## How control points are determined in bracketed shots using PTGUI

Expand Messages
• From white photo are the control points linked from one bracketed shot to another? I would assume that it would be the first bracketed shot in the series? I
Message 1 of 26 , Nov 30, 2007
From white photo are the control points linked from one bracketed
shot to another? I would assume that it would be the first bracketed
shot in the series?

I ask this because I have taken 4 series of breacketed shots ( 3, 5,
7, 9) on a heavy manfrotto tripod, con carpet with a precision 360
head. I put colored pieces of paper on the wall and on the ceilings
so that hopefully as many control points can be found very easily.
Given the stops on the detent ring may actually be different I
started and ended each series in the same place. Manual focus, manual
white balance.

All 4 series have at least 6shots -15 , 6shots +15 and 1 vertical
shot in common that they all should be used to find control points.

I have not been using this software very long so I am just trying to
get an understanding why these numbers are so different if the
program detects bracketed sets then it should only be looking for
control points in one image in the bracket set. The image is the same
in all series.

Avg Min Max
3 - 39 images =((6*3)*2)+3 2.08 0.11 32.44
5 - 65 images =((6*5)*2)+5 1.94 0.04 37.81
7 - 91 images =((6*7)*2)+7 2.05 0.05 79.80
9 - 117 images=((6*9)*2)+9 1.97 0.04 98.52

Thank you

KieranMullen
• The farther out of range from a balance exposure, to more you have clipping in shadows and highlights. It s much hard to get a control point when there is not
Message 2 of 26 , Nov 30, 2007
The farther out of range from a balance exposure, to more you have clipping

It's much hard to get a control point when there is not very much detail.

So when it does occur, it becomes more error prone.

I'm curious however, as to why you are using control points.

I thought that was the purpose of the precision 360 head. No need for
control points.

Otherwise, much cheaper heads can be used.

Sacha Griffin
Southern Digital Solutions LLC - Atlanta, Georgia
www.southern-digital.com
www.seeit360.net
www.ezphotosafe.com
404-551-4275
404-731-7798

_____

From: PanoToolsNG@yahoogroups.com [mailto:PanoToolsNG@yahoogroups.com] On
Behalf Of verifone411
Sent: Friday, November 30, 2007 5:20 PM
To: PanoToolsNG@yahoogroups.com
Subject: [PanoToolsNG] How control points are determined in bracketed shots
using PTGUI

From white photo are the control points linked from one bracketed
shot to another? I would assume that it would be the first bracketed
shot in the series?

I ask this because I have taken 4 series of breacketed shots ( 3, 5,
7, 9) on a heavy manfrotto tripod, con carpet with a precision 360
head. I put colored pieces of paper on the wall and on the ceilings
so that hopefully as many control points can be found very easily.
Given the stops on the detent ring may actually be different I
started and ended each series in the same place. Manual focus, manual
white balance.

All 4 series have at least 6shots -15 , 6shots +15 and 1 vertical
shot in common that they all should be used to find control points.

I have not been using this software very long so I am just trying to
get an understanding why these numbers are so different if the
program detects bracketed sets then it should only be looking for
control points in one image in the bracket set. The image is the same
in all series.

Avg Min Max
3 - 39 images =((6*3)*2)+3 2.08 0.11 32.44
5 - 65 images =((6*5)*2)+5 1.94 0.04 37.81
7 - 91 images =((6*7)*2)+7 2.05 0.05 79.80
9 - 117 images=((6*9)*2)+9 1.97 0.04 98.52

Thank you

KieranMullen

[Non-text portions of this message have been removed]
• I am creating templates for myself since I do have the head so I do not need to create control points in the future. However I am trying to get them as good as
Message 3 of 26 , Nov 30, 2007
I am creating templates for myself since I do have the head so I do
not need to create control points in the future. However I am trying
to get them as good as possible.

I would understand your point if the points within the series were
not linked, but they are. Thus only the first image should get
control points generated. The images images do within the bracketed
set do not need control points from example 0 & 7 (why cant we start
at 1?) have good control points set so do 1 and 8 really need them?

