"Benny" <no spam ple@se.com> wrote in message
news:ijUvj.18311$421.1251@news-server.bigpond.net.au...
>I have created a panorama using about 6 images all taken using a 17mm lense
>(with a Canon 40D).
>
> At a later date I went and took some more shots (from the same
> position/location) but with focal lengths of 35mm and 50mm.
> I now want to add these extra images into the original panorama without
> them looking disproportionate or appearing to have better or worse
> resolution.
>
> What is the math I should be applying to ensure I get the extra images at
> the same proportions as the panorama and also preferably at the same
> printable resolution?
>
> My idea is that the 50mm images will need to be reduced in size to 17/50
> (ie 0.34) to fit correctly into the panorama. The 35mm images would be
> reduced to 17/35 (ie about half) to fit into the panorama etc etc.
>
> Not too sure what to do to ensure a consistency in printable resolution by
> mixing images (of differing focal lengths) in this way.

The focal length is not an issue - what matters is that you took the images
from the same location. You are correct about the change in image size.

One consequence of this is that you will need to double (or triple) up in
the vertical direction to match the field of view of your original pano, and
there will be keystoning issues because the vertical axis has changed. To
deal with that, you could get by with the transform tool, but a tool such as
PTGUI might save you a lot of time, since it known how to align images in
spite of this rectilinear distortion.

Another issue that you will almost certainly need to deal with will be color
matching the sky and other objects in the panorama. One way to do this is
with curves, and numeric values from the info palette. Again, a panorama
savvy program may save you some effort here, though you can do this
perfectly well manually with a little effort.
--
Mike Russell - www.curvemeister.com

I have created a panorama using about 6 images all taken using a 17mm lense
(with a Canon 40D).

At a later date I went and took some more shots (from the same
position/location) but with focal lengths of 35mm and 50mm.
I now want to add these extra images into the original panorama without them
looking disproportionate or appearing to have better or worse resolution.

What is the math I should be applying to ensure I get the extra images at
the same proportions as the panorama and also preferably at the same
printable resolution?

My idea is that the 50mm images will need to be reduced in size to 17/50 (ie
0.34) to fit correctly into the panorama. The 35mm images would be reduced
to 17/35 (ie about half) to fit into the panorama etc etc.

Not too sure what to do to ensure a consistency in printable resolution by
mixing images (of differing focal lengths) in this way.

thanks

Benny

On Feb 23, 6:27 am, "Benny" <no spam p...@se.com> wrote:
> What is the math I should be applying to ensure I get the extra images at
> the same proportions as the panorama and also preferably at the same
> printable resolution?
> Benny

Hi Benny,
One possibility would be to 'place' each image as a smart object into
a PS document set to the image size you want with a canvas size large
enough to facilitate the largest. Align them. Then starting at the
next to the bottom layer use the transform tools and the shift key
(for constraint) to match the image sizes. Smart objects
transformations are nondestructive so you get as many shots as you
want. You can set the working layer to difference mode so you can tell
when the image turns black that you have it registered. After you get
them sized set the guides to percent and snap a guide to each edge.
Note the percent then drill down into the smart object and group the
document into another smart object then add guides to those locations
and snap the document size to the guides. This is a nondestructive way
to get all your documents to match perspective.

Good luck,
Ron

<ronviers@gmail.com> wrote in message
news:c1a1497b-4e51-4f67-9445-e65942d01bc4@71g2000hse.googlegroups.com...
> On Feb 23, 6:27 am, "Benny" <no spam p...@se.com> wrote:
>> What is the math I should be applying to ensure I get the extra images at
>> the same proportions as the panorama and also preferably at the same
>> printable resolution?
>> Benny
>
> Hi Benny,
> One possibility would be to 'place' each image as a smart object into
> a PS document set to the image size you want with a canvas size large
> enough to facilitate the largest. Align them. Then starting at the
> next to the bottom layer use the transform tools and the shift key
> (for constraint) to match the image sizes. Smart objects
> transformations are nondestructive so you get as many shots as you
> want. You can set the working layer to difference mode so you can tell
> when the image turns black that you have it registered. After you get
> them sized set the guides to percent and snap a guide to each edge.
> Note the percent then drill down into the smart object and group the
> document into another smart object then add guides to those locations
> and snap the document size to the guides. This is a nondestructive way
> to get all your documents to match perspective.
>
> Good luck,
> Ron

Thanks Ron
I will digest your comments and get back if I need further clarification.
regards
Benny

>> My idea is that the 50mm images will need to be reduced in size to 17/50
>> (ie 0.34) to fit correctly into the panorama. The 35mm images would be
>> reduced to 17/35 (ie about half) to fit into the panorama etc etc.
>>
>> Not too sure what to do to ensure a consistency in printable resolution
>> by mixing images (of differing focal lengths) in this way.
>
> The focal length is not an issue - what matters is that you took the
> images from the same location. You are correct about the change in image
> size.
>
> One consequence of this is that you will need to double (or triple) up in
> the vertical direction to match the field of view of your original pano,
> and there will be keystoning issues because the vertical axis has changed.
> To deal with that, you could get by with the transform tool, but a tool
> such as PTGUI might save you a lot of time, since it known how to align
> images in spite of this rectilinear distortion.
>
> Another issue that you will almost certainly need to deal with will be
> color matching the sky and other objects in the panorama. One way to do
> this is with curves, and numeric values from the info palette. Again, a
> panorama savvy program may save you some effort here, though you can do
> this perfectly well manually with a little effort.
> --
> Mike Russell - www.curvemeister.com
>

