Loading ...
Sorry, an error occurred while loading the content.

Re: zebra sky help

Expand Messages
  • Hans
    ... Here is an example. These are images from a brand new Nikon D7000 http://www.panoramas.dk/technics/bad-shutter-befor-after.jpg The after shows it after
    Message 1 of 19 , Jul 3, 2011
    • 0 Attachment
      --- In PanoToolsNG@yahoogroups.com, "Hans" <hans@...> wrote:
      >
      >
      >
      > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
      > >
      > > Thanks John. The images were shot in portrait orientation, which I think would orient the shutter movement up and down. I guess the most important thing I've taken away so far is that this is not a common problem, so worth doing some testing on my end.
      >
      > Well this is a quite common problem and its old as focal plain shutters has existed. Which means it is soon a 100 anniversary as the first Leica was in 1913.
      >
      > Almost all shutters today are vertical which means horizontal when you take the images in portrait.
      >
      > As I said the difference in exposure to make this is minimal only 10% is enough as no automatic exposure or vignetting correction helps.
      >
      > I can see you panorama may even have been taken in minus degrees which makes another trigger for it.
      > Remember that even in todays digital world the shutter is still a mechanical thing.
      >
      > I had a lot of this problem on my Pentax 67 back in the 80s.
      >
      > Hans

      Here is an example. These are images from a brand new Nikon D7000
      http://www.panoramas.dk/technics/bad-shutter-befor-after.jpg

      The after shows it after applying graduation adjustment in photoshop.

      Here is how the stitch looks from the not adjusted.
      http://www.panoramas.dk/technics/bad-shutter-result.jpg

      As you can see the difference is very small just 3% but the result on the stitched is large.

      You can not see that there is a small exposure difference without placing the images side by side.

      Hans





      >
      >
      >
      >
      > >
      > > --- In PanoToolsNG@yahoogroups.com, "John Houghton" <j.houghton@> wrote:
      > > >
      > > > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
      > > > > The images were shot with a D300 in manual mode with nothing left to chance (that I know of... but definitely not shutter speed, aperture or iso... not sure what else is left when shooting raw) and processed identically in ACR to tiff. Putting the tiffs side by side I dont notice an exposure difference,
      > > >
      > > > It's difficult to tell whether the images are in portrait or landscape orientation, so the shutter curtain explanation suggested by Hans may or may not be a possible explanation. Setting the images side-by-side won't reveal anything. They need to be overlaid as in the PSB file. I would take a small section of 3 or 4 overlapping sky images and limit the investigation to those. You could try a curves adjustment with a L-R graduated selection or mask - in 16 bit mode and with a curves adjustment layer on top to enhance the contrast so that the levels differences are magnified. And you could try different blenders: the Photoshop Auto-Blend Layers often works well.
      > > >
      > > > John
      > > >
      > >
      >
    • Ormar
      Do you tried to convert images with other converter (not ACR or Lightroom) I had similar problem and moving to Capture One fixed the problem. By my experience
      Message 2 of 19 , Jul 3, 2011
      • 0 Attachment
        Do you tried to convert images with other converter (not ACR or Lightroom) I had similar problem and moving to Capture One fixed the problem. By my experience ACR will do it, if you use strong enhancements.
      • Erik Krause
        ... Matt, what exposure time did you use? Hans is right about the shutter speed differences. However, this effect gets worse if you use shorter exposures. If
        Message 3 of 19 , Jul 3, 2011
        • 0 Attachment
          Am 03.07.2011 20:20, schrieb Hans:
          > Well this is a quite common problem and its old as focal plain
          > shutters has existed. Which means it is soon a 100 anniversary as
          > the first Leica was in 1913.

          > Almost all shutters today are vertical which means horizontal when
          > you take the images in portrait.

          > As I said the difference in exposure to make this is minimal only 10%
          > is enough as no automatic exposure or vignetting correction helps.

          Matt, what exposure time did you use? Hans is right about the shutter
          speed differences. However, this effect gets worse if you use shorter
          exposures. If you use longer exposures where the total opening time of
          the shutter is much longer than the actual shutter movement it shouldn't
          play a role anymore.

          The time one shutter needs to move is approximately half the shortest
          time that still allows flash exposure (with a conventional flash - no
          tricks like high speed sync). The reason is, that the shutter needs to
          be fully open before the flash fires and only after that can the closing
          curtain start.

          So if you shoot at this exposure time (or slower) you shouldn't get the
          effect. However, additional care must be taken to avoide camera shake at
          slow speeds with long lenses...

