Pano head

Discussion in 'Digital Photography' started by l e o, Sep 4, 2005.

  1. l e o

    l e o Guest

    What is the best one? And what is a reasonable priced one? How does the
    Kaidan kiWi perform? I am using it on 20D. Also, how does one determine
    the nodal point of a lens?
    l e o, Sep 4, 2005
    #1
    1. Advertising

  2. In article <gEBSe.4103$>, says...
    > What is the best one? And what is a reasonable priced one? How does the
    > Kaidan kiWi perform? I am using it on 20D. Also, how does one determine
    > the nodal point of a lens?
    >

    Kaidan works well but is expensive. You can get by with a bubble level
    and a regular tripod. Nodal point is not really an issue if you are
    shooting objects more than about 10 feet away. The displacement of an
    inch or two won't show up.
    A google search of nodal point will give you a few sites that show how
    to find it. If you do this you mount the camera on a slidebar and offset
    from the center of the tripod as the pivot point. The center of the lens
    is close enough, though.

    --
    Robert D Feinman
    Landscapes, Cityscapes and Panoramic Photographs
    http://robertdfeinman.com
    mail:
    Robert Feinman, Sep 4, 2005
    #2
    1. Advertising

  3. l e o

    l e o Guest

    Robert Feinman wrote:
    > In article <gEBSe.4103$>, says...
    >
    >>What is the best one? And what is a reasonable priced one? How does the
    >>Kaidan kiWi perform? I am using it on 20D. Also, how does one determine
    >>the nodal point of a lens?
    >>

    >
    > Kaidan works well but is expensive. You can get by with a bubble level
    > and a regular tripod. Nodal point is not really an issue if you are
    > shooting objects more than about 10 feet away. The displacement of an
    > inch or two won't show up.
    > A google search of nodal point will give you a few sites that show how
    > to find it. If you do this you mount the camera on a slidebar and offset
    > from the center of the tripod as the pivot point. The center of the lens
    > is close enough, though.



    I'll be shooting interior so a proper tool would be needed and I also
    would like to pivot the camera vertically.
    l e o, Sep 4, 2005
    #3
  4. l e o <> wrote:

    : I'll be shooting interior so a proper tool would be needed and I also
    : would like to pivot the camera vertically.

    This is probably not the best setup, but here is what I am in the process
    of puting together. I found a "panorama plate" at B&H that screws to the
    tripod (1/4-20) screw and has a camera screw on the top and is graduated
    in degrees. On top of this I am attaching a "monopod tilt head". This way
    I can level the tripod (with its built in bullseye level) which will make
    sure the pano plate is on a level plane. Then the tilt head will allow me
    to adjust the location with respect to the horizon that the pano strip
    will take. So all I have to do is calculate the angle of the field of view
    to know how many degrees to turn the pano plate with each image (with a
    certain percentage of overlap).

    As I said I have not completed the setup yet and so can not attest to the
    effectiveness of the idea, but this may give an idea of how to make what
    it seems you want.

    Randy

    ==========
    Randy Berbaum
    Champaign, IL
    Randy Berbaum, Sep 5, 2005
    #4
  5. "l e o" <> wrote in message
    news:VQCSe.14699$...
    > Robert Feinman wrote:

    SNIP
    > I'll be shooting interior so a proper tool would be needed and I
    > also would like to pivot the camera vertically.


    In that case:
    <http://www.manfrotto.com/Jahia/cache/offonce/pid/2356;jsessionid=4B388A8CD41C1B95DDBE0C30E682D047?livid=liv2213>

    Bart
    Bart van der Wolf, Sep 5, 2005
    #5
  6. l e o

    Bill Funk Guest

    On Mon, 5 Sep 2005 03:41:53 +0000 (UTC), Randy Berbaum
    <> wrote:

    >l e o <> wrote:
    >
    >: I'll be shooting interior so a proper tool would be needed and I also
    >: would like to pivot the camera vertically.
    >
    >This is probably not the best setup, but here is what I am in the process
    >of puting together. I found a "panorama plate" at B&H that screws to the
    >tripod (1/4-20) screw and has a camera screw on the top and is graduated
    >in degrees. On top of this I am attaching a "monopod tilt head". This way
    >I can level the tripod (with its built in bullseye level) which will make
    >sure the pano plate is on a level plane. Then the tilt head will allow me
    >to adjust the location with respect to the horizon that the pano strip
    >will take. So all I have to do is calculate the angle of the field of view
    >to know how many degrees to turn the pano plate with each image (with a
    >certain percentage of overlap).
    >
    >As I said I have not completed the setup yet and so can not attest to the
    >effectiveness of the idea, but this may give an idea of how to make what
    >it seems you want.
    >
    >Randy
    >
    >==========
    >Randy Berbaum
    >Champaign, IL


    Maybe I'm missing something, so if I am, let me know.
    Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 5, 2005
    #6
  7. "Bill Funk" <> wrote in message
    news:...
    SNIP
    > Maybe I'm missing something, so if I am, let me know.
    > Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?


    If the rotations are not done on the single same point in 3D, then
    foreground or background will not align for stitching. The overlap
    only serves to identify common points in multiple images.

    Bart
    Bart van der Wolf, Sep 5, 2005
    #7
  8. l e o

    Bill Funk Guest

    On Mon, 5 Sep 2005 21:31:18 +0200, "Bart van der Wolf"
    <> wrote:

    >
    >"Bill Funk" <> wrote in message
    >news:...
    >SNIP
    >> Maybe I'm missing something, so if I am, let me know.
    >> Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?

    >
    >If the rotations are not done on the single same point in 3D, then
    >foreground or background will not align for stitching. The overlap
    >only serves to identify common points in multiple images.
    >
    >Bart


    For close-in objects, yes. And that's not the point I was asking
    about.
    I was asking about this:
    "This is probably not the best setup, but here is what I am in the
    process of puting together. I found a "panorama plate" at B&H that
    screws to the tripod (1/4-20) screw and has a camera screw on the top
    and is graduated in degrees. On top of this I am attaching a "monopod
    tilt head". This way I can level the tripod (with its built in
    bullseye level) which will make sure the pano plate is on a level
    plane. Then the tilt head will allow me to adjust the location with
    respect to the horizon that the pano strip will take. So all I have to
    do is calculate the angle of the field of view to know how many
    degrees to turn the pano plate with each image (with a certain
    percentage of overlap). "

    The part about knowing how many degrees to turn the pano head seems
    too complex for the task.
    Thus the question, isn't an overlap of 1/2 to 2/3 accurate sufficient?

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 5, 2005
    #8
  9. l e o

    Ken Weitzel Guest

    Bill Funk wrote:

    > On Mon, 5 Sep 2005 21:31:18 +0200, "Bart van der Wolf"
    > <> wrote:
    >
    >
    >>"Bill Funk" <> wrote in message
    >>news:...
    >>SNIP
    >>
    >>>Maybe I'm missing something, so if I am, let me know.
    >>>Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?

    >>
    >>If the rotations are not done on the single same point in 3D, then
    >>foreground or background will not align for stitching. The overlap
    >>only serves to identify common points in multiple images.
    >>
    >>Bart

    >
    >
    > For close-in objects, yes. And that's not the point I was asking
    > about.
    > I was asking about this:
    > "This is probably not the best setup, but here is what I am in the
    > process of puting together. I found a "panorama plate" at B&H that
    > screws to the tripod (1/4-20) screw and has a camera screw on the top
    > and is graduated in degrees. On top of this I am attaching a "monopod
    > tilt head". This way I can level the tripod (with its built in
    > bullseye level) which will make sure the pano plate is on a level
    > plane. Then the tilt head will allow me to adjust the location with
    > respect to the horizon that the pano strip will take. So all I have to
    > do is calculate the angle of the field of view to know how many
    > degrees to turn the pano plate with each image (with a certain
    > percentage of overlap). "
    >
    > The part about knowing how many degrees to turn the pano head seems
    > too complex for the task.
    > Thus the question, isn't an overlap of 1/2 to 2/3 accurate sufficient?


    Hi..

    If it's any help... I use Oly's; with their built-in
    panorama feature.

    Just had a look through one of the (evf) viewfinders, and
    it looks to my eye that it's 'wasting' less than a
    quarter of the frame. (on each side, or top and bottom)

    If you like I'll be happy to take a pic of something with
    a tiny something right on the line and measure the pixel
    ratio for you; just holler.

