn't matter whether you turn off the camera in order to mount a lens.

Yes it will remember, and will even remember for a while when the battery is out too. I have frequently changed batteries in the camera during a session while using astrotracer and things worked just fine without having to recalibrate. On especially cold nights I may change out the battery 3 or 4 times all without issue.

This astrotracer business is very murky, and it appears that the people who designed it either don't know how it works in practice or failed to document their endeavors.

It is some black magic and one of those features that most people don't need but once you find you want to use it you see just how useful it is. It basically turns the IBIS into an alt-az mount with field rotation allowing for limited tracking of the night sky. It makes use of you GPS coordinates to know how far north or south you are so it knows where the celestial equator and poles are is, the electronic compass so it knows what cardinal direction the camera is pointed, and the electronic level sensors to know where in the sky it is pointed and how the camera is oriented. With this information it will have enough to calculate how to move the sensor for good tracking. When you read the info about the O-GPS1 it will give estimated tracking time based off of the lens length and the declination but declination is not the angle from the horizon to the zenith but is the angle from celestial equator to the celestial pole. So basically if you center polaris (the north star) in the view finder you are basically at 90 degrees for your declination, however if you center the upper most star in Orion's belt you are at 0 degrees for your declination. When you are pointed at the celectial pole (declination of 90 degrees) the movement is all rotational and in this case the IBIS can rotate the frame about 1 degree which is what gives the tracking time of 5 minutes. All lens will give a tracking time of 5 minutes if centered on the celestial pole however that assumes a good calibration. When you are shooting things along the celestial equator there is no rotation only linear movement so the sensor doesn't rotate but moves in a nice line. Here we are not concerned with the angular speed but the linear speed across the frame and the time it takes to do that. Here lens length matters and a long lens like my 400mm will allow for only a short time while something like a 28mm will still allow for something close to that 5 minutes of tracking. Anything between a declination of 0 and 90 degrees will be some combination of linear motion and rotational motion and the amount will vary. In general though you will get longer exposures by shooting closer to the celestial pole.

This however brings us to the odd case of very wide lenses. Pointing one at the celestial pole works great and getting oddities in the edges and the corners isn't a problem since the movement is all rotational. Even something like a fisheye would work well when pointed at a celestial pole even though it has a huge amount of distortion. However if you centered an ultra wide along the celestial equator you would end up with a lot of streaking of stars in the corners with it getting worse as you move away from the celestial equator. In this case the sensor is moving in a nice line along the celestial equator but the lens is so wide there is a rotational component that it isn't capturing for the parts of the frame that aren't along the celestial equator.

So you now probably have more info than you ever cared to on the inner workings of astrotracer but you know:
You can likely change lenses without having to recalibrate (if the lens has magnets in it like a PLM lens or if it has a lot of ferrous material you likely will)
You can turn the camera off and on and it will retain the calibration
You can turn the camera off and swap batteries and it will retain the calibration
It is finicky and calibrating it takes practice. It took me a good 6 months before I started getting consistently good results and at first I wasn't impressed.
Power lines, electric motors, excessive iron in the soil, your vehicle, nearby magnetic objects, and other nearby electronic devices will all throw it off.
The longest tracking regardless of lens length can be had at the celestial pole where lens length doesn't matter.
Lens length affects tracking time more the closer you get to the celestial equator.

...
So you now probably have more info than you ever cared to on the inner workings of astrotracer but you know:
...

Actually, that was exactly what I wanted to know, and the best explanation I've seen. Thanks!

[MossyRocks]

Actually, that was exactly what I wanted to know, and the best explanation I've seen. Thanks!

Well good to hear. I don't think most people need that level of info but I do agree that the documentation in lacking. Especially since in the O-GPS1 specifications they use declination but people think 90 is the zenith and 0 is the horizon which I'm sure leads to lots of confusion. They are also light on things to watch out for to ensure good tracking which would be useful. From what I can tell they never intended it to be For me it is one of the most used features of the camera but I know I am likely in the minority.