Image-stitching and virtual tour solutions My account Updates
It is currently Wed Oct 01, 2014 5:02 am

All times are UTC + 1 hour




Post new topic Reply to topic  [ 3 posts ] 
Author Message
 Post subject: APP Benchmarking
PostPosted: Wed Apr 30, 2008 5:19 pm 
Offline
Member
User avatar

Joined: Tue Mar 11, 2008 10:49 pm
Posts: 511
Location: Bath, UK
I have built a new machine for running APP and want to compare its speed against other machines, and its speed on different numbers of images on the same machine. I know next to nothing about benchmarking other than it is very difficult to do because of the huge number of variables involved, especially when different machines are involved. Nevertheless I intend to try because I am a geek. I would also like some reassurance that the money I spent on the new machine was not wasted, but I want to do this mainly because I am a geek.

What I thought people here might be able to tell me is to what extent does the type of image determine run-times?

For example, is it harder for APP to detect panos in say low-contrast images? Does contrast/sharpness/whatever have an effect on blending times? Or how about the field of view? I can make a 1-D pano out of say 20 images, but the FoV depends on the zoom setting, so maybe the trigonometric calculations are different.

Aeris

_________________
Robopan 3.1 - Sony DSC H7.


Top
 Profile  
 
 Post subject:
PostPosted: Thu May 01, 2008 10:08 am 
Offline
Member

Joined: Tue Dec 06, 2005 1:57 pm
Posts: 2946
Location: Grenoble
Aeriscera wrote:
What I thought people here might be able to tell me is to what extent does the type of image determine run-times?

A brief question where a complete and sensible answer is certainly not possible: too many intervening variables exist.

Elapsed time before you press the Render button and rendering time are to be set aside.

What counts before rendering:
- number of source images
- source images size (MP)
- most Detection and Optimization settings
- RAW / 16 bit TIFF / JPEG
... and much more.

What counts for rendering:
- output size (if 100% corresponds to 10 minutes then 25% often corresponds to less than 1 minute)
- color depth
- used blender
- used interpolator
- number of layers
... and much more.

A comparative study between using Smartblend and using Multiband on different machines (disk system included) for a large spectrum of pano sizes (8bit or 16bit color depth) would be, probably, the most useful and the most appraised one ...but, hopefully, would be obsolete when completed :)

_________________
Georges


Top
 Profile  
 
 Post subject:
PostPosted: Wed May 07, 2008 11:23 am 
Offline
Member
User avatar

Joined: Fri May 05, 2006 8:16 am
Posts: 1228
Location: Bulgaria
There is at least two other topics on benchmarking with quite a bit of discussion, but I cannot seem to find them now.

The best suggestion IMHO is a separate application, that runs tests with predefined settings on sample image sets, downloadable separately.

_________________
Some of my panoramas, posted in the Autopano Pro flickr group.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 3 posts ] 

All times are UTC + 1 hour


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group