Weird-looking Fast Preview window

98 views
Skip to first unread message

John Muccigrosso

unread,
Nov 6, 2014, 12:34:08 PM11/6/14
to hugi...@googlegroups.com
Hugin 2014 RC4 on a MacBook Pro running OS X 10.9.5 with Intel HD Graphics 3000 512 MB

Not always, but often I get weird previews in the Fast Preview window. See the attached. The stitching will go just fine, but of course I can't use any of the functions in Fast Preview.

Any help?
Screenshot 2014-11-06 12.26.44.png

Frederic Da Vitoria

unread,
Nov 6, 2014, 12:55:42 PM11/6/14
to hugin-ptx
I sometimes get this too on Windows. I may be wrong, but I recently managed to get the preview correctly by just closing Hugin and then loading it again. But I think (I did not pay much attention) that sometimes this did not work.

--
Frederic Da Vitoria
(davitof)

Membre de l'April - « promouvoir et défendre le logiciel libre » - http://www.april.org

Michael Perry

unread,
Nov 6, 2014, 12:59:28 PM11/6/14
to hugi...@googlegroups.com
Me too on OSX 10.7.5. Quite irregular, though I will pay more attention next time

John Muccigrosso

unread,
Nov 6, 2014, 1:09:15 PM11/6/14
to hugi...@googlegroups.com
I've tried to quit and restart the app, even boosting the image cache memory to 2048 when I did it. No luck.

Now, I just added some more control points and re-optimized and all is well. Not sure those steps had anything to do with it though.

Frederic Da Vitoria

unread,
Nov 6, 2014, 1:44:47 PM11/6/14
to hugin-ptx
2014-11-06 19:09 GMT+01:00 John Muccigrosso <jmuc...@gmail.com>:
I've tried to quit and restart the app, even boosting the image cache memory to 2048 when I did it. No luck.

Now, I just added some more control points and re-optimized and all is well. Not sure those steps had anything to do with it though.

I am almost sure that some (most?) optimization results will never trigger this bug, so that re-optimizing quite probably was why the issue disappeared. I always use the "progressive" optimization technique (where you first optimize by positions, then by positions-and-view and so on and I often noticed that when this bug appeared, I could avoid it by resetting all optimization parameters and then stopping at the last step which did not trigger the bug. So my workflow when seeing this issue would be:
- use one optimization level
- check in the preview if the bug does not appear
- use the next optimization level
...
- if the bug appears, undo the last optimization.

John Muccigrosso

unread,
Nov 6, 2014, 1:58:07 PM11/6/14
to hugi...@googlegroups.com
Good to know. Did you file a bug report?

Frederic Da Vitoria

unread,
Nov 6, 2014, 2:23:59 PM11/6/14
to hugin-ptx
2014-11-06 19:58 GMT+01:00 John Muccigrosso <jmuc...@gmail.com>:
Good to know. Did you file a bug report?

No. I know, I should have. Maybe someone else did... No, I just checked and couldn't find anything. I could enter a bug, but I currently don't have any pto file to show the issue, so it would be useless. I'll do it next time it happens, if no one beats me to it.

Marius Loots

unread,
Nov 7, 2014, 4:57:31 AM11/7/14
to hugi...@googlegroups.com
Hallo Everyone

Thursday, November 6, 2014, 7:59:28 PM, you wrote:
Michael> Me too on OSX 10.7.5. Quite irregular, though I will pay more attention
Michael> next time

>>> Not always, but often I get weird previews in the Fast Preview window.
>>> See the attached. The stitching will go just fine, but of course I can't
>>> use any of the functions in Fast Preview.
>>
>> I sometimes get this too on Windows. I may be wrong, but I recently
>> managed to get the preview correctly by just closing Hugin and then loading
>> it again. But I think (I did not pay much attention) that sometimes this
>> did not work.

On Windows 7. Have seen this, but never thought that it could be a
bug. Mostly restart the project from scratch, assuming it to be a
problem with a control point or two. Impression I had, but never
actively followed up on, is that two images are connected with a
faulty control point and that hugin attempts to bridge them via the
spokes that are criss-crossing the field.

I also use 2012 and 2013 interchangeably, so not able to pin it on a
specific version.

Will attempt to reproduce if I see it again.


