thanks. i'll give this a try at higher resolution, but possibly not
for a week or so.
On Jun 17, 12:27 pm, Bart van Andel wrote:
> I just got a perfect result from your image set. What I did was the
> following:
>
> 1. Have Hugin generate the remapped images, let's assume those are
> located in
I just got a perfect result from your image set. What I did was the
following:
1. Have Hugin generate the remapped images, let's assume those are
located in folder1;
2. Find the images containing a full ring (3 of them in your image
set) and move them to another folder, folder2;
3. For each of th
finally converted the project to jpg. i had to split the images into
two archives do you file size limits on yousendit.
they can be downloaded here:
part 1: https://download.yousendit.com/UmNKcHBPUzdFc0pMWEE9PQ
part 2: https://download.yousendit.com/UmNKcHBDVnN6NExIRGc9PQ
On Jun 11, 6:31 pm,
i tried opening the ring images, no change in results of the final
pano, though.
regarding ptmasker and ptroller, do i reprocess the entire project
(i.e., change soem settings in hugin to use ptmasker and ptroller) or
only the ring images or ...??? i took a quick look at the man pages
and its no
ok, i'll try to remove a small segment of the mask and re-run enblend
on them.
regarding ptmasker and ptroller, i've never used either of these but
i'll see if i can figure it out.
thanks
On Jun 11, 3:09 am, Bart van Andel wrote:
> > when you mention opening up the ring masks above, are you re
> when you mention opening up the ring masks above, are you referring to
> increasing the inside diameter of the ring, thus having a thinner
> sliver which will be covered by another image?
By "opening up" I mean creating a gap in the ring, so that it's no
longer a ring, which will mean that othe
I think it will an interesting exercise to run them through PTmasker
and PTroller to see if the mask computation works better. This might
help create a patch for enblend.
On Thu, Jun 11, 2009 at 7:54 AM,
slaterson wrote:
>
> i took a look at all three ring images in this project as well as a
> f
i took a look at all three ring images in this project as well as a
few that are 'missing' (in the black areas) in the problematic pano.
the masks appear to be pixel perfect in line with the remapped image.
when you mention opening up the ring masks above, are you referring to
increasing the insid
this certainly does help, i'll try to have a look today.
once i get the project converted to jpg i'll send a link, also...
thanks!
On Jun 9, 2:00 pm, Bart van Andel wrote:
> Converting from a TIFF project to JPG is pretty easy. Just batch
> convert all the tiffs to jpg (only change the file ex
Converting from a TIFF project to JPG is pretty easy. Just batch
convert all the tiffs to jpg (only change the file extension), open
the hugin .pto file with your favourite text editor and replace all
".tiff" occurences with ".jpg". No need to recompute any keypoints or
realign the images.
An alp
i can work on making a jpg project out of this. due to work and some
upcoming business travel & vacation i may not be able to get to it for
a couple weeks, though.
regarding checking the alpha channels/masks, i'd like to check them
but need a little help in how to do this. can you offer a littl
Of course this is a solution (well, more of a workaround actually),
but a) it requires an extra rendering step which shouldn't be
necessary and b) the more reprojection, the more quality loss because
of interpolation and extrapolation. The latter can be circumvented
partly by rendering the equirec
There was a thread on this topic entitled "Hugin/enblend error?" dated
Feb 10th 2008. The simplest solution is to create a single
equirectangular image in hugin and then remap that (in hugin).
Peter.
From: Bart van Andel
Subject: [hugin-ptx] Re: lambert equal area az projection
Currently I don't have time for extensive testing, but if you were to
upload your project somewhere, I think for testing jpg offers enough
quality to find out what's going wrong. For a 1:20 compression (pretty
typical for jpg, 100% quality setting not really necessary) you end up
with only about 1
anything else i can do to try to trouble shoot? i have tried cropping
the 'ring' images in hugin, but that doesn't help at all. i am happy
to upload my images and project file somewhere, however the entire set
will be approx 2gb.