It is more work that doesnt need to be computed. In my humble (not

KM

--- In PanoToolsNG@yahoogroups.com, "Sacha Griffin"
<sachagriffin@...> wrote:
>
> The farther out of range from a balance exposure, to more you have
clipping
>
> It's much hard to get a control point when there is not very much
detail.
>
> So when it does occur, it becomes more error prone.
>
>
>
> I'm curious however, as to why you are using control points.
>
> I thought that was the purpose of the precision 360 head. No need
for
> control points.
>
> Otherwise, much cheaper heads can be used.
>
>
>
> Sacha Griffin
> Southern Digital Solutions LLC - Atlanta, Georgia
> www.southern-digital.com
> www.seeit360.net
> www.ezphotosafe.com
> 404-551-4275
> 404-731-7798
>
> _____
>
> From: PanoToolsNG@yahoogroups.com
[mailto:PanoToolsNG@yahoogroups.com] On
> Behalf Of verifone411
> Sent: Friday, November 30, 2007 5:20 PM
> To: PanoToolsNG@yahoogroups.com
> Subject: [PanoToolsNG] How control points are determined in
bracketed shots
> using PTGUI
>
>
>
> From white photo are the control points linked from one bracketed
> shot to another? I would assume that it would be the first
bracketed
> shot in the series?
>
> I ask this because I have taken 4 series of breacketed shots ( 3,
5,
> 7, 9) on a heavy manfrotto tripod, con carpet with a precision 360
> head. I put colored pieces of paper on the wall and on the ceilings
> so that hopefully as many control points can be found very easily.
> Given the stops on the detent ring may actually be different I
> started and ended each series in the same place. Manual focus,
manual
> white balance.
>
> All 4 series have at least 6shots -15 , 6shots +15 and 1 vertical
> shot in common that they all should be used to find control points.
>
> I have not been using this software very long so I am just trying
to
> get an understanding why these numbers are so different if the
> program detects bracketed sets then it should only be looking for
> control points in one image in the bracket set. The image is the
same
> in all series.
>
> Avg Min Max
> 3 - 39 images =((6*3)*2)+3 2.08 0.11 32.44
> 5 - 65 images =((6*5)*2)+5 1.94 0.04 37.81
> 7 - 91 images =((6*7)*2)+7 2.05 0.05 79.80
> 9 - 117 images=((6*9)*2)+9 1.97 0.04 98.52
>
> Thank you
>
> KieranMullen
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
• OK sure, but you don t need to start using hdr, to create a template since, as you said they are linked. You can apply on cp template from an ldr to an hdr
Message 4 of 26 , Nov 30, 2007
OK sure, but you don't need to start using hdr, to create a template since,
as you said they are linked.

You can apply on cp template from an ldr to an hdr project, and then link
and save, and it should work fine.

And yes, the cp generator wastes time computing CP on images it will never
use.

Sacha Griffin
Southern Digital Solutions LLC - Atlanta, Georgia
www.southern-digital.com
www.seeit360.net
www.ezphotosafe.com
404-551-4275
404-731-7798

_____

From: PanoToolsNG@yahoogroups.com [mailto:PanoToolsNG@yahoogroups.com] On
Behalf Of verifone411
Sent: Friday, November 30, 2007 10:27 PM
To: PanoToolsNG@yahoogroups.com
Subject: [PanoToolsNG] Re: How control points are determined in bracketed
shots using PTGUI

I am creating templates for myself since I do have the head so I do
not need to create control points in the future. However I am trying
to get them as good as possible.

I would understand your point if the points within the series were
not linked, but they are. Thus only the first image should get
control points generated. The images images do within the bracketed
set do not need control points from example 0 & 7 (why cant we start
at 1?) have good control points set so do 1 and 8 really need them?

It is more work that doesnt need to be computed. In my humble (not

com/preview/183/Untitled.jpg

KM

--- In PanoToolsNG@ <mailto:PanoToolsNG%40yahoogroups.com> yahoogroups.com,
"Sacha Griffin"
<sachagriffin@...> wrote:
>
> The farther out of range from a balance exposure, to more you have
clipping
>
> It's much hard to get a control point when there is not very much
detail.
>
> So when it does occur, it becomes more error prone.
>
>
>
> I'm curious however, as to why you are using control points.
>
> I thought that was the purpose of the precision 360 head. No need
for
> control points.
>
> Otherwise, much cheaper heads can be used.
>
>
>
> Sacha Griffin
> Southern Digital Solutions LLC - Atlanta, Georgia
> www.southern-digital.com
> www.seeit360.net
> www.ezphotosafe.com
> 404-551-4275
> 404-731-7798
>
> _____
>
> From: PanoToolsNG@ <mailto:PanoToolsNG%40yahoogroups.com> yahoogroups.com
[mailto:PanoToolsNG@ <mailto:PanoToolsNG%40yahoogroups.com> yahoogroups.com]
On
> Behalf Of verifone411
> Sent: Friday, November 30, 2007 5:20 PM
> To: PanoToolsNG@ <mailto:PanoToolsNG%40yahoogroups.com> yahoogroups.com
> Subject: [PanoToolsNG] How control points are determined in
bracketed shots
> using PTGUI
>
>
>
> From white photo are the control points linked from one bracketed
> shot to another? I would assume that it would be the first
bracketed
> shot in the series?
>
> I ask this because I have taken 4 series of breacketed shots ( 3,
5,
> 7, 9) on a heavy manfrotto tripod, con carpet with a precision 360
> head. I put colored pieces of paper on the wall and on the ceilings
> so that hopefully as many control points can be found very easily.
> Given the stops on the detent ring may actually be different I
> started and ended each series in the same place. Manual focus,
manual
> white balance.
>
> All 4 series have at least 6shots -15 , 6shots +15 and 1 vertical
> shot in common that they all should be used to find control points.
>
> I have not been using this software very long so I am just trying
to
> get an understanding why these numbers are so different if the
> program detects bracketed sets then it should only be looking for
> control points in one image in the bracket set. The image is the
same
> in all series.
>
> Avg Min Max
> 3 - 39 images =((6*3)*2)+3 2.08 0.11 32.44
> 5 - 65 images =((6*5)*2)+5 1.94 0.04 37.81
> 7 - 91 images =((6*7)*2)+7 2.05 0.05 79.80
> 9 - 117 images=((6*9)*2)+9 1.97 0.04 98.52
>
> Thank you
>
> KieranMullen
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