Groetnis
Marius
mailto:mlo...@medic.up.ac.za
--
add some chaos to your life and put the world in order
http://www.mapungubwe.co.za/
http://www.chaos.co.za/
skype: marius_loots

Hierdie boodskap en aanhangsels is aan 'n vrywaringsklousule
onderhewig. Volledige besonderhede is by
www.it.up.ac.za/documentation/governance/disclaimer/
beskikbaar.

John Muccigrosso

unread,
Nov 7, 2014, 9:49:55 AM11/7/14
to hugi...@googlegroups.com, mlo...@medic.up.ac.za
On Friday, November 7, 2014 4:57:31 AM UTC-5, Marius Loots wrote:
Hallo Everyone

On Windows 7. Have seen this, but never thought that it could be a
bug. Mostly restart the project from scratch, assuming it to be a
problem with a control point or two. Impression I had, but never
actively followed up on, is that two images are connected with a
faulty control point and that hugin attempts to bridge them via the
spokes that are criss-crossing the field.

I also use 2012 and 2013 interchangeably, so not able to pin it on a
specific version.

I got it to stop on this project by adding more CPs, not taking any away. 

John Muccigrosso

unread,
Feb 26, 2015, 4:19:28 PM2/26/15
to hugi...@googlegroups.com
Got another one like this. Interestingly if I skip the fast preview window and just go to the Stitcher pane, I can get a rectilinear output just fine.

The effect appears only with that project (and a little on triplane). All the others are fine.

Contents of the pro are in-line below.


# hugin project file
#hugin_ptoversion 2
p f2 w8338 h2063 v194  E13.9366 R0 S42,7332,611,1826 n"TIFF_m c:LZW"
m g1 i0 f0 m2 p0.00784314

# image lines
#-hugin  cropFactor=4.8
i w3264 h2448 f0 v71.569775446991 Ra0 Rb0 Rc0 Rd0 Re0 Eev13.9366378213403 Er1 Eb1 r0 p0 y-80.5404573778347 TrX0 TrY0 TrZ0 Tpy0 Tpp0 j0 a0 b-0.0875484292446481 c0 d0 e0 g0 t0 Va1 Vb0 Vc0 Vd0 Vx0 Vy0  Vm5 n"/Users/john_muccigrosso/Pictures/Vicus images/Masters/2009/15 June 2009_2/100_2568.JPG"
#-hugin  cropFactor=4.80000019073486
i w3264 h2448 f0 v=0 Ra=0 Rb=0 Rc=0 Rd=0 Re=0 Eev14.6147098272241 Er1 Eb1 r3.19585354253604 p-4.18424447236812 y-37.917728259893 TrX0 TrY0 TrZ0 Tpy0 Tpp0 j0 a=0 b=0 c=0 d=0 e=0 g=0 t=0 Va=0 Vb=0 Vc=0 Vd=0 Vx=0 Vy=0  Vm5 n"/Users/john_muccigrosso/Pictures/Vicus images/Masters/2009/15 June 2009_2/100_2569.JPG"
#-hugin  cropFactor=4.80000019073486
i w3264 h2448 f0 v=0 Ra=0 Rb=0 Rc=0 Rd=0 Re=0 Eev13.9366378213403 Er1 Eb1 r1.42381380953124 p-7.27818500819498 y4.98069633229237 TrX0 TrY0 TrZ0 Tpy0 Tpp0 j0 a=0 b=0 c=0 d=0 e=0 g=0 t=0 Va=0 Vb=0 Vc=0 Vd=0 Vx=0 Vy=0  Vm5 n"/Users/john_muccigrosso/Pictures/Vicus images/Masters/2009/15 June 2009_2/100_2570.JPG"
#-hugin  cropFactor=4.80000019073486
i w3264 h2448 f0 v=0 Ra=0 Rb=0 Rc=0 Rd=0 Re=0 Eev14.6147098272241 Er1 Eb1 r-4.31763249067973 p-6.05011075808428 y39.6694040914257 TrX0 TrY0 TrZ0 Tpy0 Tpp0 j0 a=0 b=0 c=0 d=0 e=0 g=0 t=0 Va=0 Vb=0 Vc=0 Vd=0 Vx=0 Vy=0  Vm5 n"/Users/john_muccigrosso/Pictures/Vicus images/Masters/2009/15 June 2009_2/100_2571.JPG"


