Pentax O-GPS1 accuracy

Paschalis · 11-10-2011, 07:53 AM

Hello,
I recently got my O-GPS1.
After following all the instructions of the operating manual carefully, I compared (camera was the K5) the coordinates for a given position with those of Google Earth and my Navigon GPS device as well.
To my great surprise, there is a significant difference between O-GPS1 and the other two (Google Earth's and Navigon's coordinates were pretty much the same).
It seems that O-GPS1 is falling 480 to 500 meters outside (towards North-East).
(I have done the AstroTracer calibration too, with the hope that the positioning woould be closer to the other two values, but nothing changed).
Note that the satelite icon in the Display Menu was green (4 or more satelites were detected).
Any idea or opinion for this difference ? Is the O-GPS1 faulty, is this its normal accuracy, or am I missing something?
Your help will be much appreciated

Regards, Paschalis

Gimbal · 11-10-2011, 08:21 AM

No that isn't normal. I have however noticed that one program I used (maybe it was irfanview, but I'm not sure) did something wrong when sending/translating the coordinates from the exif to some map-provider, and this came out as a large error in position.

Paschalis · 11-10-2011, 08:39 AM

I forgot to mention that I read the O-GPS1 coordinates of my images, only from ACDSee (Properties - EXIF)

Gimbal · 11-10-2011, 12:36 PM

Perhaps you could try some other software, for instance geosetter, see GeoSetter

Paschalis · 11-10-2011, 01:05 PM

Thank you very much Gimbal!
Finally, the problem was with ACDSee. GeoSetter is a great application!

KansasHorizons.com · 11-17-2011, 07:55 AM

I am going to post here in an existing thread since I am getting little response in the accessory section.

I cannot get this unit to work correctly. Last night was the third night of failures which ended with the unit refusing to do a precise calibration at all. I came home and decided to give it another try from the front yard. I did the precise calibration again and again and received the OK from the camera. The unit continued to produce trails which were variable and some were even curved. Some were close and a few looked like I was shooting without the unit at all. I tried various lenses, new batteries etc and it all still remained variable. I even tried repeating shots without changing anything to see the trail length change.

After realizing that there have been thousands of the units sold and seeing very few images that have worked I decided to look for anything I could find on the internet this morning related to the unit. With the exception of a few pictures that were really good and a few ok pictures there is little if any overall success with the unit. The best pictures were stacked with 40 and more short duration shots.

One thing many reported was that anything over about 45 seconds started trailing regardless of the reported maximum time. The people who had only shot wide field were generally happy because they don't see the trailing there anyways though many have noticed edge distortion in wide field shots.

In my case either the unit or the cameras SR is bad. I bought mine from Amazon so I will give it one more chance and get it replaced. Sadly though I think this thing is very hit and miss with a very, very small chance of success. Many people keep assuming that they must be messing something up according to the forums but they never posted success and I couldn't find that they ever did succeed, I think they just gave up.

I was so excited when this came out, I wasn't going to have to rebuild my wooden double arm tracker that had got wet and warped but at this time I can say that even with it warped I was getting far better shots with it than I am with this unit.

If you are having better luck then by all means encourage me, if not then I cam curious if your story is the same.

Gimbal · 11-17-2011, 08:38 AM

What lenses and how long/short exposures have you tried?
Have you checked that the compass gives you a reasonably correct reading after calibration?

KansasHorizons.com · 11-17-2011, 09:28 AM

Originally posted by Gimbal

What lenses and how long/short exposures have you tried?
Have you checked that the compass gives you a reasonably correct reading after calibration?

I will say that the compass generally stinks. It seems to read the East and West very accurately but North and South can be off by 30 degrees and rarely seems to settle down right. I have read that this was not connected to the astrotracer but if it really is then this might be the issue.

1 through 4 min exposures. Lenses have varied from the tamron di90 to a 50 mm pentax to manual 135 mms etc.

I am having another sent to try out. Hopefully it will work better.

Gimbal · 11-17-2011, 11:53 AM

The compass is a very critical part of the astrotracer. The apparent movement of the stars is way different depending on in which direction you look. So a successful calibration of the compass is important.

I'm still trying to figure out the best way of doing it. I usually spin the camera over 360 degrees in a steady pace, (instead of the +-180 degrees). Sometimes I spin it while on the tripod, (theoretically that would be the ideal way of doing it since any magnetic disturbances from the tripod would be part of the calibration) but when doing the rest of the calibration I have to detach it from the tripod and this is bit of a mess.