[Non-text portions of this message have been removed]
• Kiran, I have read the other answers but just want to address one issue: You say heavy manfrotto tripod CON CARPET Carpets are the worst and your
Message 5 of 26 , Nov 30, 2007
Kiran,

heavy manfrotto tripod CON CARPET" Carpets are the worst and your
360precision will be useless on a carpet. I once attached a laser to the
camera, pointing it upward to the ceiling. When the tripod is on a carpet
you can see the laser move up to 1 inch on the ceiling. Even on a wooden
floor without carpet there is still a lot of movement while you turn with
the camera to stay out of the shot.

Carel Struycken
--
View this message in context: http://www.nabble.com/How-control-points-are-determined-in-bracketed-shots-using-PTGUI-tf4926132.html#a14102301
Sent from the PanoToolsNG mailing list archive at Nabble.com.
• Well ow we are going to the question of why I did what I did, rather than why does it do what it does... I am just trying to figure out why it does what it
Message 6 of 26 , Dec 1, 2007
Well ow we are going to the question of why I did what I did, rather
than why does it do what it does... I am just trying to figure out
why it does what it does. I doubt I will take 9 HDR image shots too
often anyway.

I would also seem to think that if one could pre align the pictures
in the program via low res preview that woudl save computing time as
well.

Or does it do it already based on the order of which they are loaded?

Thank you

KieranMullen

--- In PanoToolsNG@yahoogroups.com, "Sacha Griffin"
<sachagriffin@...> wrote:
>
> OK sure, but you don't need to start using hdr, to create a
template since,
> as you said they are linked.
>
> You can apply on cp template from an ldr to an hdr project, and
> and save, and it should work fine.
>
> And yes, the cp generator wastes time computing CP on images it
will never
> use.
>
>
>
> Sacha Griffin
> Southern Digital Solutions LLC - Atlanta, Georgia
> www.southern-digital.com
> www.seeit360.net
> www.ezphotosafe.com
> 404-551-4275
> 404-731-7798
>
• I don t fully understand the work flow for bracketed shots in PTgui. I too have problems If I bring all my bracket shots in at once and try to generate control
Message 7 of 26 , Dec 1, 2007
I don't fully understand the work flow for bracketed shots in PTgui.
I too have problems If I bring all my bracket shots in at once and
try to generate control points. It seems that PTgui should only
generate points for one bracketed series rather than all of them. If
you have a lot of images, it can get kind of confusing sorting through
all those points.

So here is what I have been doing, with good results:

I bring in my main bracketed series (usually the medium bracket). I
stitch that as normal in PTGui. When satisfied, I then bring in the
other bracketed shots using the Source Images tab and position the
images accordingly using the Up and Down arrows.

After bring in the other bracketed series of shots, click on the HDR
tab & enable HDR stitching. PTGui recognizes the brackets as an HDR
sequence. From there you can either link the images so that the other
brackets stitch exactly the same as your main bracket, or you can
keep the images unlinked, at which point PTGu is supposed to align
them as best as possible.

At this point, however, I have my own question:

When rendering a HDR sequence with unlinked images - do the other
bracketed sequences need their own control points? Or will the
unlinked images with no control points be generated as if they were
linked, and PTGui will then attempt to align those as best as possible?

I did a test on a HDR pano that had control points associated with
only one bracket. I did two renderings, one as linked and one
unlinked. There was no difference I could see in the final output.

Can anyone offer more work flow tips on using unlinked images in PTgui?

--- In PanoToolsNG@yahoogroups.com, "verifone411" <kieranmullen@...>
wrote:
>
> Well ow we are going to the question of why I did what I did, rather
> than why does it do what it does... I am just trying to figure out
> why it does what it does. I doubt I will take 9 HDR image shots too
> often anyway.
>
> I would also seem to think that if one could pre align the pictures
> in the program via low res preview that woudl save computing time as
> well.
>
> Or does it do it already based on the order of which they are loaded?
>
> Thank you
>
> KieranMullen
>
• ... Because in over- and underexposed areas there are no control points possible. I suspect one shoots HDR because there is a large dynamic range. If one side
Message 8 of 26 , Dec 1, 2007
On Saturday, December 01, 2007 at 3:27, verifone411 wrote:

> Thus only the first image should get control points generated. The
> images images do within the bracketed set do not need control points

Because in over- and underexposed areas there are no control points
possible. I suspect one shoots HDR because there is a large dynamic
range. If one side of the first image of a bracketed set is
completely overexposed but the other isn't, PTGui needs to use the
darker images instead to generate relevant CPs. But on the darker one
the other side might be underexposed, hence the control points from
the first image are needed...

