Very interesting idea. Perhaps worth mentioning there is an excellent tool
for stabilizing video itself.
DaVinci Resolve is a gold standard and has a free option.
https://motionarray.com/learn/davinci-resolve/davinci-resolve-stabilize/
Another tool from a few years back but still functionality is
Thanks for the suggestions:
1) No luck on --primary-seam-generator=nearest-feature-transform. Did not
change anything.
2) No luck with or without --pre-assemble.
3) Always wanted to try multiblend. Never found adequate instructions for
running in windows. Tried and failed several times over
On very large renders (>15K pixels/side), I often see hugin "skipping" an
image or two and leaving a blank space.
Example at https://photos.app.goo.gl/sEqkSeJbqYSGBGTb8
Rendering at lower resolution fixes the problem.
At full resolution, by trial and error I re-arrange the order of the images
The problem is the line "echo @echo off > $$$_enfuse_temp_$$$.bat".
I suppose windows won't let the script create a .bat file.
The enfuse_additional_parameters are incorrect, but trivial to correct.
I AM able to make enfuse_align_droplet.bat work by dropping a collection of
files.
This isn't my s
to date.
I am surprised not to find any further discussion about this tool in the
newsgroup, as it is exactly what is needed to save huge amounts of time in
some workflows.
Is there a newer way to do things that I have missed?
TIA,
Rich MacDonald
--
A list of frequently asked questions is av
>If I understand your process correctly, you are using hugin to create and
save a set of exposure fused images which you then load into hugin as new
project, and stitch.
Yes.
>Does hugin automatically detect the fov of each of the exposure fused
images ...
It fails to detect the fovs,
I do a lot of panoramas with multi-row and multi-exposure. I usually put
all the images into a single file with good results, but when the enblend
"excessive overlap" hits, I cannot figure out how to recover.
Recently I switched to a more manual approach: I first build the stacks
individually a
Afaik, the only way to determine if you have "unconnected image groups" is
to open the fast preview->assistant tab and read the text message. However,
for large projects, that text message goes offscreen, so I can never find
which items need to be connected.
Is there another places to see this?
I've loved hugin and used it for a few years. This is the first time I've
had a real problem. But I create a stitcher job and it goes to the batch
processor properly, but then the log shows that it fails in Nona. Looks
like right after the OpenGL GPU code is created...See the log
--
A list of
rename it
to enfuse_openmp.exe
Good luck. I actually have to do this myself today, so I'll report back
again if it screws up.
On Monday, July 1, 2013 8:30:14 AM UTC-5, Rich MacDonald wrote:
>
> You are probably running enfuse with the openmap option turned on.
>
>
> http:
You are probably running enfuse with the openmap option turned on.
http://wiki.panotools.org/Hugin_FAQ#Enblend-Enfuse_OpenMP_SSE_GPU:_which_one_is_the_right_one_for_me.3F
Try switching to enfuse with the "image cache" option instead (incompatible
with openmap, so it has to be one or the other).
Having just run into the dreaded "Mask is entirely black, but white image
was not identified as redundant" fatal error on the 4th processing day of
large job, I did a search for ways to restart the build from an
intermediate point. I found the articles in this forum explaining how to
use the ma
A very useful feature for me would be to add a checkbox in the Optimizer
tab that tells hugin to save the project file between every iteration. This
would help with the very-large projects where the app has pegged the CPU,
taken up all the RAM, and become unresponsive. For example, I have a
pro
I found this post looking for the same feature :-) Agreed it would be very
useful to automatically "fit the gray" pairs.
--
You received this message because you are subscribed to the Google Groups
"Hugin and other free panoramic software" group.
A list of frequently asked questions is availabl
se get them to the same place?
>
> Yuv (trying to make sense of this part of the snapshot release process
> that I was not aware of)
--
Rich
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"hugin and o
On 2009.07.21. 15:46, Bruno Postle wrote:
> On Tue 21-Jul-2009 at 15:30 +0300, Rich wrote:
>> what about 2009.07.dev- ?
>> that would make it easier for casual users to figure out what's what
>
> Sorry this stuff is actually used by Linux packaging systems, please
>
27;m quite sure that we will never
> release 4075 fixes to the current stable so it wouldn't interfere with
> stable.
while that might confuse a user here or there, i believe that't better
than .00.0 syntax.
what about 2009.07.dev- ?
that would make it easier for casual u
On 2009.07.20. 20:06, Rich wrote:
> On 2009.07.20. 19:33, Bruno Postle wrote:
>> On Mon 20-Jul-2009 at 00:17 -0400, Yuval Levy wrote:
>>> bloat? I strongly disagree. a branch or a tag is very cheap and it is,
>>> in such a case, a bookmark in history.