# specify variables that should be optimized
v v0
v Ra0
v Rb0
v Rc0
v Rd0
v Re0
v Vb0
v Vc0
v Vd0
v Eev1
v r1
v p1
v y1
v Eev2
v r2
v p2
v y2
v Eev3
v r3
v p3
v y3
v


# control points
c n2 N3 x2429.03224564391 y1749.22080566635 X967.961580162426 Y1759.3544274765 t0
c n2 N3 x2556.5689911605 y1752.71057217885 X1095.36522482161 Y1765.02219867082 t0
c n2 N3 x2617.81385013405 y1762.86737893789 X1159.39515422852 Y1775.76429174853 t0
c n2 N3 x2739.48498222567 y1683.56937782587 X1278.66599118363 Y1695.00294556666 t0
c n2 N3 x2805.72411763484 y1775.59793546303 X1332.56249155008 Y1788.46915346177 t0
c n2 N3 x2956.64931179055 y1714.58645904045 X1472.54895196052 Y1726.7704340417 t0
c n1 N2 x2878.23754001693 y1404.68040993574 X993 Y1341 t0
c n1 N2 x2648.67488071321 y1663.82012193501 X872 Y1615 t0
c n1 N2 x2198.08247266611 y1047.72327598955 X313 Y1004 t0
c n1 N2 x2945.04079809108 y1671.09102372899 X1199 Y1623 t0
c n1 N2 x3053 y1222 X1132.755758106 Y1143.80834650593 t0
c n2 N3 x1522.05724081172 y1655.1940033898 X64 Y1664 t0
c n2 N3 x1677 y1058 X83 Y1093 t0
c n2 N3 x2642 y821 X1241.23943661972 Y821.746478873239 t0
c n0 N1 x2398 y1860 X628.86730215175 Y1734.95027118349 t0
c n0 N1 x1965.17935092488 y853.098266356968 X282 Y753 t0
c n0 N1 x2801 y1209 X937 Y1075 t0
c n0 N1 x3117 y412 X1352.32580453812 Y263.943909736371 t0
c n2 N3 x1838 y1657 X360.141267681625 Y1662.54461926042 t0
c n2 N3 x2155 y1649 X663.171687009634 Y1657.64213081665 t0
c n1 N2 x2350 y1672 X552.925592891253 Y1622.43180757434 t0
c n1 N2 x2035 y1660 X238.280014333458 Y1628.21356042751 t0
c n1 N2 x1926 y837 X45.9718309859155 Y804.507042253521 t0
c n2 N3 x2481 y890 X856.225352112676 Y902.197183098592 t0
c n2 N3 x2729 y1637 X1275.33193927573 Y1647.1096622596 t0
c n2 N3 x2108 y1358 X476.957746478873 Y1367.66197183099 t0
c n1 N2 x3231 y1695 X1522.1410512552 Y1642.62610795079 t0
c n2 N3 x2491.40107856588 y1657.1518831527 X1007 Y1666 t0

# masks
k i0 t0 p"3297 -41 -32 -36 -32 2538 3297 2538 3280 1903 1009 1876 941 441 3324 387"
k i1 t0 p"-36 1773 3347 1786 3315 2528 -45 2515"
k i1 t0 p"-41 -41 3292 -36 3301 481 -59 445"
k i1 t0 p"-32 1290 982 1715 1000 1969 -77 1965"
k i2 t0 p"-50 1724 3319 1742 3313.89 2538 2267.51 2538 -27 2492"
k i2 t0 p"-50 -50 3297 -50 3306 494 -59 423"
k i2 t0 p"-32 1317 110 1371 311 1411 562 1487 781 1554 906 1576 848 1733 816 1804 -27 1831"
k i3 t0 p"-41 1750 3301 1720 3297 2537 -41 2475 -41 2479"
k i3 t0 p"-40 329 -41 -32 3297 -14 3310 334 3310 334 3306 334"
k i3 t0 p"2832 -68 3319 -72 3310 2609 2810 2622 2810 2622"