You could argue that it would be sufficient to generate CPs in other
but the first image only if in the first one there weren't relevant
ones. But if the user decides to unlink images after that, there are
no control points in lots of images. I guess that is why Joost
decided to create CPs for all of them, even if they are linked...

> from example 0 & 7 (why cant we start at 1?)

For historical reasons: C programmers always start counting with 0.
;-)

Best regards
--
http://www.erik-krause.de
• Hi Carel, I normally try and stay out of these debates but making such a broad sweeping statement about the 360Precision head is simply not acceptable. One
Message 9 of 26 , Dec 2, 2007
Hi Carel,

I normally try and stay out of these debates but making such a broad
acceptable. One thing I've learnt whilst dealing with thousands of
photographers over the past 3-4 years is experience means nothing.
Just because someone has been involved in panoramic photography since
it's inception doesn't mean they actually know what they're talking
about. In-fact it quite often means there so stuck in their ways that
they're not willing to accept change.

I've seen complete beginners create technically perfect panoramas
after a weekend of training yet others take years. Shooting on carpet
if you have a properly set-up tripod and head shouldn't pose any real
problems.

The twin bearing set-up in the 360Precision Absolute and Adjuste base
is designed specifically to isolate as much head vibration and
movement from effecting or moving the tripod. It's this very reason
that no other head whether the manufacturer calls it a precision or
not, will never work as well. One sealed bearing or plastic bush will
never work as well, we know, we tried the same designs.

If your tripod or set-up is moving on carpet and hard-floors then I
happen. Another cause is standing directly next to the tripod leg when
shooting. Try to stand as far away from the tripod as possible and
always use mirror lock-up and a remote release if possible.

A better test is to shoot 3-4 panoramas in succession and overlay the
individual images in Photoshop. Switch the blend mode of the top image
to difference, the results should be as close to a black image as
possible. The blend image will never be completely black as two images
will never be 100% identical.

Matt
360Precision Ltd
http://www.360precision.com

On 1 Dec 2007, at 05:02, Carel wrote:

>
> Kiran,
>
> You say "
> heavy manfrotto tripod CON CARPET" Carpets are the worst and your
> 360precision will be useless on a carpet. I once attached a laser to
> the
> camera, pointing it upward to the ceiling. When the tripod is on a
> carpet
> you can see the laser move up to 1 inch on the ceiling. Even on a
> wooden
> floor without carpet there is still a lot of movement while you turn
> with
> the camera to stay out of the shot.
>
> Carel Struycken
> --
> View this message in context: http://www.nabble.com/How-control-points-are-determined-in-bracketed-shots-using-PTGUI-tf4926132.html#a14102301
> Sent from the PanoToolsNG mailing list archive at Nabble.com.
>
>
>
• I understand what you and others have said about the over and under exposed images not having enough control points, however shouldnt the first image of a
Message 10 of 26 , Dec 2, 2007
I understand what you and others have said about the over and under
exposed images not having enough control points, however shouldnt the
first image of a bracketed set, the metered image, be able to
generate sufficient control points?

Would it make any difference to give the user the option to only use
control points from the metered shot whaterv order it is in ? eg. 0,-
,+ or -,0,+ etc just to take the 0 and use that as a model? Then
perhaps give the user the option to over ride the master image in
case like you say the image is deleted.

Thank you for your information! 3am time for bed.

KieranMullen

--- In PanoToolsNG@yahoogroups.com, "Erik Krause" <erik.krause@...>
wrote:
>
> On Saturday, December 01, 2007 at 3:27, verifone411 wrote:
>
> > Thus only the first image should get control points generated. The
> > images images do within the bracketed set do not need control
points
>
> Because in over- and underexposed areas there are no control points
> possible. I suspect one shoots HDR because there is a large dynamic
> range. If one side of the first image of a bracketed set is
> completely overexposed but the other isn't, PTGui needs to use the
> darker images instead to generate relevant CPs. But on the darker
one
> the other side might be underexposed, hence the control points from
> the first image are needed...
>
> You could argue that it would be sufficient to generate CPs in
other
> but the first image only if in the first one there weren't relevant
> ones. But if the user decides to unlink images after that, there
are
> no control points in lots of images. I guess that is why Joost
> decided to create CPs for all of them, even if they are linked...
>
> > from example 0 & 7 (why cant we start at 1?)
>
> For historical reasons: C programmers always start counting with 0.
> ;-)
>
> Best regards
> --
> http://www.erik-krause.de
>
• ... If you shoot f.e. EV 0, -2, 2 this should be the case, but what if you shoot -2, 0, 2 as some cameras do? Or even more brackets? For a standard room with a
Message 11 of 26 , Dec 2, 2007
On Sunday, December 02, 2007 at 11:24, verifone411 wrote:

> I understand what you and others have said about the over and under
> exposed images not having enough control points, however shouldnt the
> first image of a bracketed set, the metered image, be able to
> generate sufficient control points?

