mis-colored lines across stitched image  

This forum is dedicated to Autopano Pro / Giga bug reports and features discussion.
Please read the posting rules before starting a topic!
no avatar
wjh31
Member
 
Topic author
Posts: 455
Likes: 0 post
Liked in: 0 post
Joined: Thu Apr 09, 2009 11:23 pm
Location: Surrey, UK
Info

mis-colored lines across stitched image

by wjh31 » Thu Jun 14, 2012 7:40 pm

On a large panorama im trying to stitch, Im having a little problem with stability. As you can see in the attatched image, although the image initially appears to have stitched well, the final psb file, when opened to edit had a number of corrupted horizontal lines of pixels across it in various places. Has anyone come across this before and/or have any suggested solutions?

APG 2.6.3, Win7Pro64Bit



no avatar
leifs
Member
 
Posts: 648
Likes: 1 post
Liked in: 0 post
Joined: Sun Sep 06, 2009 12:08 pm
Location: Ørsta Norway
Info

by leifs » Thu Jun 14, 2012 8:58 pm

I have seen something like it. See the images below.
It was a pain in the *** and I had to do a lot of thinking before I found the explanation. First (ofcourse) I thought it was APG. But before throwing accusations against Kolor I explored other options.

It was disk error !
I have a RAID of 4 SSD's on a LSI controller and there was a hardware error. After a low level format the problem disappeared.

I don't know if this is the source of your problem, but you should check it out.

leifs




Olympus OM-D E-M1, Leica 25mm f1.4, fisheye, zooms, primes
Seitz VRdrive2
2x Intel Xeon E5-2660, 128GB RAM, Win7 64bit, SSD RAIDs

no avatar
wjh31
Member
 
Topic author
Posts: 455
Likes: 0 post
Liked in: 0 post
Joined: Thu Apr 09, 2009 11:23 pm
Location: Surrey, UK
Info

by wjh31 » Thu Jun 14, 2012 9:16 pm

I did think it was probably a hardware fault but thought id check here first as there's alot of hardware to check. That particular render performed read and/or write on every disk. Ive started a new render missing out one of the disks and work from there to find the fault.

Interesting that formatting was sufficient to remove the error

no avatar
wjh31
Member
 
Topic author
Posts: 455
Likes: 0 post
Liked in: 0 post
Joined: Thu Apr 09, 2009 11:23 pm
Location: Surrey, UK
Info

by wjh31 » Sat Jun 16, 2012 5:39 pm

Ive run memcheck and removed one bad stick of ram, Ive also formatted my SSD scratch disk and run the render several times avoiding each possible disk and im still getting issues. Ive got a 4Gigapixel image which has stitched ok but the 20Gigapixel image is still comming out plastered in corruption, any other ideas?

no avatar
wjh31
Member
 
Topic author
Posts: 455
Likes: 0 post
Liked in: 0 post
Joined: Thu Apr 09, 2009 11:23 pm
Location: Surrey, UK
Info

by wjh31 » Sat Jun 16, 2012 11:07 pm

small update. I rendered the 20 gigapixel image at 1 gigapixel to see if the corruption was still present, and it came out clean, so i rendered at 85% giving 15 gigapixels and that too came out clean so i will use that for now. However such a result does throw the source of error back into question

User avatar
AlexandreJ
Kolor Team
 
Posts: 5987
Likes: 7 posts
Liked in: 10 posts
Joined: Mon Nov 14, 2005 4:56 pm
Location: Francin, France
Info

by AlexandreJ » Wed Jun 20, 2012 8:39 am

We did have only a few time this kind of report and every time, it was a hardware issue. THe fact that you don't have it with smaller version does work with this hypothesis : you just didn't use as much cache as with the full version and so didn't reach the hardware sector that has issue.

no avatar
wjh31
Member
 
Topic author
Posts: 455
Likes: 0 post
Liked in: 0 post
Joined: Thu Apr 09, 2009 11:23 pm
Location: Surrey, UK
Info

by wjh31 » Sat Jun 23, 2012 9:57 am

Ill have another go at memcheck then. Can anyone reccomend a way of finding disc errors?


Who is online

Users browsing this forum: No registered users and 1 guest