#hugin_optimizeReferenceImage 0
#hugin_blender enblend
#hugin_remapper nona
#hugin_enblendOptions 
#hugin_enfuseOptions 
#hugin_hdrmergeOptions -m avg -c
#hugin_outputLDRBlended true
#hugin_outputLDRLayers false
#hugin_outputLDRExposureRemapped false
#hugin_outputLDRExposureLayers false
#hugin_outputLDRExposureBlended false
#hugin_outputLDRStacks false
#hugin_outputLDRExposureLayersFused false
#hugin_outputHDRBlended false
#hugin_outputHDRLayers false
#hugin_outputHDRStacks false
#hugin_outputLayersCompression LZW
#hugin_outputImageType png
#hugin_outputImageTypeCompression LZW
#hugin_outputJPEGQuality 90
#hugin_outputImageTypeHDR exr
#hugin_outputImageTypeHDRCompression LZW
#hugin_outputStacksMinOverlap 0.7
#hugin_outputLayersExposureDiff 0.5
#hugin_optimizerMasterSwitch 6
#hugin_optimizerPhotoMasterSwitch 21

Terry Duell

unread,
Feb 26, 2015, 4:34:29 PM2/26/15
to hugi...@googlegroups.com
Hello John,

On Fri, 27 Feb 2015 08:19:28 +1100, John Muccigrosso <jmuc...@gmail.com>
wrote:

> Got another one like this. Interestingly if I skip the fast preview
> window and just go to the Stitcher pane, I can get a rectilinear output
> just fine.
>
> The effect appears only with that project (and a little on triplane).
> All the others are fine.
>

triplane can be a bit problematic.

> Contents of the pro are in-line below.

Probably best to attach the .pto as a text file. When pasted into an email
it ends up being buggered up by spurious characters being inserted by the
mailer formatting.


Cheers,
--
Regards,
Terry Duell

John Muccigrosso

unread,
Feb 26, 2015, 4:37:36 PM2/26/15
to hugi...@googlegroups.com
Yeah, I just tried out all the projections to see what they looked like. It surprised me that rectilinear was so different from the others.

pto attached now.
US 89.pto

Terry Duell

unread,
Feb 26, 2015, 4:44:12 PM2/26/15
to hugi...@googlegroups.com
Hello John,

On Fri, 27 Feb 2015 08:19:28 +1100, John Muccigrosso <jmuc...@gmail.com>
wrote:

> Got another one like this. Interestingly if I skip the fast preview
> window
> and just go to the Stitcher pane, I can get a rectilinear output just
> fine.
>
> The effect appears only with that project (and a little on triplane). All
> the others are fine.
>
> Contents of the pro are in-line below.
>

Loading the .pto into hugin using dummy images, it looks like there are
some spurious control points linking images 3 and 2.

John Muccigrosso

unread,
Feb 26, 2015, 4:52:40 PM2/26/15
to hugi...@googlegroups.com
Terry,

How did you see that? Distances on the control points tab?

Thanks. 

Terry Duell

unread,
Feb 26, 2015, 5:00:24 PM2/26/15
to hugi...@googlegroups.com
Hello John,

On Fri, 27 Feb 2015 08:37:36 +1100, John Muccigrosso <jmuc...@gmail.com>
wrote:

> Yeah, I just tried out all the projections to see what they looked like.
> It surprised me that rectilinear was so different from the others.
> pto attached now.

I get a better result with this copy of the .pto.
Looking at the Fast Panorama preview, using rectilinear projection, it
initially looks pretty crook, but can be dragged into submission. Changing
to other projections looks OK.
The fit isn't real good, there are some nasty control points.
The rectilinear looks nasty simply because of the overall FOV which is
beyond the limits for this projection.
Here it doesn't look all that bad, apart from the few bad control points.
Is this in accord with what you're seeing? If not, do a screen capture.

Terry Duell

unread,
Feb 26, 2015, 5:06:30 PM2/26/15
to hugi...@googlegroups.com
On Fri, 27 Feb 2015 08:52:40 +1100, John Muccigrosso <jmuc...@gmail.com>
wrote:
Looking at the arrangement of CPs in the images in CP tab, as well as
'show control points' in the Preview, but I may not have been seeing the
real thing.
Your later .pto provided as an attachment gave a different (better) result
here, as discussed in my previous email.

John Muccigrosso

unread,
Feb 26, 2015, 5:09:52 PM2/26/15
to hugi...@googlegroups.com
Do you mean you're eyeballing it to find bad control points?