If you shoot f.e. EV 0, -2, 2 this should be the case, but what if
you shoot -2, 0, 2 as some cameras do? Or even more brackets? For a
standard room with a view you most often need 5 brackets (at 2 EV
steps).

There is absolutely no guarantee which image is the "correct"
exposure, since shooting a more than 3 brackets HDR will involve to
meter the darkest spot that still needs texture, meter the lightest
spot and choose the amount of brackets. It could well be that in such
a panorama most of the images in the first two brackets are almost
completely black, since you need the short exposure times only for
the windows and eventually lamps... In any case it is a good strategy
to get CPs from as many images as possible.

best regards
--
http://www.erik-krause.de
• I don t think Carel intended to say anything negative about the 360Precision at all Matt, the way I read it is more like even a panohead as good as the
Message 12 of 26 , Dec 2, 2007
I don't think Carel intended to say anything negative about the
360Precision at all Matt, the way I read it is more like "even a
panohead as good as the 360Precision cannot do its job properly on a
carpet", and I think he is perfectly right about that.
You need to use spiked tripod legs on surfaces like thick carpets or
soft soil and push them down really hard. When you can't do that
because of possible damage to the carpet you should at least make your
tripod as heavy as possible. Flexing wooden floors have been a serious
problem for me as well, standing as far away from the tripod legs as
possible helps, a motorized panohead rotator and/or a wireless remote
shutter release would be ideal for this.

Regards,

erik leeman

(www.erikleeman.com)
• ... Matt, Carel is in no way saying anything about the precision of the 360precision. He just says that there are other factors which can very easy change this
Message 13 of 26 , Dec 2, 2007
--- In PanoToolsNG@yahoogroups.com, Matthew Rogers <matthew@...> wrote:
>
> Hi Carel,
>
> I normally try and stay out of these debates but making such a broad
> acceptable.

Matt,
Carel is in no way saying anything about the precision of the 360precision.

He just says that there are other factors which can very easy change this precision.

I just did a very simple test.
My floor is standard 22mm plates with a carpet on and I placed my large heavy Manfrotto
tripod 058 on it.
I took a monopod and fastened it at a stable shelf 1,5 m from the tripod and moved it
until it almost touched the top of the tripod vertical pole.

The movement when walking around as you would normally do for taking the panorama
was 1mm.
That means that you get different pitches on the images and using template stitching will
give you errors.
And that has nothing to do with the panohead.

You say the same yourself below so why are you complaining.

Hans

One thing I've learnt whilst dealing with thousands of
> photographers over the past 3-4 years is experience means nothing.
> Just because someone has been involved in panoramic photography since
> it's inception doesn't mean they actually know what they're talking
> about. In-fact it quite often means there so stuck in their ways that
> they're not willing to accept change.
>
> I've seen complete beginners create technically perfect panoramas
> after a weekend of training yet others take years. Shooting on carpet
> if you have a properly set-up tripod and head shouldn't pose any real
> problems.
>
> The twin bearing set-up in the 360Precision Absolute and Adjuste base
> is designed specifically to isolate as much head vibration and
> movement from effecting or moving the tripod. It's this very reason
> that no other head whether the manufacturer calls it a precision or
> not, will never work as well. One sealed bearing or plastic bush will
> never work as well, we know, we tried the same designs.
>
> If your tripod or set-up is moving on carpet and hard-floors then I
> happen. Another cause is standing directly next to the tripod leg when
> shooting. Try to stand as far away from the tripod as possible and
> always use mirror lock-up and a remote release if possible.
>
> A better test is to shoot 3-4 panoramas in succession and overlay the
> individual images in Photoshop. Switch the blend mode of the top image
> to difference, the results should be as close to a black image as
> possible. The blend image will never be completely black as two images
> will never be 100% identical.
>
> Matt
> 360Precision Ltd
> http://www.360precision.com
>
> On 1 Dec 2007, at 05:02, Carel wrote:
>
> >
> > Kiran,
> >
> > You say "
> > heavy manfrotto tripod CON CARPET" Carpets are the worst and your
> > 360precision will be useless on a carpet. I once attached a laser to
> > the
> > camera, pointing it upward to the ceiling. When the tripod is on a
> > carpet
> > you can see the laser move up to 1 inch on the ceiling. Even on a
> > wooden
> > floor without carpet there is still a lot of movement while you turn
> > with
> > the camera to stay out of the shot.
> >
> > Carel Struycken
• ... ...or simply the self timer of the camera. 10 seconds should be enough to walk away from the tripod and eventually even close the door behind you... best
Message 14 of 26 , Dec 2, 2007
On Sunday, December 02, 2007 at 11:44, erik leeman wrote:

> Flexing wooden floors have been a serious
> problem for me as well, standing as far away from the tripod legs as
> possible helps, a motorized panohead rotator and/or a wireless remote
> shutter release would be ideal for this.