    Take care.

    Ken
    Ken Weitzel, Sep 5, 2005
    #9
  10. l e o

    Bill Funk Guest

    On Mon, 05 Sep 2005 20:25:09 GMT, Ken Weitzel <>
    wrote:

    >
    >
    >Bill Funk wrote:
    >
    >> On Mon, 5 Sep 2005 21:31:18 +0200, "Bart van der Wolf"
    >> <> wrote:
    >>
    >>
    >>>"Bill Funk" <> wrote in message
    >>>news:...
    >>>SNIP
    >>>
    >>>>Maybe I'm missing something, so if I am, let me know.
    >>>>Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?
    >>>
    >>>If the rotations are not done on the single same point in 3D, then
    >>>foreground or background will not align for stitching. The overlap
    >>>only serves to identify common points in multiple images.
    >>>
    >>>Bart

    >>
    >>
    >> For close-in objects, yes. And that's not the point I was asking
    >> about.
    >> I was asking about this:
    >> "This is probably not the best setup, but here is what I am in the
    >> process of puting together. I found a "panorama plate" at B&H that
    >> screws to the tripod (1/4-20) screw and has a camera screw on the top
    >> and is graduated in degrees. On top of this I am attaching a "monopod
    >> tilt head". This way I can level the tripod (with its built in
    >> bullseye level) which will make sure the pano plate is on a level
    >> plane. Then the tilt head will allow me to adjust the location with
    >> respect to the horizon that the pano strip will take. So all I have to
    >> do is calculate the angle of the field of view to know how many
    >> degrees to turn the pano plate with each image (with a certain
    >> percentage of overlap). "
    >>
    >> The part about knowing how many degrees to turn the pano head seems
    >> too complex for the task.
    >> Thus the question, isn't an overlap of 1/2 to 2/3 accurate sufficient?

    >
    >Hi..
    >
    >If it's any help... I use Oly's; with their built-in
    >panorama feature.
    >
    >Just had a look through one of the (evf) viewfinders, and
    >it looks to my eye that it's 'wasting' less than a
    >quarter of the frame. (on each side, or top and bottom)
    >
    >If you like I'll be happy to take a pic of something with
    >a tiny something right on the line and measure the pixel
    >ratio for you; just holler.
    >
    >Take care.
    >
    >Ken


    Well, OK, if you would...
    The question, though, is this:
    For a panorama, is it really necessary to determine how many degrees
    to turn the pano head?
    Isn't 1/3 to 1/2 overlap sufficiently vague and serviceable enough?
    Is anything gained by making the extra effort?
    Am I not getting the question over well?

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #10
  11. l e o

    BC Guest

    Bill Funk wrote:
    >
    > Well, OK, if you would...
    > The question, though, is this:
    > For a panorama, is it really necessary to determine how many degrees
    > to turn the pano head?
    > Isn't 1/3 to 1/2 overlap sufficiently vague and serviceable enough?
    > Is anything gained by making the extra effort?
    > Am I not getting the question over well?
    >


    For single-row panoramas with relatively few images you can probably
    get away without knowing the angular coordinates of each shot.

    However, for multirow panoramas with lots of images knowing the
    approximate values of pitch/yaw/roll for each image can make all the
    difference in terms of stitching accuracy. I find this to be
    particularly true when using long focal length lenses. As a result, I
    use only two or three different lenses for shooting panoramas, and I
    always keep track of the pitch and yaw angles for every shot.

    Bear in mind that my experience is almost entirely with Panorama Tools.
    Other software may act differently.