>> Actually
hile others work on the shiny new features.
i agree - trunk freezing strategy mostly helps to motivate developers
for the release, otherwise it becomes a neverending trunk development.
--
Rich
--~--~-~--~~~---~--~~
You received this message because you are subscribed
enerated files when the matching phase fails (because of
> memory issues, etc.).
i second this usage. .key files have also been useful to me in case i
want to stop the process - then i can resume it later.
> Of course, if there are no memory problems, the need for those files
> disappear
ed
but not yet assigned to a dev bugs.
people could subscribe to any of such lists and get mails of new,
untriaged bugs or new, confirmed bugs.
> comments welcome. the page on the wiki is editable, feel free to improve
> or as questions.
>
> Yuv
--
Rich
--~--~-~--~~-
se candidate, i.e. The final release may be identical.
>
> Changes since 0.8.0 rc3:
>
> * Minor Hungarian translation update.
>
> * Fixes for Ev value bug that manifests as 'white' images.
HOORAY !
thanks :)
this was
he H-Fov corresponding to
> the long side of the image, but now that it's turned the H-fov is the
> short side of the image. )
--
Rich
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"hugin and other
lue set to some ridiculous value.
simply resetting it manually in "camera and lens" -> "photometric"
results in a usable pano for me.
--
Rich
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
&q
On 2009.03.06. 14:15, paul womack wrote:
> Rich wrote:
>> On 2009.03.06. 12:40, Alexandre Duret-Lutz wrote:
>>> On Wed, Mar 4, 2009 at 10:50 AM, Bruno Postle wrote:
>>>> It's nice to see some hugin pictures in the 'mainstream media',
>>>&g
y.
i'd suggest (politely, of course ;) ) emailing & explaining that you
feel it was an important information lost.
unfortunately, most journalists tend to be extremely ignorant about
topics they write - maybe you can get them to showcase another
w
On 2009.02.05. 03:11, Bruno Postle wrote:
...
> On Sun 25-Jan-2009 at 11:07 +0200, Rich wrote:
>> what about hugin setting EV to some ridiculously large value for opened
>> photos ?
>> for me, opening a few images exposure value is set to 13.something -
>> took me s
t;important, scale up" and low pixel values
> mean "unimportant, scale down".
>
> So I could paint the mask file and apply "the magic"
>
> Does anyone know where I get "the magic" ?
>
>BugBear
--
Rich
--~--~-~--~~---
ey can be considered the only thing most users will use.
faster release cycle pushes new features to users faster, generates
interest etc.
while i personally started using trunk soon after discovering hugin, i
wouldn't recommend that to most persons i know - but then they miss the
neat fea
rge value for opened
photos ?
for me, opening a few images exposure value is set to 13.something -
took me some time to find out why my resulting pano is all-white. i'd
expect a novice user to be quite annoyed about that...
...
--
Rich
--~--~-~--~~~---~--
world, but specific enough so that clear requirements for
submitted works are set.
given that this was suggested for hugin, i'd suggest positioning this as
so that whole process has to be done using opensource tools (well,
except camera firmware maybe... ;) ).
though sometimes (especially with larger
panos) using cropped images can speed up things a lot.
if you happen to be in such a situation, you can save offset from a
remapped image, edit it with gimp, then restore offset - i think
tiffinfo and tiffset comm
ed to support Windows.
> Remember the users' outcry when I questioned whether PowerPC Mac should
> be maintained? PPC users stepped up to the task and helped themselves.
which would be awesome to happen in this case as well :)
...
> Yuv
--
Rich
--~--~-~--~~
mage ?
>
>
> At the risk of sounding old-fashioned, can the command be run
> under a debugger, or the core dump analysed after the fact, to give
> a stack trace of the error?
>
>BugBear
--
Rich
--~--~-~--~~~---~--~~
You received
nt vendors.
personally, i would have no objections for small, unobtrusive ads in
these two locations (i would compare that to sun logo in oo.org
splashscreen).
--
Rich
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Group
s for the mentioned reasoning that 'later' might give impression that
particular bugs might not be fixed for a long time, that would (imho)
still be better than assigning a version, then changing it to a later
one (this _will_ happen), as in this case you are not meeting the
expectati
s and based on that
> comment on some more.
it probably would be hard to implement technically, but having all
related tools (nona, enblend etc) able to use distcc... that would be
kinda awesome and allow reuse of existing distcc infrastructure =)
...
--
Rich
--~--~-~--~~--
IFT an 84x8mp shot, that just errored out with "could
> not run".
>
> I've seen mention of people leaving hugin running for hours on end,
> are you swapping? the speed just goes into the dirt when I hit
> swap...maybe I could try putting
38 matches
Mail list logo
