1
General / Re: TEST TEXTURE 1.0
« on: January 20, 2014, 12:37:40 PM »
maybe so 'looks best
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: [1]
2
General / Re: TEST TEXTURE 1.0
« on: January 20, 2014, 12:27:49 PM »
with version 0.9 were seen only video with version 1.0 you can also see on the printed pattern.
3
General / Re: TEST TEXTURE 1.0
« on: January 20, 2014, 11:32:42 AM »
hi Wishgranter
thanks for the information.
I downloaded the software and effectively with the format. fbx can
upload files with 8k texture (I have a ZPrint 450).
This afternoon I try to print a file processed with version 1.0 format .fbx.
I wanted to know if you're experiencing problems with scratches in the print version 1.0
thanks roby
thanks for the information.
I downloaded the software and effectively with the format. fbx can
upload files with 8k texture (I have a ZPrint 450).
This afternoon I try to print a file processed with version 1.0 format .fbx.
I wanted to know if you're experiencing problems with scratches in the print version 1.0
thanks roby
4
General / TEST TEXTURE 1.0
« on: January 17, 2014, 08:47:07 PM »
Dear all good morning,
I use AGISOFT like BODYSCANNER to make miniature printing with a colour Zprinter (like the most of you).
Switching from version 0.9 to 1.00 i found easyest to make MESH (more volume precision) but big problems to working the TEXTURE.
I saw on the forum that this problem was already discussed but i desire to share with you the results of some of my tests to understand if the workflow is correct.
With the old 0.9 version the workflow was:
photo 100
1 align photo - hight - generic
2 build geometry medium - 1000000 face
3 build texture generic -mosaic- 4096
The file obtained was exported in format: OBJ and TEXTURE.TIF
Importing the file in ZBRUSH it was smooth and the textures were optimized using PHOTOSHOP.
From ZBRUSH we exported the file in format: WRLM
On ZPRINT we imported the format WRLM loading the modified TIF and in this way the printing were good.
Using the latest version of Agisoft we did some tests. You can see the resultas listed below in order from the first one we made (the one with big problems for textures ) to the last (the one that is more near to the results obtained with the previous version of Agisoft ) .
1° TEST
N° PHOTO 210
1 align photo - hight - generic
2 DENS CLOUD medium
3 build geometry medium
4 build texture generic -mosaic- 4096
result vertices 180456 -face 375289 - point 341111- dense cloud point 16888089
TEXTURE BLACK !!
2° TEST
N° PHOTO 210
1 align photo - hight - generic
2 DENS CLOUD LOW
3 build geometry medium
4 build texture generic -mosaic- 4096
result vertices 143746 -face 288372 - point 341111- dense cloud point 4297264
TEXTURE acceptable but very fragmented
Following the discussions on the forum, we realized that the texture problem is perhaps linked to the fact of having a model with a lot of mesh vertices and then from the first to the last test we tried to gradually reduce the mesh trying to maintain a quality on the volumes of the model acceptable .
We verified that the quality of the texture improves by switching from 4096 to 8192 but unfortunately the printer reads a maximum of 4096 .
We still need to fix the fact that with the texture of the latest version of the prints semms to be streaked , but we hope to solve the problem soon.
To conclude I would say that this latest version (apart from the texture) greatly improves the final results both in quality ' of volume and processing time.
With the inclusion of dense clou we are no longer forced to divide the cunk into 2 thing (before with 200 photos and 16 GB we were forced to do).
I Take this opportunity to thank everyone for the advice taken from this forum and especially to Lee for the many suggestions and last but not least the video posted on tripleangares .
Any consideration or advice will be wellcome
SORRY FOR THE BAD ENGLISH
Greatings Roby
I use AGISOFT like BODYSCANNER to make miniature printing with a colour Zprinter (like the most of you).
Switching from version 0.9 to 1.00 i found easyest to make MESH (more volume precision) but big problems to working the TEXTURE.
I saw on the forum that this problem was already discussed but i desire to share with you the results of some of my tests to understand if the workflow is correct.