An alternate method is holding the camera steadily in front of you and rotate your whole body (and camera) 360 degrees in a steady pace. Smooth and steady is the key when calibrating, the violent throwing around of the camera that can be seen on some youtube clips are just ridiculous.

Start with shorter exposures and work your way up until failure. It seems the precision of the sensors aren't really up to the task for maximum lengths of exposure without traces.

Also check that your virtual horizons are ok, apparently some K5 have a tilted sensor and that will of course not help one bit with the astrotracer either.

KansasHorizons.com · 11-17-2011, 01:10 PM

Originally posted by Gimbal

Also check that your virtual horizons are ok, apparently some K5 have a tilted sensor and that will of course not help one bit with the astrotracer either.

My horizon is off on my current 5 and the 5 I had replaced in the beginning and my 7 was also off. (Is there a Pentax camera that the horizon is correct? LOL) I really think it is the sensor that is tilted rather than the level off. It was my understanding that the camera does not have to be level because the original calculations to the GPS have already accounted for its position? I would be curious if this has affected anyone else regarding the camera being level.

**jerm1386** · 11-17-2011, 01:30 PM

Not sure the O-GPS1 interfaces with horizon correction, since it works fully on the K-r which doesn't have horizon correction. Can someone confirm?. My guess would be that it has built-in accelerometer that is calibrated to a higher precision when leaving the factory. It shouldnt need to know the side-to-side level, but it does need to know how high up into the sky you're pointed.

KansasHorizons.com · 11-17-2011, 02:10 PM

Originally posted by jerm1386

Not sure the O-GPS1 interfaces with horizon correction, since it works fully on the K-r which doesn't have horizon correction. Can someone confirm?. My guess would be that it has built-in accelerometer that is calibrated to a higher precision when leaving the factory. It shouldnt need to know the side-to-side level, but it does need to know how high up into the sky you're pointed.

My biggest trailing failures were higher up. If one thing was shown to be affecting the trails more it was height. Andromeda was totally trashed, in fact after trying to shoot it was when the GPS wouldn't even re-calibrate for a while.

As far as horizon I am sure if it were important it would have been mentioned somewhere.

Gimbal · 11-17-2011, 02:31 PM

Not with the horizon correction function but with the same accelerometers that measures the left/right tilt angle. And of course it has to know the left/right tilt angle, (it's even part of the calibration procedure) if you tilt the camera 10 degrees the stars will also change direction with 10 degrees.

So to track the stars the camera has to know both the up/down angle and the left/right tilt angle, as well as the shooting heading, the position on earth and the focal length. So there are a lot of measurements that can go wrong.

If you check your exifs (from at least the K5) you can see the up/down and left/right angle has been recorded in every picture from the very first shoot. Kind of cool, but perhaps not that useful info.

Gimbal · 11-17-2011, 02:40 PM

I've noticed that something happens with the compass heading when pointing the camera upwards past 45 degrees. There is some sort of limit there when something suddenly changes and sometimes it affects the heading with up to 20 degrees. Don't know what to do about it though.

KansasHorizons.com · 11-17-2011, 04:32 PM

Originally posted by Gimbal

I've noticed that something happens with the compass heading when pointing the camera upwards past 45 degrees. There is some sort of limit there when something suddenly changes and sometimes it affects the heading with up to 20 degrees. Don't know what to do about it though.

I think there is some serious compass issues. I went outside just to see if the N S offset was a given and the E W wouldn't even line up.... That is still level, not sure if it would be off by more if at a higher elevation or not.

11-10-2011, 07:53 AM	#1
Paschalis New Member Join Date: Feb 2011 Location: Thessaloniki Posts: 14	Pentax O-GPS1 accuracy Hello, I recently got my O-GPS1. After following all the instructions of the operating manual carefully, I compared (camera was the K5) the coordinates for a given position with those of Google Earth and my Navigon GPS device as well. To my great surprise, there is a significant difference between O-GPS1 and the other two (Google Earth's and Navigon's coordinates were pretty much the same). It seems that O-GPS1 is falling 480 to 500 meters outside (towards North-East). (I have done the AstroTracer calibration too, with the hope that the positioning woould be closer to the other two values, but nothing changed). Note that the satelite icon in the Display Menu was green (4 or more satelites were detected). Any idea or opinion for this difference ? Is the O-GPS1 faulty, is this its normal accuracy, or am I missing something? Your help will be much appreciated Regards, Paschalis