    Brian
    BC, Sep 6, 2005
    #11
  12. l e o

    Ken Weitzel Guest

    Bill Funk wrote:
    > On Mon, 05 Sep 2005 20:25:09 GMT, Ken Weitzel <>
    > wrote:
    >
    >
    >>
    >>Bill Funk wrote:
    >>
    >>
    >>>On Mon, 5 Sep 2005 21:31:18 +0200, "Bart van der Wolf"
    >>><> wrote:
    >>>
    >>>
    >>>
    >>>>"Bill Funk" <> wrote in message
    >>>>news:...
    >>>>SNIP
    >>>>
    >>>>
    >>>>>Maybe I'm missing something, so if I am, let me know.
    >>>>>Is all this necessary? Isn't an overlap of 1/3 to 1/2 sufficient?
    >>>>
    >>>>If the rotations are not done on the single same point in 3D, then
    >>>>foreground or background will not align for stitching. The overlap
    >>>>only serves to identify common points in multiple images.
    >>>>
    >>>>Bart
    >>>
    >>>
    >>>For close-in objects, yes. And that's not the point I was asking
    >>>about.
    >>>I was asking about this:
    >>>"This is probably not the best setup, but here is what I am in the
    >>>process of puting together. I found a "panorama plate" at B&H that
    >>>screws to the tripod (1/4-20) screw and has a camera screw on the top
    >>>and is graduated in degrees. On top of this I am attaching a "monopod
    >>>tilt head". This way I can level the tripod (with its built in
    >>>bullseye level) which will make sure the pano plate is on a level
    >>>plane. Then the tilt head will allow me to adjust the location with
    >>>respect to the horizon that the pano strip will take. So all I have to
    >>>do is calculate the angle of the field of view to know how many
    >>>degrees to turn the pano plate with each image (with a certain
    >>>percentage of overlap). "
    >>>
    >>>The part about knowing how many degrees to turn the pano head seems
    >>>too complex for the task.
    >>>Thus the question, isn't an overlap of 1/2 to 2/3 accurate sufficient?

    >>
    >>Hi..
    >>
    >>If it's any help... I use Oly's; with their built-in
    >>panorama feature.
    >>
    >>Just had a look through one of the (evf) viewfinders, and
    >>it looks to my eye that it's 'wasting' less than a
    >>quarter of the frame. (on each side, or top and bottom)
    >>
    >>If you like I'll be happy to take a pic of something with
    >>a tiny something right on the line and measure the pixel
    >>ratio for you; just holler.
    >>
    >>Take care.
    >>
    >>Ken

    >
    >
    > Well, OK, if you would...
    > The question, though, is this:
    > For a panorama, is it really necessary to determine how many degrees
    > to turn the pano head?
    > Isn't 1/3 to 1/2 overlap sufficiently vague and serviceable enough?
    > Is anything gained by making the extra effort?
    > Am I not getting the question over well?


    Hi Bill...

    Did the experiment; here's how.

    First, the Oly display draws an H shaped set of lines on the
    display in Pano mode. Shooting left to right horizontally,
    the left vertical of the H is - "waste" or overlap to the left
    of it, and usable picture to the right.

    So, I ointed the camera best I could at the rec room wall,
    and lined up the left side of a photo frame with the left
    vertical of the picture frame. Shot it.

    Downloaded it. The whole image was 1600 pixels wide.
    Cropped the picture frame and removed it, leaving only
    wall (the waste) It was then 265 pixels wide.
    Keep in mind, that's difficult to do, so not precise.

    In Oly's manual, which I haven't looked at in ages,
    they say (fair use claimed) "Overlap the left end of the picture
    with the right end of the previous picture (when shooting
    to the right)"

    Perhaps I shouldn't use the word waste. Their demo
    sequence of shots shows a single tree in the waste
    right hand space of the first shot. Then they've
    composed the next shot so that the same tree is in
    the waste left hand space of the next.

    In other words, the single tree isn't in either of the
    pictures (inside the H), but does appear in the final
    picture :)

    Hope this helps.

    Take care.

    Ken
    Ken Weitzel, Sep 6, 2005
    #12
  13. l e o

    Bill Funk Guest

    On 5 Sep 2005 16:57:59 -0700, "BC" <> wrote:

    >
    >Bill Funk wrote:
    >>
    >> Well, OK, if you would...
    >> The question, though, is this:
    >> For a panorama, is it really necessary to determine how many degrees
    >> to turn the pano head?
    >> Isn't 1/3 to 1/2 overlap sufficiently vague and serviceable enough?
    >> Is anything gained by making the extra effort?
    >> Am I not getting the question over well?
    >>

    >
    >For single-row panoramas with relatively few images you can probably
    >get away without knowing the angular coordinates of each shot.
    >
    >However, for multirow panoramas with lots of images knowing the
    >approximate values of pitch/yaw/roll for each image can make all the
    >difference in terms of stitching accuracy. I find this to be
    >particularly true when using long focal length lenses. As a result, I
    >use only two or three different lenses for shooting panoramas, and I
    >always keep track of the pitch and yaw angles for every shot.
    >
    >Bear in mind that my experience is almost entirely with Panorama Tools.
    > Other software may act differently.
    >
    >Brian


    OK, I see what you're saying.
    So you take your first row, and then pitch up enough to overlap 1/2 to
    1/3, then yaw as you take the next row. Repeat as necessary.
    What am I missing? Where's the roll?