          --
          Erik Krause
          http://www.erik-krause.de
        • Erik Krause
          ... Brand new? The effect in those images seems not to be simple speed differences in opening and closing curtain, it looks like the shutter hangs in some
          Message 4 of 19 , Jul 3, 2011
          • 0 Attachment
            Am 03.07.2011 20:39, schrieb Hans:
            > Here is an example. These are images from a brand new Nikon D7000
            > http://www.panoramas.dk/technics/bad-shutter-befor-after.jpg

            Brand new? The effect in those images seems not to be simple speed
            differences in opening and closing curtain, it looks like the shutter
            hangs in some point.

            I had this in my very old (analog) EOS 600 and the reason was a piece of
            foam rubber intended to brake the shutter. This piece disintegrated into
            a gluey smear which sticked the shutter blades together - a common
            problem for older EOS cameras. I managed to clean the shutter with some
            cleaning solvent and the camera worked ok for another couple of years.

            If this happens in a brand new camera it should be returned in my opinion.

            --
            Erik Krause
            http://www.erik-krause.de
          • Mark Fink
            Hi Matt, I don t know if this is contributing to the problem, but why are you converting to sRGB TIFF? I convert to ProPhoto TIFF based on something I heard
            Message 5 of 19 , Jul 4, 2011
            • 0 Attachment
              Hi Matt,

              I don't know if this is contributing to the problem, but why are you
              converting to sRGB TIFF? I convert to ProPhoto TIFF based on something I
              heard this guy say a couple years ago:

              http://schewephoto.com/sRGB-VS-PPRGB/

              It would be interesting to redo a section of your pano in ProPhoto and
              compare via overlay with your sRGB.

              Mark

              -----Original Message-----
              From: PanoToolsNG@yahoogroups.com [mailto:PanoToolsNG@yahoogroups.com] On
              Behalf Of web@...
              Sent: Sunday, July 03, 2011 6:46 AM
              To: PanoToolsNG@yahoogroups.com
              Subject: [PanoToolsNG] zebra sky help

              I've noticed on several of my gigapixel panoramas that there is a vertical
              banding problem, something like zebra stripes. It seems like every other
              image or so is brighter than the other, so that from left-to-right there is
              not a seamless sky but rather alternating light and dark. It is subtle, but
              clear, especially when you zoom out. It shows up in snow fields too.

              Here is an example (not sure whether this works on a Mac):
              http://www.drmattnolan.org/photography/2011/ahab_gigatest/ahabridge_20k.htm

              This is a 20k wide version of a 120k pixel wide original; the zebra pattern
              is present in the full size original in Photoshop, so its not a resizing or
              HDview problem. Workflow was 16 bit throughout, from NEF to sRGB tiffs to
              PSB, then output. The images were shot with a D300 in manual mode with
              nothing left to chance (that I know of... but definitely not shutter speed,
              aperture or iso... not sure what else is left when shooting raw) and
              processed identically in ACR to tiff. Putting the tiffs side by side I dont
              notice an exposure difference, but I havent tested it well either. In
              PTgui, I tried with and without the exposure optimization tool, with not
              much difference. I guess I'm thinking I'm either missing some setting in
              PTgui or some setting on the camera, but before I spend much time with it I
              was hoping someone might already be familiar with this issue and could let
              me know the cause?

              Thanks,
              Matt





              ------------------------------------

              --




              -----
              No virus found in this message.
              Checked by AVG - www.avg.com
              Version: 10.0.1388 / Virus Database: 1516/3741 - Release Date: 07/03/11
            • web@drmattnolan.org
              Thanks Hans. I still havent done any testing of my images, but it does seem that you have identified the issue. It s a bit distressing, as this panorama,
              Message 6 of 19 , Jul 4, 2011
              • 0 Attachment
                Thanks Hans. I still havent done any testing of my images, but it does seem that you have identified the issue. It's a bit distressing, as this panorama, like many others of mine, has about 400 images in it, and even if I had the time, I'm not sure I'd now where to begin, as if I understand correctly this is a somewhat random process affecting each image differently. If I understood your fix correctly, it was not a simply expsoure adjustment applied to the whole image, but rather a gradient?
                -Matt