11-17-2011, 09:28 AM	#8
KansasHorizons.com Senior Member Join Date: Jan 2010 Location: Kansas Photos: Gallery Posts: 183	Originally posted by Gimbal What lenses and how long/short exposures have you tried? Have you checked that the compass gives you a reasonably correct reading after calibration? I will say that the compass generally stinks. It seems to read the East and West very accurately but North and South can be off by 30 degrees and rarely seems to settle down right. I have read that this was not connected to the astrotracer but if it really is then this might be the issue. 1 through 4 min exposures. Lenses have varied from the tamron di90 to a 50 mm pentax to manual 135 mms etc. I am having another sent to try out. Hopefully it will work better. Last edited by KansasHorizons.com; 11-17-2011 at 09:41 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
O-GPS1 = new Pentax flash system?	DAZ	Pentax News and Rumors	7	10-22-2011 01:40 AM
Pentax O-GPS1 GPS - USA Availability	KansasHorizons.com	Pentax K-5 & K-5 II	47	10-18-2011 07:36 PM
Pentax O-GPS1	Photodana	Pentax K-5 & K-5 II	52	10-04-2011 05:28 AM
Pentax O-GPS1 out of stock	VE2CJW	Pentax Camera and Field Accessories	3	09-20-2011 08:51 PM
Math behind Pentax O-GPS1	kamayok3	Pentax K-5 & K-5 II	8	09-09-2011 05:54 AM

11-10-2011, 08:21 AM	#2
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	No that isn't normal. I have however noticed that one program I used (maybe it was irfanview, but I'm not sure) did something wrong when sending/translating the coordinates from the exif to some map-provider, and this came out as a large error in position.

11-10-2011, 08:39 AM	#3
Paschalis New Member Join Date: Feb 2011 Location: Thessaloniki Posts: 14 Original Poster	I forgot to mention that I read the O-GPS1 coordinates of my images, only from ACDSee (Properties - EXIF)

11-10-2011, 12:36 PM	#4
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	Perhaps you could try some other software, for instance geosetter, see GeoSetter

11-10-2011, 01:05 PM	#5
Paschalis New Member Join Date: Feb 2011 Location: Thessaloniki Posts: 14 Original Poster	Thank you very much Gimbal! Finally, the problem was with ACDSee. GeoSetter is a great application!

11-17-2011, 07:55 AM	#6
KansasHorizons.com Senior Member Join Date: Jan 2010 Location: Kansas Photos: Gallery Posts: 183	I am going to post here in an existing thread since I am getting little response in the accessory section. I cannot get this unit to work correctly. Last night was the third night of failures which ended with the unit refusing to do a precise calibration at all. I came home and decided to give it another try from the front yard. I did the precise calibration again and again and received the OK from the camera. The unit continued to produce trails which were variable and some were even curved. Some were close and a few looked like I was shooting without the unit at all. I tried various lenses, new batteries etc and it all still remained variable. I even tried repeating shots without changing anything to see the trail length change. After realizing that there have been thousands of the units sold and seeing very few images that have worked I decided to look for anything I could find on the internet this morning related to the unit. With the exception of a few pictures that were really good and a few ok pictures there is little if any overall success with the unit. The best pictures were stacked with 40 and more short duration shots. One thing many reported was that anything over about 45 seconds started trailing regardless of the reported maximum time. The people who had only shot wide field were generally happy because they don't see the trailing there anyways though many have noticed edge distortion in wide field shots. In my case either the unit or the cameras SR is bad. I bought mine from Amazon so I will give it one more chance and get it replaced. Sadly though I think this thing is very hit and miss with a very, very small chance of success. Many people keep assuming that they must be messing something up according to the forums but they never posted success and I couldn't find that they ever did succeed, I think they just gave up. I was so excited when this came out, I wasn't going to have to rebuild my wooden double arm tracker that had got wet and warped but at this time I can say that even with it warped I was getting far better shots with it than I am with this unit. If you are having better luck then by all means encourage me, if not then I cam curious if your story is the same.

11-17-2011, 08:38 AM	#7
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	What lenses and how long/short exposures have you tried? Have you checked that the compass gives you a reasonably correct reading after calibration?