With the old 0.9 version the workflow was:
photo 100
1 align photo - hight - generic
2 build geometry medium - 1000000 face
3 build texture generic -mosaic- 4096
The file obtained was exported in format: OBJ and TEXTURE.TIF
Importing the file in ZBRUSH it was smooth and the textures were optimized using PHOTOSHOP.
From ZBRUSH we exported the file in format: WRLM
On ZPRINT we imported the format WRLM loading the modified TIF and in this way the printing were good.
Using the latest version of Agisoft we did some tests. You can see the resultas listed below in order from the first one we made (the one with big problems for textures ) to the last (the one that is more near to the results obtained with the previous version of Agisoft ) .
1° TEST
N° PHOTO 210
1 align photo - hight - generic
2 DENS CLOUD medium
3 build geometry medium
4 build texture generic -mosaic- 4096
result vertices 180456 -face 375289 - point 341111- dense cloud point 16888089
TEXTURE BLACK !!
2° TEST
N° PHOTO 210
1 align photo - hight - generic
2 DENS CLOUD LOW
3 build geometry medium
4 build texture generic -mosaic- 4096
result vertices 143746 -face 288372 - point 341111- dense cloud point 4297264
TEXTURE acceptable but very fragmented
Following the discussions on the forum, we realized that the texture problem is perhaps linked to the fact of having a model with a lot of mesh vertices and then from the first to the last test we tried to gradually reduce the mesh trying to maintain a quality on the volumes of the model acceptable .
We verified that the quality of the texture improves by switching from 4096 to 8192 but unfortunately the printer reads a maximum of 4096 .
We still need to fix the fact that with the texture of the latest version of the prints semms to be streaked , but we hope to solve the problem soon.
To conclude I would say that this latest version (apart from the texture) greatly improves the final results both in quality ' of volume and processing time.
With the inclusion of dense clou we are no longer forced to divide the cunk into 2 thing (before with 200 photos and 16 GB we were forced to do).
I Take this opportunity to thank everyone for the advice taken from this forum and especially to Lee for the many suggestions and last but not least the video posted on tripleangares .
Any consideration or advice will be wellcome
SORRY FOR THE BAD ENGLISH
Greatings Roby
5
Face and Body Scanning / export file
« on: July 31, 2013, 04:51:11 PM »
Hello,
while using agisoft standard as bodyscanner I obtain a very small wrlm size of the model when exporting (person 1800 mm height - model 20 mm).
Is there a way to obtain a better size of the model before exporting or should I use wrong setups?
Thank you.
Roby
while using agisoft standard as bodyscanner I obtain a very small wrlm size of the model when exporting (person 1800 mm height - model 20 mm).
Is there a way to obtain a better size of the model before exporting or should I use wrong setups?
Thank you.
Roby
6
Face and Body Scanning / Re: Samsung NX1000
« on: June 10, 2013, 12:40:40 PM »
Dear Mala,
many thanks for your kind reply and precious informations.
Unfortunately i'm a beginner and could be possible i will disturb you again in the future.
Below you can find the links .
As you can see from the Twenga website the prices changes week by week.
From the graphic (up on your right) you can see that some days ago the price was 200€ (2013/06/05).
I created a price allert inserting my email. Every time that the price go down an e-mail will be sent to you.
Today the cheapest price is 239€
I ask you if you have some suggests to give me about the Agisoft set up for the scansion of the human figures.
Best regards
roby
http://www.twenga.it/samsung-nx1000.html
http://www.resetdigitale.it/it/samsung/30306-fotocamera-samsung-nx1000-nx-1000-nx-1000-kit-20-50mm-20-50-nero-black-8806085044227.html
many thanks for your kind reply and precious informations.
Unfortunately i'm a beginner and could be possible i will disturb you again in the future.
Below you can find the links .
As you can see from the Twenga website the prices changes week by week.
From the graphic (up on your right) you can see that some days ago the price was 200€ (2013/06/05).
I created a price allert inserting my email. Every time that the price go down an e-mail will be sent to you.
Today the cheapest price is 239€
I ask you if you have some suggests to give me about the Agisoft set up for the scansion of the human figures.