Thanks Mike
I actually used PTGUI (for the first time) to create the pano. I will now
look further into it's other capabilities.
As an aside, I also created an equivalent pano using CS3's Photomerge and
found some interesting differences in image and file sizes etc, as noted
below.
PTGUI produced an image size of 3490 x 1147 pixels @ 72 ppi and a saved
filesize of 4880 KB although in CS3 the doc size and pixel dimensions are
shown as 11.5MB.
CS3's Photomerge produced an image size of 12681 x 4253 pixels @ 72ppi and a
saved filesize of only 1400KB. CS3 indicates a doc size and pixel dims of
154.3MB. The image is about 3.6 times larger but the file is about 3.5 times
smaller???
I will need to investigate further to sort these figure differences out.
In CS3 when I zoom to fill window to view them both at the same screen size
I notice that PTGUI zoom is 26.25% and CS3's is only 7.22%. I also need to
zoom in around 300% for the PTGUI pano to obtain the same size as the 100%
CS3 zoom. The PTGUI pano pixelates a loot sooner than the CS3 pano. My
leaning is therefore towards the CS3 pano because of this. But it may just
be a setting in PTGUI that will resolve this issue.
I presume with the extra images I want to insert into the pano, I will need
to maintain the same pixel resolution (or better) as that of the pano?
regards
Benny

"Benny" <no spam ple@se.com> wrote in message
news:IV2wj.18419$421.7234@news-server.bigpond.net.au...
>>> My idea is that the 50mm images will need to be reduced in size to 17/50
>>> (ie 0.34) to fit correctly into the panorama. The 35mm images would be
>>> reduced to 17/35 (ie about half) to fit into the panorama etc etc.
>>>
>>> Not too sure what to do to ensure a consistency in printable resolution
>>> by mixing images (of differing focal lengths) in this way.
>>
>> The focal length is not an issue - what matters is that you took the
>> images from the same location. You are correct about the change in image
>> size.
>>
>> One consequence of this is that you will need to double (or triple) up in
>> the vertical direction to match the field of view of your original pano,
>> and there will be keystoning issues because the vertical axis has
>> changed. To deal with that, you could get by with the transform tool, but
>> a tool such as PTGUI might save you a lot of time, since it known how to
>> align images in spite of this rectilinear distortion.
>>
>> Another issue that you will almost certainly need to deal with will be
>> color matching the sky and other objects in the panorama. One way to do
>> this is with curves, and numeric values from the info palette. Again, a
>> panorama savvy program may save you some effort here, though you can do
>> this perfectly well manually with a little effort.
>> --
>> Mike Russell - www.curvemeister.com
>>
>
> Thanks Mike
> I actually used PTGUI (for the first time) to create the pano. I will now
> look further into it's other capabilities.
> As an aside, I also created an equivalent pano using CS3's Photomerge and
> found some interesting differences in image and file sizes etc, as noted
> below.
> PTGUI produced an image size of 3490 x 1147 pixels @ 72 ppi and a saved
> filesize of 4880 KB although in CS3 the doc size and pixel dimensions are
> shown as 11.5MB.
> CS3's Photomerge produced an image size of 12681 x 4253 pixels @ 72ppi and
> a saved filesize of only 1400KB. CS3 indicates a doc size and pixel dims
> of 154.3MB. The image is about 3.6 times larger but the file is about 3.5
> times smaller???
> I will need to investigate further to sort these figure differences out.
> In CS3 when I zoom to fill window to view them both at the same screen
> size I notice that PTGUI zoom is 26.25% and CS3's is only 7.22%. I also
> need to zoom in around 300% for the PTGUI pano to obtain the same size as
> the 100% CS3 zoom. The PTGUI pano pixelates a loot sooner than the CS3
> pano. My leaning is therefore towards the CS3 pano because of this. But it
> may just be a setting in PTGUI that will resolve this issue.
> I presume with the extra images I want to insert into the pano, I will
> need to maintain the same pixel resolution (or better) as that of the
> pano?
> regards
> Benny
>

Autopano is worth checking out. I also use PTGui and CS3 photomerge. It can
be a lot of work, but who wants to sleep anyway. You can do that when you
are dead.

r.

>>> The focal length is not an issue - what matters is that you took the
>>> images from the same location. You are correct about the change in image
>>> size.
>>>
>>> Mike Russell - www.curvemeister.com

Say WHAT ???

You may know curves, but you do not know optics!

Differing focal lengths will alter the visual relationship between objects
not to mention barrel and other lens distortions.

Technically you can make the pan, but the eye will wonder "What's WRONG?".

For visually good pans, you should shoot in portrait (vertical) mode, lens
of 55mm or better, overlap of 15% and USE A TRIPOD with the cameral level!

From: "Johan W. Elzenga" <nomail@please.invalid>
....
> Once again, you may read this (which you 'conveniently' deleted when
> quoting me): http://www.dpreview.com/learn/?/key=perspective


LOL - Sir F. A. Rien, le nom dit tous: il ne ce fait pas.
---
Mike Russell - www.curvemeister.com