(This is where I have to admit that I don't understand how Hugin works well enough.)

I'm going to combine my response to your other message here too:

> Looking at the Fast Panorama preview, using rectilinear projection, it iinitially looks pretty crook, but can be dragged into submission. 

What do you mean here? What are you dragging?

> Changing to other projections looks OK. 

Yes.

> The fit isn't real good, there are some nasty control points.

Based on what? (Not that I disagree!)

> The rectilinear looks nasty simply because of the overall FOV which is beyond the limits for this projection. 

Yes, these are some old photos I am trying to play with. I didn't take them with this process in mind exactly.

> Here it doesn't look all that bad, apart from the few bad control points.

Yes, I can get a not-so-bad version out of it, but not with the Fast Preview process.

Terry Duell

unread,
Feb 26, 2015, 5:50:27 PM2/26/15
to hugi...@googlegroups.com
Hello John,
A couple of our emails crossed, so I'll try to clarify as well as I can by
responding to your questions below.

On Fri, 27 Feb 2015 09:09:52 +1100, John Muccigrosso <jmuc...@gmail.com>
wrote:

> On Thursday, February 26, 2015 at 5:06:30 PM UTC-5, Tduell wrote:

>> > How did you see that? Distances on the control points tab?
>> >
>>
>> Looking at the arrangement of CPs in the images in CP tab, as well as
>> 'show control points' in the Preview, but I may not have been seeing the
>> real thing.
>> Your later .pto provided as an attachment gave a different (better)
>> result
>>
>> here, as discussed in my previous email.
>>
>>
> Do you mean you're eyeballing it to find bad control points?

I don't see your images, as I'm using dummy images (generated by ptodummy,
part of Bruno Postle's Perl Panotools Script), which allows me to load the
.pto and see everything as it would be but with plain coloured images.
When I looked at the control points tab, I can see the arrangement of the
CPs in overlapping image pairs, if the arrangement isn't the same in each
of the images there is a problem. I also look at the error distances
reported.
I also look at the Fast Panorama preview window, using Preview and "show
control points", this shows the bad control points pretty clearly.
But...the really nasty control points I was seeing were an artefact of
using the .pto that I extracted from your initial post, and it all looked
quite a bit better using the .pto you attached in the later
post...although that does still have a few CPs that are pretty poor.

>
> (This is where I have to admit that I don't understand how Hugin works
> well enough.)
>
> I'm going to combine my response to your other message here too:
>
>> Looking at the Fast Panorama preview, using rectilinear projection, it
> iinitially looks pretty crook, but can be dragged into submission.
>
> What do you mean here? What are you dragging?

In the Fast Panorama preview I select the "Move/Drag" tab and then select
the displayed pano (anywhere) with the left mouse and drag it around until
it looks a bit more like it should. Sometimes the pano can be way off
centre and if in rectilinear projection it can become horribly distorted.
You can also select with the right mouse and dragging will rotate the pano.

>
>> Changing to other projections looks OK.
>
> Yes.
>
>> The fit isn't real good, there are some nasty control points.
>
> Based on what? (Not that I disagree!)

Based on what the Fast Panorama preview tells me in the "Assistant" tab,
and also looking at the distances in the Control Points tab, and seeing
the differences displayed in the Fast Panorama preview in the Preview tab
with "show control points" selected.

>
>> The rectilinear looks nasty simply because of the overall FOV which is
> beyond the limits for this projection.
>
> Yes, these are some old photos I am trying to play with. I didn't take
> them with this process in mind exactly.
>
>> Here it doesn't look all that bad, apart from the few bad control
>> points.
>
> Yes, I can get a not-so-bad version out of it, but not with the Fast
> Preview process.

OK, there may be some differences in behaviour of the Fast Panorama
preview due to differences in OS and version of hugin. I am running Fedora
21 and hugin-2014.1.0 (a build of the current development source).

Hope this helps.

John Muccigrosso

unread,
Feb 26, 2015, 5:54:58 PM2/26/15
to hugi...@googlegroups.com
OK. Thanks. I'm on a Mac. I'm going to play with this a bit and see what I can do/learn.

Appreciate the help.
Reply all
Reply to author
Forward
0 new messages