Agisoft PhotoScan 1.1.0 pre-release

[Alexey Pasumansky]

Hello Mark,

Then please check that you are not using zero instead of O letter.

[ppkong]

Can I export model tiles in python script 1.1?

[r0xx]

Hello r0xx,

Could you please send the screenshots of the network settings on root and client to support@agisoft.com? Please also specify the OS you are using.

Will do right now! Thanks!

[SuperNewb]

Hi guys,

Maybe I'm missing something.  It seems that the markers are unable to be toggled on and off in this version.

Am I doing something wrong or is this purposefully shut down?

[mwillis]

My issue turned out to be a mistake on my part.  I downloaded the standard version of Photoscan by accident.  Sorry for my confusion.

[Alexey Pasumansky]

Hello SuperNewb,

Only markers with any source coordinate values can be checked/unchecked. As soon as any data is input for the marker it will be checked on automatically.

[bmc130]

Is there any speed gains in 1.1 compared to the older version? I mean faster processing mesh and generating point clouds and so on?

[SuperNewb]

Hello SuperNewb,

Only markers with any source coordinate values can be checked/unchecked. As soon as any data is input for the marker it will be checked on automatically.

Thanks Alexey,

Another question though.  I use Tile export in obj format for use in Geoverse.  It works great for the converter package.  The Pre Release does not seem to offer the same export method, leaving only the option to export photoscan tiles for the viewer package. 

Is there any hope that the final release will hold on to the functionality to export the obj tiles that I need?

Thanks again

[Alexey Pasumansky]

Hello SuperNewb,

If you open TLS file in Agisoft Viewer, you will be able to convert it into one of the supported formats (like OBJ, COLLADA, OpenCTM and etc.) by choosing Convert option in the context menu after clicking on the *.tls line in the Playlist pane.

[Alexey Pasumansky]

Hello bmc130,

There were speed improvements for Estimating Scene Structure step for large datasets.

[orpheas]

Hi,

I have updated from Photoscan Pro 1.0.4 to version 1.1.0 in order to use my GPU, as version 1.0.4 was showing the same issue as describe here: http://www.agisoft.com/forum/index.php?topic=561.0

Once I have updated, photoscan was able to use the GPU but when I closed the application and restarted it, it no longer sees the opencl device and outputs the following in the console when I go to Preferences -> OpenCL
clGetPlatformIDs failed: error -1001

What I have tried so far is:
  - uninstall / reinstall photoscan (both version 1.0.4 and 1.1.0), same results
  - uninstall and purge old drivers and reinstall latest nvidia drivers (344.48), same results
  - found this related article in the forum, http://www.agisoft.com/forum/index.php?topic=2977.msg15682 , tried to install to a different path and still get the same results.

Since this happened when I updated to version 1.1.0 I thought I'd ask here.
I have had the same behavior on two systems and I was wondering if someone would have a suggestion on how to solve this.

Thanks in advance.


Orpheas

[Alexey Pasumansky]

Hello Orpheas,

Thank you for reporting. What graphic card you are using? Probably you can provide the log file from the beggining (open project and start dense cloud generation)?

We have tested latest nVidia drivers for several graphic cards with PhotoScan 1.1 and have not got such problems.

[Helsky]

Dear Agisoft Team,

I was very pleased reading "Added support for Cross (non-coded) target detection". However, do they need to be colored or does this work for gray scale images as well. Moreover, how many pixels should a target at least be in the image?
Thank you very much

[orpheas]

Hi Alexey,

Thanks for the quick response.
I have finished diagnosing this issue this weekend and it is not related to photoscan version 1.1.0.

This behavior is due to remote access of the computer via MS Terminal Services.

Sorry for reporting this before finishing the troubleshooting.


Orpheas

[mwillis]

In the current stable version of Photoscan Pro, you can "rotate object" on model that has ground control.  In the new beta, you can not rotate objects that have ground control.

I often assign arbitrary ground control to an object to establish scale and then rotate the object so it is in the proper orientation.  Would you please consider adding "rotate object" back as on option in future versions of PhotoScan?

Thanks