On Jun 8, 4:33 pm, dmg wrote:
> Thanks! This is great help isol
Thanks! This is great help isolating the problem. It sounds like it is
an enblend issue
with some features of these images.
--dmg
On Tue, Jun 9, 2009 at 4:42 AM, slaterson wrote:
>
> removing all three 'ring' images gets rid of the black areas.
> however, the final pano has a small area at the t
removing all three 'ring' images gets rid of the black areas.
however, the final pano has a small area at the top and bottom that
are missing. also, there are a couple areas around the outer
circumference of the final pano where the seams are showing up.
On Jun 8, 10:46 am, slaterson wrote:
>
i ran another test this morning. all the remapped images open in the
gimp and display properly.
bart: yes, there are three images just as you describe below. they
are 'rings', a 'sliver' of image around the outside of the circle with
nothing in the center. i just started another run, excluding
Are one or more of the remapped images fully contained within another
remapped image perhaps (e.g. the containing remapped images forms a
circle with a hole in the middle)? If masks aren't taken in
consideration properly, Enblend might decide to skip it. By cropping a
part of the output image (and
on previous runs i have looked at the remapped images (nautilus
preview) while enblend was running (launched from hugin, of course)
and they all look fine. i will run it again tomorrow (monday), save
the remapped images and give them a closer look, though.
On Jun 7, 10:04 pm, dmg wrote:
> for
for the time being just output separate images and look at them.
If you like them then run enblend -h and it it will be self-descriptive.
--dmg
On Mon, Jun 8, 2009 at 12:46 PM,
slaterson wrote:
>
> i haven't tried that. happy to try it, but need a little help in
> exactly how i should run enbl
i haven't tried that. happy to try it, but need a little help in
exactly how i should run enblend, images to include, command line
options, etc. is it possible to have hugin save the enblend command
it is running so i can run it manually later?
thanks
On Jun 7, 8:39 pm, dmg wrote:
> Have you
Have you tried outputing independent images and then feeding them
manually to enblend?
Your problem looks like a masking problem.
-dmg
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"hugin and other free panoramic
i have uploaded some images, check here:
http://www.flickr.com/photos/24148...@n05/3605374851/
i have made an equirectangular pano from these and then tried lambert
eaa projection on it, hugin spits out exactly what i am looking for.
all three images are at the link above (cropped lambert - no d
On Mon, Jun 8, 2009 at 8:06 AM, slaterson wrote:
> i will post the output image and send a link in a little bit. need to
> reboot before i can get to the image.
>
> thanks
Have you tried to map the images first to a equirectangular, and then
to the Lambert?
This will allow us to know if it is a
i will post the output image and send a link in a little bit. need to
reboot before i can get to the image.
thanks
On Jun 7, 4:02 pm, Bruno Postle wrote:
> On Sun 07-Jun-2009 at 15:56 -0700, slaterson wrote:
>
>
>
> >where is the best place to upload the images? there are a total of
> >61, ea
On Sun 07-Jun-2009 at 15:56 -0700, slaterson wrote:
>
>where is the best place to upload the images? there are a total of
>61, each is 36 megs (tifs). happy to upload them.
I was thinking that an output image showing the result of the
problem would probably be enough to start with.
--
Bruno
where is the best place to upload the images? there are a total of
61, each is 36 megs (tifs). happy to upload them.
On Jun 7, 3:42 pm, Bruno Postle wrote:
> On Sun 07-Jun-2009 at 13:31 -0700, slaterson wrote:
>
>
>
> >i recently discovered the lambert equal area azimuthal project, it
> >prod
On Sun 07-Jun-2009 at 13:31 -0700, slaterson wrote:
>
>i recently discovered the lambert equal area azimuthal project, it
>produces a very pleasing image. :) i am having a problem with it
>though. if i chop off the top and bottom of the final image (by
>changing the height of the pano) the pano
29 matches
Mail list logo