    Maybe my laziness and "form follows function" mindset is hiding
    something from me, but it seems that this is being made much harder
    than it needs to be.

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #13
  14. l e o

    Bill Funk Guest

    On Tue, 06 Sep 2005 00:06:43 GMT, Ken Weitzel <>
    wrote:

    >> Well, OK, if you would...
    >> The question, though, is this:
    >> For a panorama, is it really necessary to determine how many degrees
    >> to turn the pano head?
    >> Isn't 1/3 to 1/2 overlap sufficiently vague and serviceable enough?
    >> Is anything gained by making the extra effort?
    >> Am I not getting the question over well?

    >
    >Hi Bill...
    >
    >Did the experiment; here's how.
    >
    >First, the Oly display draws an H shaped set of lines on the
    >display in Pano mode. Shooting left to right horizontally,
    >the left vertical of the H is - "waste" or overlap to the left
    >of it, and usable picture to the right.
    >
    >So, I ointed the camera best I could at the rec room wall,
    >and lined up the left side of a photo frame with the left
    >vertical of the picture frame. Shot it.
    >
    >Downloaded it. The whole image was 1600 pixels wide.
    >Cropped the picture frame and removed it, leaving only
    >wall (the waste) It was then 265 pixels wide.
    >Keep in mind, that's difficult to do, so not precise.
    >
    >In Oly's manual, which I haven't looked at in ages,
    >they say (fair use claimed) "Overlap the left end of the picture
    >with the right end of the previous picture (when shooting
    >to the right)"
    >
    >Perhaps I shouldn't use the word waste. Their demo
    >sequence of shots shows a single tree in the waste
    >right hand space of the first shot. Then they've
    >composed the next shot so that the same tree is in
    >the waste left hand space of the next.
    >
    >In other words, the single tree isn't in either of the
    >pictures (inside the H), but does appear in the final
    >picture :)
    >
    >Hope this helps.
    >
    >Take care.
    >
    >Ken
    >

    That "waste" is overlap. No need to remove it, the pano software will
    handle that.
    Overlap is needed; it's how the software knows where to stitch and
    align the different shots.
    This still doesn't answer why the exact amount of overlap is
    important.

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #14
  15. l e o

    Paul Mitchum Guest

    Bill Funk <> wrote:

    > On 5 Sep 2005 16:57:59 -0700, "BC" <> wrote:

    [..]
    > >For single-row panoramas with relatively few images you can probably get
    > >away without knowing the angular coordinates of each shot.
    > >
    > >However, for multirow panoramas with lots of images knowing the
    > >approximate values of pitch/yaw/roll for each image can make all the
    > >difference in terms of stitching accuracy. I find this to be
    > >particularly true when using long focal length lenses. As a result, I
    > >use only two or three different lenses for shooting panoramas, and I
    > >always keep track of the pitch and yaw angles for every shot.
    > >
    > >Bear in mind that my experience is almost entirely with Panorama Tools.
    > >Other software may act differently.

    >
    > OK, I see what you're saying.
    > So you take your first row, and then pitch up enough to overlap 1/2 to
    > 1/3, then yaw as you take the next row. Repeat as necessary.
    > What am I missing? Where's the roll?
    >
    > Maybe my laziness and "form follows function" mindset is hiding
    > something from me, but it seems that this is being made much harder
    > than it needs to be.


    You *can* take a bunch of pictures and then stitch them together however
    you want. It's just easier to get things started if you're methodical,
    and already know the orientation of each picture.

    You could even dump a bunch of pictures into something like autopano and
    let it make all the decisions for you and get reasonable results.