11-17-2011, 11:53 AM	#9
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	The compass is a very critical part of the astrotracer. The apparent movement of the stars is way different depending on in which direction you look. So a successful calibration of the compass is important. I'm still trying to figure out the best way of doing it. I usually spin the camera over 360 degrees in a steady pace, (instead of the +-180 degrees). Sometimes I spin it while on the tripod, (theoretically that would be the ideal way of doing it since any magnetic disturbances from the tripod would be part of the calibration) but when doing the rest of the calibration I have to detach it from the tripod and this is bit of a mess. An alternate method is holding the camera steadily in front of you and rotate your whole body (and camera) 360 degrees in a steady pace. Smooth and steady is the key when calibrating, the violent throwing around of the camera that can be seen on some youtube clips are just ridiculous. Start with shorter exposures and work your way up until failure. It seems the precision of the sensors aren't really up to the task for maximum lengths of exposure without traces. Also check that your virtual horizons are ok, apparently some K5 have a tilted sensor and that will of course not help one bit with the astrotracer either.

11-17-2011, 01:10 PM	#10
KansasHorizons.com Senior Member Join Date: Jan 2010 Location: Kansas Photos: Gallery Posts: 183	Originally posted by Gimbal Also check that your virtual horizons are ok, apparently some K5 have a tilted sensor and that will of course not help one bit with the astrotracer either. My horizon is off on my current 5 and the 5 I had replaced in the beginning and my 7 was also off. (Is there a Pentax camera that the horizon is correct? LOL) I really think it is the sensor that is tilted rather than the level off. It was my understanding that the camera does not have to be level because the original calculations to the GPS have already accounted for its position? I would be curious if this has affected anyone else regarding the camera being level.

11-17-2011, 01:30 PM	#11
jerm1386 Site Supporter Join Date: Aug 2011 Location: Baltimore, MD Photos: Albums Posts: 45	Not sure the O-GPS1 interfaces with horizon correction, since it works fully on the K-r which doesn't have horizon correction. Can someone confirm?. My guess would be that it has built-in accelerometer that is calibrated to a higher precision when leaving the factory. It shouldnt need to know the side-to-side level, but it does need to know how high up into the sky you're pointed.

11-17-2011, 02:10 PM	#12
KansasHorizons.com Senior Member Join Date: Jan 2010 Location: Kansas Photos: Gallery Posts: 183	Originally posted by jerm1386 Not sure the O-GPS1 interfaces with horizon correction, since it works fully on the K-r which doesn't have horizon correction. Can someone confirm?. My guess would be that it has built-in accelerometer that is calibrated to a higher precision when leaving the factory. It shouldnt need to know the side-to-side level, but it does need to know how high up into the sky you're pointed. My biggest trailing failures were higher up. If one thing was shown to be affecting the trails more it was height. Andromeda was totally trashed, in fact after trying to shoot it was when the GPS wouldn't even re-calibrate for a while. As far as horizon I am sure if it were important it would have been mentioned somewhere.

11-17-2011, 02:31 PM	#13
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	Not with the horizon correction function but with the same accelerometers that measures the left/right tilt angle. And of course it has to know the left/right tilt angle, (it's even part of the calibration procedure) if you tilt the camera 10 degrees the stars will also change direction with 10 degrees. So to track the stars the camera has to know both the up/down angle and the left/right tilt angle, as well as the shooting heading, the position on earth and the focal length. So there are a lot of measurements that can go wrong. If you check your exifs (from at least the K5) you can see the up/down and left/right angle has been recorded in every picture from the very first shoot. Kind of cool, but perhaps not that useful info.

11-17-2011, 02:40 PM	#14
Gimbal Pentaxian Join Date: Apr 2007 Location: Sweden Posts: 2,106	I've noticed that something happens with the compass heading when pointing the camera upwards past 45 degrees. There is some sort of limit there when something suddenly changes and sometimes it affects the heading with up to 20 degrees. Don't know what to do about it though.

11-17-2011, 04:32 PM	#15
KansasHorizons.com Senior Member Join Date: Jan 2010 Location: Kansas Photos: Gallery Posts: 183	Originally posted by Gimbal I've noticed that something happens with the compass heading when pointing the camera upwards past 45 degrees. There is some sort of limit there when something suddenly changes and sometimes it affects the heading with up to 20 degrees. Don't know what to do about it though. I think there is some serious compass issues. I went outside just to see if the N S offset was a given and the E W wouldn't even line up.... That is still level, not sure if it would be off by more if at a higher elevation or not.