...or simply the self timer of the camera. 10 seconds should be
enough to walk away from the tripod and eventually even close the
door behind you...

best regards
--
http://www.erik-krause.de
• Sorry, but that wouldn t help much because you d have to approach and leave the camera for every single shot. Every move you make can (and will) disturb the
Message 15 of 26 , Dec 2, 2007
Sorry, but that wouldn't help much because you'd have to approach and
leave the camera for every single shot. Every move you make can (and
will) disturb the position of the camera under such circumstances.
Because the combined center of gravity of the camera + panohead is
never aligned with the vertical axis of rotation of the tripod some
displacement will always occur when the tripod is not stable enough,
therefore even a remotely controlled rotator might not prevent errors
completely.

Regards,

erik leeman

(www.erikleeman.com)
• Hi Hans, I do realise the context or Carel s statement but I m sick of wasting my time defending comments made on forums and email lists that ultimately have
Message 16 of 26 , Dec 2, 2007
Hi Hans,

I do realise the context or Carel's statement but I'm sick of wasting
ultimately have nothing to do with our products. My new tactic is to
have to explain to 20 people individually that you can in-fact use a
360precision set-up on carpet. A lot of the time people don't realise
how highly regarded comments on lists like this are taken, whether
they're correct or not.

The thing is Hans, you may realise what Carel meant. Believe me
though, there are a lot of people that will take his statement to mean
you can't use 360Precision with a template when shooting on carpet and
this is b.s.

Like I said, if you guys are having problems don't blame the head or
the carpet, blame your set-up or technique. Personally I've shot
hundreds of locations for real estate tours with 1-2 templates and
have not noticed any major problems. And I've shot rooms with carpet,
tiles, floorboards, slate, glass, I've even shot in swimming pools and
hot tubs.

A large part of this comes down to understanding how precise you
actually need to be when shooting.

Matt

On 2 Dec 2007, at 12:06, Hans Nyberg wrote:

> --- In PanoToolsNG@yahoogroups.com, Matthew Rogers <matthew@...>
> wrote:
> >
> > Hi Carel,
> >
> > I normally try and stay out of these debates but making such a broad
> > acceptable.
>
> Matt,
> Carel is in no way saying anything about the precision of the
> 360precision.
>
> He just says that there are other factors which can very easy change
> this precision.
>
> I just did a very simple test.
> My floor is standard 22mm plates with a carpet on and I placed my
> large heavy Manfrotto
> tripod 058 on it.
> I took a monopod and fastened it at a stable shelf 1,5 m from the
> tripod and moved it
> until it almost touched the top of the tripod vertical pole.
>
> The movement when walking around as you would normally do for taking
> the panorama
> was 1mm.
> That means that you get different pitches on the images and using
> template stitching will
> give you errors.
> And that has nothing to do with the panohead.
>
> You say the same yourself below so why are you complaining.
>
> Hans
>
• ... I don t see that as pessimistic as you. Although it won t help under certain circumstances, it might well help to use a self timer and walk away under
Message 17 of 26 , Dec 2, 2007
On Sunday, December 02, 2007 at 12:39, erik leeman wrote:

> Sorry, but that wouldn't help much because you'd have to approach and
> leave the camera for every single shot.

I don't see that as pessimistic as you. Although it won't help under
certain circumstances, it might well help to use a self timer and
walk away under other ones - f.e. in case of an elastic wooden floor.

And there certainly are other possibilities, like f.e. to weigh the
tripod down by hanging a heavy bag under it etc...

BTW.: I (and many others too) would appreciate if you quote what you
refer to. It makes reading of an answer much easier if the context is
http://www.netmeister.org/news/learn2quote.html

best regards

--
http://www.erik-krause.de
• ... Interesting list! My personal more extreme shooting environments so far have been the sea (meaning the shore, not the deeps), on rickety scaffolding
Message 18 of 26 , Dec 2, 2007
Sometime around 2/12/07 (at 12:54 +0000) Matthew Rogers said:

>I've shot rooms with carpet,
>tiles, floorboards, slate, glass, I've even shot in swimming pools and
>hot tubs.

Interesting list!
My personal more extreme shooting environments so far have been the
sea (meaning the shore, not the deeps), on rickety scaffolding
boards, deep mud, and in the middle of hard-core dance festival
crowds. At times those last two were together. And yes, the bottom of
my tripod legs *are* still just a bit muddy.
I've also shot from a paraglider, but no tripod or head was involved there. ;-)

I wonder what other extreme conditions people have been in when
shooting panos? Probably some would put my examples to shame!

Anyway, thanks to the head and a modicum of care, some of the pano
shots from festivals were exceptionally successful. One was used for
the cover of my recent CS3 Integration book.