    As far as the roll: If you have a steady pano head on a level tripod,
    there won't be any.
    Paul Mitchum, Sep 6, 2005
    #15
  16. l e o

    ASAAR Guest

    On Mon, 05 Sep 2005 19:44:55 -0700, Bill Funk wrote:

    > Overlap is needed; it's how the software knows where to stitch and
    > align the different shots.
    > This still doesn't answer why the exact amount of overlap is
    > important.


    It may not be for single rows. For multiple rows the overlap
    shouldn't have to be exact, but the more precise the better, because
    errors tend to be additive. If you start taking shots left to
    right, when it's time to stitch them together, the leftmost column
    will be aligned pretty well, but as you progress to the right, the
    alignment errors might get large enough so that image N+1 is
    practically above image N + 2 in the row below. I don't see that
    this would create insurmountable problems but it might make
    assembling the panorama a bit trickier. Or not. Or there could be
    a different reason why precision is preferred. One being that with
    less "waste" you might be able to get by with fewer images to deal
    with. The fewer the better as that allows all of the shots to be
    taken in less time, and for shots taken outside, that gives the sun
    less time to move, less time for the weather to change, and less
    time for someone from Homeland Security to approach and ask what
    you're doing and explain why you're there taking pictures of that
    particular resevoir, etc.
    ASAAR, Sep 6, 2005
    #16
  17. l e o

    Bill Funk Guest

    On Mon, 5 Sep 2005 20:36:06 -0700, 0m (Paul Mitchum)
    wrote:

    >> OK, I see what you're saying.
    >> So you take your first row, and then pitch up enough to overlap 1/2 to
    >> 1/3, then yaw as you take the next row. Repeat as necessary.
    >> What am I missing? Where's the roll?
    >>
    >> Maybe my laziness and "form follows function" mindset is hiding
    >> something from me, but it seems that this is being made much harder
    >> than it needs to be.

    >
    >You *can* take a bunch of pictures and then stitch them together however
    >you want. It's just easier to get things started if you're methodical,
    >and already know the orientation of each picture.


    Maybe I'm just doing it a way that works.
    When I take a multi-row pano (and they are only two rows for me), I
    use my tripod, level it, and start taking shots from the left,
    overlapping by about 1/3 for each successive shot. When I reach the
    end of that row, I pitch the camera up to overlap the lower pic by
    about 1/3, and proceed to take shots to the left, overlapping by about
    1/3.
    Maybe I'm just making it sound easier than it is. I don't know. But I
    see no need to measure anything except by the Mk I eyeball.
    >
    >You could even dump a bunch of pictures into something like autopano and
    >let it make all the decisions for you and get reasonable results.


    You get better results by measuring than by eyeballing, if the overlap
    is sufficient by eyeballing?
    >
    >As far as the roll: If you have a steady pano head on a level tripod,
    >there won't be any.


    I know; I didn't bring up roll. Paul did.

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #17
  18. l e o

    Bill Funk Guest

    On Tue, 06 Sep 2005 01:21:31 -0400, ASAAR <> wrote:

    >On Mon, 05 Sep 2005 19:44:55 -0700, Bill Funk wrote:
    >
    >> Overlap is needed; it's how the software knows where to stitch and
    >> align the different shots.
    >> This still doesn't answer why the exact amount of overlap is
    >> important.

    >
    > It may not be for single rows. For multiple rows the overlap
    >shouldn't have to be exact, but the more precise the better, because
    >errors tend to be additive. If you start taking shots left to
    >right, when it's time to stitch them together, the leftmost column
    >will be aligned pretty well, but as you progress to the right, the
    >alignment errors might get large enough so that image N+1 is
    >practically above image N + 2 in the row below. I don't see that
    >this would create insurmountable problems but it might make
    >assembling the panorama a bit trickier. Or not. Or there could be
    >a different reason why precision is preferred. One being that with
    >less "waste" you might be able to get by with fewer images to deal
    >with. The fewer the better as that allows all of the shots to be
    >taken in less time, and for shots taken outside, that gives the sun
    >less time to move, less time for the weather to change, and less
    >time for someone from Homeland Security to approach and ask what
    >you're doing and explain why you're there taking pictures of that
    >particular resevoir, etc.


    Maybe I just don't put enopugh shots in my panos for this to matter.
    In mine, I always end up with the same number of shots in the two
    rows.
    Maybe I just have a good eye for it.