Best regards
roby
http://www.twenga.it/samsung-nx1000.html
http://www.resetdigitale.it/it/samsung/30306-fotocamera-samsung-nx1000-nx-1000-nx-1000-kit-20-50mm-20-50-nero-black-8806085044227.html
7
Face and Body Scanning / Re: Samsung NX1000
« on: June 09, 2013, 06:00:37 PM »
Hi everybody!
I finally managed to make some proofs with my nikon nx1000.
The three proofs have been made with one camera and one rounding table.
In picture 1 I used 2 500watt flashes with 60x90 soft box.(point 24593)
In picture 2 I used 2 500watt flashes with umbrellas pointed directly on the object.(point 27011)
In picture 3 I used 2 500watt flashes with umbrellas directed on the contrary using reflected light.(point 47970)
I think that third solution is the one giving best results. The problem is being able to synchronize more cameras at the same time.
I contacted Samsung Italy and one technician told me that no synchronization is foreseen for this kind of devices.
Has anybody experienced this kind of thing?
I found some Nikon nx1000 at Euros 200 on the internet therefore it could be interesting to find the solution to use more cameras at the same time.
I recently bought the Agisoft standard
[/img][/img]licence and would like to know from those who are expert in this field which are the best set ups on Agisoft for this kind of work.
1) in preferences which maximum point number we have to set and with which model generation (medium or aggressive).
2) in build geometry which are the best parameters to obtain a good result.
Any suggestion on the use of Agisoft is highly appreciated.
I will soon publish further results.
Bye,
Roby
I finally managed to make some proofs with my nikon nx1000.
The three proofs have been made with one camera and one rounding table.
In picture 1 I used 2 500watt flashes with 60x90 soft box.(point 24593)
In picture 2 I used 2 500watt flashes with umbrellas pointed directly on the object.(point 27011)
In picture 3 I used 2 500watt flashes with umbrellas directed on the contrary using reflected light.(point 47970)
I think that third solution is the one giving best results. The problem is being able to synchronize more cameras at the same time.
I contacted Samsung Italy and one technician told me that no synchronization is foreseen for this kind of devices.
Has anybody experienced this kind of thing?
I found some Nikon nx1000 at Euros 200 on the internet therefore it could be interesting to find the solution to use more cameras at the same time.
I recently bought the Agisoft standard
1) in preferences which maximum point number we have to set and with which model generation (medium or aggressive).
2) in build geometry which are the best parameters to obtain a good result.
Any suggestion on the use of Agisoft is highly appreciated.
I will soon publish further results.
Bye,
Roby
8
Face and Body Scanning / Re: Samsung NX1000
« on: April 12, 2013, 12:33:17 PM »
Dear all,
I'm writting from Italy. I desire to thank you for the good suggests you gave me.
I also tested many different systems (at the moment i'm using Artec with Kinect).
Some time ago i made some tests with Agi and Canon Reflex with good results.
I read the post written by Coadey; i bought a Samsung NX 1000 camera and i built a rotating platform and tomorrow i will make some tests.
I desire to know by Coadey the correct distance between the camera and the subject.
I desire to know this to understand better the space that i will need for doing my test (in this way i will avoid problem of space).
As soon as possible i will post my results.
Best regards and thanks for all.
I'm writting from Italy. I desire to thank you for the good suggests you gave me.
I also tested many different systems (at the moment i'm using Artec with Kinect).
Some time ago i made some tests with Agi and Canon Reflex with good results.
I read the post written by Coadey; i bought a Samsung NX 1000 camera and i built a rotating platform and tomorrow i will make some tests.
I desire to know by Coadey the correct distance between the camera and the subject.
I desire to know this to understand better the space that i will need for doing my test (in this way i will avoid problem of space).
As soon as possible i will post my results.
Best regards and thanks for all.
9
Face and Body Scanning / Re: Body Scanner Kickstarter
« on: March 24, 2013, 11:07:02 PM »
thanks for information and see you soon.
roby
roby
10
Face and Body Scanning / Re: Body Scanner Kickstarter
« on: March 24, 2013, 10:59:38 AM »
Dear lee please accept my compliments for your results. I am also trying to do some shots in Turin. Would you please let me know which dslr you use and if you block the rounding table at every photo shot or if you leave it moving. Thanks by roby 
Pages: [1]