My tripod is also solid, although I do wish I had a lighter one.
If the surface my tripod is on isn't solid, then I need to watch for
this and also be ready to deal with this afterwards when stitching.
Basic common sense.

I try to avoid introducing instability as I walk around the tripod,
but there are times when this can't be avoided entirely, solid
head/tripod or not. And that's what control points in PTGui help to
counteract.

k
• Ideally, you should NOT be trying to optimize your lens every single time you shoot. You should do your optimization one time, in an ideal environment. HDR is
Message 19 of 26 , Dec 2, 2007
Ideally, you should NOT be trying to optimize your lens every single time
you shoot.

You should do your optimization one time, in an ideal environment. HDR is
unnecessary for this step.

1 set of images around with excellently placed control points, will optimize

Then you only need to apply this template to ANY other shooting situation if
you're using the almighty precision head.

You only need 3-4 cps per image set for stitching, with an already optimized
lens. And I have always assumed.. 0 cps with the precision head.

One should only need to attempt cp's across and hdr set with an uncalibrated

When looking at your ptgui project, you'll find many cps that are not
optimized/used.

Sacha Griffin
Southern Digital Solutions LLC - Atlanta, Georgia
www.southern-digital.com
www.seeit360.net
www.ezphotosafe.com
404-551-4275
404-731-7798

I understand what you and others have said about the over and under
exposed images not having enough control points, however shouldnt the
first image of a bracketed set, the metered image, be able to
generate sufficient control points?

Would it make any difference to give the user the option to only use
control points from the metered shot whaterv order it is in ? eg. 0,-
,+ or -,0,+ etc just to take the 0 and use that as a model? Then
perhaps give the user the option to over ride the master image in
case like you say the image is deleted.

Thank you for your information! 3am time for bed.

KieranMullen

--- In PanoToolsNG@ <mailto:PanoToolsNG%40yahoogroups.com> yahoogroups.com,
"Erik Krause" <erik.krause@...>
wrote:
>
> On Saturday, December 01, 2007 at 3:27, verifone411 wrote:
>
> > Thus only the first image should get control points generated. The
> > images images do within the bracketed set do not need control
points
>
> Because in over- and underexposed areas there are no control points
> possible. I suspect one shoots HDR because there is a large dynamic
> range. If one side of the first image of a bracketed set is
> completely overexposed but the other isn't, PTGui needs to use the
> darker images instead to generate relevant CPs. But on the darker
one
> the other side might be underexposed, hence the control points from
> the first image are needed...
>
> You could argue that it would be sufficient to generate CPs in
other
> but the first image only if in the first one there weren't relevant
> ones. But if the user decides to unlink images after that, there
are
> no control points in lots of images. I guess that is why Joost
> decided to create CPs for all of them, even if they are linked...
>
> > from example 0 & 7 (why cant we start at 1?)
>
> For historical reasons: C programmers always start counting with 0.
> ;-)
>
> Best regards
> --
> http://www.erik- <http://www.erik-krause.de> krause.de
>

[Non-text portions of this message have been removed]
• ... Hi Matt, I made a simple observation that one should not try to figure out what is going on with mis-aligned control points by doing a test on a carpet.
Message 20 of 26 , Dec 2, 2007
Matthew Rogers-2 wrote:
>
> Hi Carel,
>
> I normally try and stay out of these debates but making such a broad
> acceptable. One thing I've learnt whilst dealing with thousands of
> photographers over the past 3-4 years is experience means nothing.
> Just because someone has been involved in panoramic photography since
> it's inception doesn't mean they actually know what they're talking
> about. In-fact it quite often means there so stuck in their ways that
> they're not willing to accept change.
>

Hi Matt,

I made a simple observation that one should not try to figure out what is
going on with mis-aligned control points by doing a test on a carpet. You
can use the universe's best pano head and it will still not make a
difference as the errors caused by movement of the carpet & floor will be
much bigger. It is not that hard to shoot a pano in such circumstances, but
any reasonably decent panohead will do. I have had many situations where I
wished I had a 360Precision, but a carpeted floor would not have been one of
those. The reason that I have not bought a 360Precision yet is that I use
two cameras and 3 lenses and often decide on the spot which one I will use.
It is not that I am stuck in my ways, as you opine, it's that I so often use
different setups, while the 360Precision only allows one.

Carel Struycken

--
View this message in context: http://www.nabble.com/How-control-points-are-determined-in-bracketed-shots-using-PTGUI-tf4926132.html#a14119009
Sent from the PanoToolsNG mailing list archive at Nabble.com.
• Hi Carel, Your email proves my point. We launched our fully adjustable precision head over 12 months ago, so why not buy one of those ? I just don t know why
Message 21 of 26 , Dec 2, 2007
Hi Carel,

precision head over 12 months ago, so why not buy one of those ? I
just don't know why you have so much trouble shooting panos on
carpet ? I can understand shooting on 10mm pile carpet but we ain't
living in the 70s.