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #18
  19. l e o

    Paul Mitchum Guest

    Bill Funk <> wrote:

    > On Mon, 5 Sep 2005 20:36:06 -0700, 0m (Paul Mitchum)
    > wrote:
    >
    > >> OK, I see what you're saying.
    > >> So you take your first row, and then pitch up enough to overlap 1/2 to
    > >> 1/3, then yaw as you take the next row. Repeat as necessary.
    > >> What am I missing? Where's the roll?
    > >>
    > >> Maybe my laziness and "form follows function" mindset is hiding
    > >> something from me, but it seems that this is being made much harder
    > >> than it needs to be.

    > >
    > >You *can* take a bunch of pictures and then stitch them together however
    > >you want. It's just easier to get things started if you're methodical,
    > >and already know the orientation of each picture.

    >
    > Maybe I'm just doing it a way that works.
    > When I take a multi-row pano (and they are only two rows for me), I
    > use my tripod, level it, and start taking shots from the left,
    > overlapping by about 1/3 for each successive shot. When I reach the
    > end of that row, I pitch the camera up to overlap the lower pic by
    > about 1/3, and proceed to take shots to the left, overlapping by about
    > 1/3.
    > Maybe I'm just making it sound easier than it is. I don't know. But I
    > see no need to measure anything except by the Mk I eyeball.


    Then don't. :)


    > >You could even dump a bunch of pictures into something like autopano and
    > >let it make all the decisions for you and get reasonable results.

    >
    > You get better results by measuring than by eyeballing, if the overlap
    > is sufficient by eyeballing?


    If you measure, you can get repeatable results. You can run a script on
    PanoTools which you've set up beforehand, and not have to set up control
    points and all that jazz. If you need a quick turnaround on the job,
    it's better to work it that way.

    If you're just shooting landscapes for fun or whatever, then eyeball it
    if you want to. The real limitation comes from parallax errors, and if
    you've got a pano head, that limitation goes away. Taking a lot of
    pictures with a lot of overlap won't hurt the result, but if you're not
    methodical, you might get back home and find a gap or two. And that'd
    suck.
    Paul Mitchum, Sep 6, 2005
    #19
  20. l e o

    Bill Funk Guest

    On Tue, 6 Sep 2005 12:19:31 -0700, 0m (Paul Mitchum)
    wrote:

    >> You get better results by measuring than by eyeballing, if the overlap
    >> is sufficient by eyeballing?

    >
    >If you measure, you can get repeatable results. You can run a script on
    >PanoTools which you've set up beforehand, and not have to set up control
    >points and all that jazz. If you need a quick turnaround on the job,
    >it's better to work it that way.
    >
    >If you're just shooting landscapes for fun or whatever, then eyeball it
    >if you want to. The real limitation comes from parallax errors, and if
    >you've got a pano head, that limitation goes away. Taking a lot of
    >pictures with a lot of overlap won't hurt the result, but if you're not
    >methodical, you might get back home and find a gap or two. And that'd
    >suck.


    Now THAT is useful and to the point! Thanks!
    Yes, I shoot for fun. If I were shooting for serious money, maybe I'd
    do it differently.

    --
    Bill Funk
    Replace "g" with "a"
    funktionality.blogspot.com
    Bill Funk, Sep 6, 2005
    #20
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Silverstrand

    GeForce 7800 GTX Head-to-Head @ TrustedReviews

    Silverstrand, Sep 13, 2005, in forum: Front Page News
    Replies:
    0
    Views:
    728
    Silverstrand
    Sep 13, 2005
  2. Paul Cordes

    Re: Good value Pano head for my PowershotG1 please

    Paul Cordes, Aug 3, 2003, in forum: Digital Photography
    Replies:
    19
    Views:
    665
    Rusty Wright
    Aug 5, 2003
  3. oli
    Replies:
    0
    Views:
    1,162
  4. bob

    easy free/cheap pano head

    bob, Sep 7, 2004, in forum: Digital Photography
    Replies:
    1
    Views:
    509
  5. Brian Stirling

    Pano head for Gitzo 1325

    Brian Stirling, Nov 22, 2004, in forum: Digital Photography
    Replies:
    1
    Views:
    537
    Clyde
    Nov 22, 2004
Loading...

Share This Page