Messages

1

General / Re: Agisoft PhotoScan 1.3.0 pre-release

« on: January 03, 2017, 01:17:23 AM »

Hello,
thanks for this new beta release, it seems that the key finding and matching use GPU now, the gain in time is considerable, but I noticed some differences in the result quality comparing to the pure CPU method.
Point color seems to be not correct and it is very difficult (sometime impossible) to have all the images matched with GPU when there is no problem with CPU, so the new method seems to be more demanding about shooting quality.
If this issue is not solvable, you could consider to give the possibility to use GPU only in dense reconstruction in case of difficulty with GPU matching.
Thanks.
Pascal
   

2

Bug Reports / Re: problem with paths

« on: October 08, 2015, 05:01:09 PM »

Alexey,
thank you, I am very sorry, I think there is a misunderstanding, we have only one subset in photoscan pointing in several folders in windows, we merged the folder in windows, so we get wrong path in photoscan for some images.
Selecting all the camera in photoscan and change the path doesn't not apply the correct path for all the photos, some path remains unchanged that it is not normal I think (selecting the unchanged camera individually and changing the path works but it is very long). Excuse my English, and thank you for your patience, but we still think that it is a problem in photoscan.
Thanks.
Pascal

3

Bug Reports / Re: problem with paths

« on: October 08, 2015, 03:56:27 PM »

thank you CC for this tip, Alexey, I do not understand your proposal, If I am not mistaken, putting the camera in a group doesn't offer the possibility of changing the path for a group, and we have again to manually select the photos to group them ,the problem remains the same I think. The images was already aligned (several days of process) so we would like not to redo the entire process.
This is not an isolate case, as we have often to complete dense clouds (whole or alignment problems) and that we want to keep our new images in separate folders we use several path but at the end we reassemble all the photos together.
Sometime the new images are easily selectable because their number are different but in this case they are perfectly mixed.
May be you can do something to simplify our live here.
Thank you.
Pascal
   

4

Bug Reports / Re: problem with paths

« on: October 08, 2015, 12:47:30 PM »

Hi Alexey,
all my images have the same extension and now are in the same location (as some has been moved they appear with a wrong location), the problem appears when we select both images with good and wrong path (in fact all the photos to go faster) and when we try to affect the correct path to all the selection.
Thanks.
Pascal

5

Bug Reports / problem with paths

« on: October 08, 2015, 12:16:57 PM »

Hello,
I encountered a problem with image path, if for some reasons some images of a project have wrong path, affecting the correct path to all the images doesn't not work. The wrong path remains even if you have affected the correct path to all the camera and you have to affect individually the path ... this work but can be totally boring when you have hundred of photos.
I think it is a problem with photoscan, but if you have a temporary solution I am interested.
Thanks.
Pascal

6

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: October 02, 2015, 05:18:46 PM »

Hello,
I encountered a problem with the last beta release, photoscan freezes (not responding during about thirty seconds) every minute regardless the action I do.
May be this is related with the lag observed (and reported ) before, but now it is really impossible to work with such issue.
All my pc is up to date (windows seven recently updated, graphic drivers too). The projects I tested are not big and during the freezes, the processor is not used. Do you have an idea?
Thanks.
Pascal

PS: after a quick reprocessing of partial set of photos no freeze was observed for the moment, does this mean that all the old calculation are to be reprocessed if some additional work is needed? I will try to re-launch a problematic chunk with the last Photoscan version, and I will let you know ....

7

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 18, 2015, 11:28:13 AM »

Hello Alexey,
the 'lag' I observed in the model view seems not to be dur to the mesh, it appears also with
a sparse cloud, but not immediately after opening a project just some minutes after ...
I restarted the pc and there is no other heavy programs running.
It appends also on light projects (less than 40 images),it appends also if I move the photoscan's windows (stop moving during 1 sec and move again, this do not append with windows from other programs)
May be do you have an idea?
Thanks.
Pascal

8

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 15, 2015, 11:56:29 AM »

Hello Alexey,
I tested the last Photoscan release and I experienced some lag (the display freeze during one second every two seconds approximatively)  in model and photos display.
This appends only after having opened a mesh model, before doing this all was ok.
This is a little bit annoying in big project when you have to navigate between photos or to do selection in cloud model.
May be you can do something to fix it.
Thanks.
Pascal

9

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 08, 2015, 03:27:18 PM »

Hi Alexey,
ooopps, sorry, thank you for your help.
Regards.
Pascal

10

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 08, 2015, 12:54:53 PM »

here is the screenshot of the windows I get (when I launch the shortcut in a shell).
thanks.
Pascal

11

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 08, 2015, 11:57:09 AM »

Thank you for this quick reply, but the problem appear when I launch the server (using an appropriate shortcut) not a job.
As the server cannot be launched, the nodes cannot connect to it and thus I can't launch a job or connect the monitor ....  I think that saving the project in the new format cannot help very much.
thanks.
Pascal
 

12

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: September 08, 2015, 11:02:43 AM »

Hello Alexey,
I cannot start a server in the last beta release,
I use the method described in the help files, but
the server window are immediately closed with an 'unrecognised parameter' error message.
may be do you modified the network processing ...
Think you.
Regards.
Pascal
 

13

General / Re: Clipped depth maps ....

« on: March 18, 2015, 05:56:33 PM »

Thank you Alexey,
I send you a private message with some photos for testing.
Regards;,
Pascal

14

General / Clipped depth maps ....

« on: March 11, 2015, 09:38:13 PM »

Hello,
I have a problem with the exported depth and normal maps, they are totally black (depth maps) or grey (normal map) or clipped (only a part of the map is present), this appears in relatively simple areas. In the joined files it is a wall in a room that is partially clipped (there is a flat wall in front of the camera in the clipped area).
May be an idea ? The dense cloud is reconstructed but I have some artefacts, may be there is e link with these clipped maps....
Thanks.
Pascal

15

General / Re: Ripple / staircase artifacts on 3D reconstruction

« on: February 05, 2015, 06:28:55 PM »

Hello,
I continue this old topic, because I am not sure it is really closed ...
As discussed before, ripples in dense clouds may result from stereo base but may be not only.
In the attached file you can see two application cases, a complex and a very simple.
In the second one, the stereo base is about 20cm for a subject distance of about 1m,
and very similar to the others cameras.
So a too small stereo base seems not to be the correct explanation.
This issue is very critical in some application, notably when you want to use Photoscan to digitize
very fine engraving on stone for example (in archaeological research) because the risk of having false information is very important.
Can this problem come from the bit depth of the depth map? filtering or not filtering these maps do not change the result.
I am curious from the real explanation of this issue and I hope there is a solution in a near future. Others photogrammetric solutions based on PMVS do not produce such artefact, so a solution certainly exists.
Thank. Pascal