Matt

> Hi Matt,
>
> I made a simple observation that one should not try to figure out
> what is
> going on with mis-aligned control points by doing a test on a
> carpet. You
> can use the universe's best pano head and it will still not make a
> difference as the errors caused by movement of the carpet & floor
> will be
> much bigger. It is not that hard to shoot a pano in such
> circumstances, but
> any reasonably decent panohead will do. I have had many situations
> where I
> wished I had a 360Precision, but a carpeted floor would not have
> been one of
> those. The reason that I have not bought a 360Precision yet is that
> I use
> two cameras and 3 lenses and often decide on the spot which one I
> will use.
> It is not that I am stuck in my ways, as you opine, it's that I so
> often use
> different setups, while the 360Precision only allows one.
>
> Carel Struycken
• ... Matt, Why do you keep mis-interpreting what I say? I have no trouble shooting panos on a carpet. I regularly shoot from a 6m hand held pole that sways in
Message 22 of 26 , Dec 2, 2007
Matthew Rogers-2 wrote:
>
> Hi Carel,
>
> precision head over 12 months ago, so why not buy one of those ? I
> just don't know why you have so much trouble shooting panos on
> carpet ? I can understand shooting on 10mm pile carpet but we ain't
> living in the 70s.
>
> Matt
>
>

Matt,

Why do you keep mis-interpreting what I say? I have no trouble shooting
panos on a carpet. I regularly shoot from a 6m hand held pole that sways in
the wind (another scenario where a 360Precision would not be of much use).
There are situations where I wish for a template-able panohead like the
360Precision, but nowhere on your website have I found information that the

Carel

--
View this message in context: http://www.nabble.com/How-control-points-are-determined-in-bracketed-shots-using-PTGUI-tf4933166.html#a14120732
Sent from the PanoToolsNG mailing list archive at Nabble.com.
• Prickles all round, eh? Personally, I find Carel s various posts to be very interesting (the Mt. Wilson Observatory object movie was wonderful!), and I find
Message 23 of 26 , Dec 2, 2007
Prickles all 'round, eh?

Personally, I find Carel's various posts to be very interesting (the
Mt. Wilson Observatory object movie was wonderful!), and I find
Matt's products to be superb. And you both give great advice. For me,
the disagreements and the rest aren't important. I just put it down
to the kind of free-thinking individualist that gets into panorama
creation... :-)

Carel, I'm about to test an Adjuste, as soon as I finish marking some
MA dissertations. I haven't tried it yet, but I have examined it
carefully. Yes, it should be every bit as template-able as the
Absolute. You'd just need to note down or mark the right positions
for your different combinations of camera/lens equipment.
Here's a pre-test assessment if you want one: It is a serious
investment to make, but I reckon it'd let you have your cake and eat
it too.

But don't go using it on trampolines, ok? (sorry, silly joke ;-)

k
--

Keith Martin
Author, Creative Suite 3 Integration (Focal Press)
Senior Lecturer in Publishing, LCC (University of the Arts, London)
Technical Editor, MacUser magazine
design, programming, VR photography, consultancy

http://www.thesmallest.com
iChat/AIM & Skype ThatKeith
• For those having issues with panos on carpet and flexing wooden floors.. Is it a matter of not finding any auto stich points? Are you use ing PTGUi? if so does
Message 24 of 26 , Dec 4, 2007
For those having issues with panos on carpet and flexing wooden floors..

Is it a matter of not finding any auto stich points?

Are you use ing PTGUi? if so does it rate it ias very good, good, bad

Thank you

Kieran
• ... Yes, of course. That are those between images inside a linked set - as long as this set stays linked. They get used as soon as you unlink the images and
Message 25 of 26 , Dec 4, 2007
On Sunday, December 02, 2007 at 12:47, Sacha Griffin wrote:

> When looking at your ptgui project, you'll find many cps that are not
> optimized/used.

Yes, of course. That are those between images inside a linked set -
as long as this set stays linked. They get used as soon as you unlink
the images and optimize again. This might be necessary sometimes if
you discover that there are HDR merging errors between those
images...

best regards
--
http://www.erik-krause.de
• I think the reference to carpets and flexing wooden floors, is regarding using a fixed PTGui template and pin registered panohead giving the (possible) ability
Message 26 of 26 , Dec 4, 2007
I think the reference to carpets and flexing wooden floors, is regarding
using a fixed PTGui template and pin registered panohead giving the
(possible) ability to not need optimising.
ie. Load images into a pre-prepared ptgui template and just stitch. So, no
cp's and no optimising.
(you 'do' have to follow a fairly rigourous set of repeatable rules for this
to work well)

In general carpet and wooden floors etc. are not a problem when you are
going to insert cp's and optimise.
(unless there is something 'very' close to the camera :-)

Cheers,
Darren.

)-----Original Message-----
)From: verifone411
)
)For those having issues with panos on carpet and flexing
)wooden floors..
)
)Is it a matter of not finding any auto stich points?
)
)Are you use ing PTGUi? if so does it rate it ias very good,