                --- In PanoToolsNG@yahoogroups.com, "Hans" <hans@...> wrote:
                >
                >
                >
                > --- In PanoToolsNG@yahoogroups.com, "Hans" <hans@> wrote:
                > >
                > >
                > >
                > > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
                > > >
                > > > Thanks John. The images were shot in portrait orientation, which I think would orient the shutter movement up and down. I guess the most important thing I've taken away so far is that this is not a common problem, so worth doing some testing on my end.
                > >
                > > Well this is a quite common problem and its old as focal plain shutters has existed. Which means it is soon a 100 anniversary as the first Leica was in 1913.
                > >
                > > Almost all shutters today are vertical which means horizontal when you take the images in portrait.
                > >
                > > As I said the difference in exposure to make this is minimal only 10% is enough as no automatic exposure or vignetting correction helps.
                > >
                > > I can see you panorama may even have been taken in minus degrees which makes another trigger for it.
                > > Remember that even in todays digital world the shutter is still a mechanical thing.
                > >
                > > I had a lot of this problem on my Pentax 67 back in the 80s.
                > >
                > > Hans
                >
                > Here is an example. These are images from a brand new Nikon D7000
                > http://www.panoramas.dk/technics/bad-shutter-befor-after.jpg
                >
                > The after shows it after applying graduation adjustment in photoshop.
                >
                > Here is how the stitch looks from the not adjusted.
                > http://www.panoramas.dk/technics/bad-shutter-result.jpg
                >
                > As you can see the difference is very small just 3% but the result on the stitched is large.
                >
                > You can not see that there is a small exposure difference without placing the images side by side.
                >
                > Hans
                >
                >
                >
                >
                >
                > >
                > >
                > >
                > >
                > > >
                > > > --- In PanoToolsNG@yahoogroups.com, "John Houghton" <j.houghton@> wrote:
                > > > >
                > > > > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
                > > > > > The images were shot with a D300 in manual mode with nothing left to chance (that I know of... but definitely not shutter speed, aperture or iso... not sure what else is left when shooting raw) and processed identically in ACR to tiff. Putting the tiffs side by side I dont notice an exposure difference,
                > > > >
                > > > > It's difficult to tell whether the images are in portrait or landscape orientation, so the shutter curtain explanation suggested by Hans may or may not be a possible explanation. Setting the images side-by-side won't reveal anything. They need to be overlaid as in the PSB file. I would take a small section of 3 or 4 overlapping sky images and limit the investigation to those. You could try a curves adjustment with a L-R graduated selection or mask - in 16 bit mode and with a curves adjustment layer on top to enhance the contrast so that the levels differences are magnified. And you could try different blenders: the Photoshop Auto-Blend Layers often works well.
                > > > >
                > > > > John
                > > > >
                > > >
                > >
                >
              • web@drmattnolan.org
                Erik, I used a shutter speed of 1/1600s, f/7.1, ISO 200. So I cant go any lower ISO, but I could decrease f/ a stop and slow the shutter by half, but beyond
                Message 7 of 19 , Jul 4, 2011
                • 0 Attachment
                  Erik,

                  I used a shutter speed of 1/1600s, f/7.1, ISO 200. So I cant go any lower ISO, but I could decrease f/ a stop and slow the shutter by half, but beyond that I guess I would have to go to a neutral density filter? Seems counter-intuitive, but rationally it makes sense that a small percent error in time has less impact with a longer interval. Certainly it seems worth maximing shutter speed in the future.

                  I do a lot of vertical aerial photography too, I'd hate to have to slow down shutter speeds to make better mosaics, as camera shake here is a major concern...!

                  -Matt


                  --- In PanoToolsNG@yahoogroups.com, Erik Krause <erik.krause@...> wrote:
                  >
                  > Am 03.07.2011 20:20, schrieb Hans:
                  > > Well this is a quite common problem and its old as focal plain
                  > > shutters has existed. Which means it is soon a 100 anniversary as
                  > > the first Leica was in 1913.
                  >
                  > > Almost all shutters today are vertical which means horizontal when
                  > > you take the images in portrait.
                  >
                  > > As I said the difference in exposure to make this is minimal only 10%
                  > > is enough as no automatic exposure or vignetting correction helps.
                  >
                  > Matt, what exposure time did you use? Hans is right about the shutter
                  > speed differences. However, this effect gets worse if you use shorter
                  > exposures. If you use longer exposures where the total opening time of
                  > the shutter is much longer than the actual shutter movement it shouldn't
                  > play a role anymore.
                  >
                  > The time one shutter needs to move is approximately half the shortest
                  > time that still allows flash exposure (with a conventional flash - no
                  > tricks like high speed sync). The reason is, that the shutter needs to
                  > be fully open before the flash fires and only after that can the closing
                  > curtain start.
                  >
                  > So if you shoot at this exposure time (or slower) you shouldn't get the
                  > effect. However, additional care must be taken to avoide camera shake at
                  > slow speeds with long lenses...
                  >
                  > --
                  > Erik Krause
                  > http://www.erik-krause.de
                  >
                • Hans
                  ... I actually used the Camera Raw adjustment brush when converting them. However a gradient on the converted tifs might work just as good. It takes some
                  Message 8 of 19 , Jul 4, 2011
                  • 0 Attachment
                    --- In PanoToolsNG@yahoogroups.com, web@... wrote:
                    >
                    >
                    > Thanks Hans. I still havent done any testing of my images, but it does seem that you have identified the issue. It's a bit distressing, as this panorama, like many others of mine, has about 400 images in it, and even if I had the time, I'm not sure I'd now where to begin, as if I understand correctly this is a somewhat random process affecting each image differently. If I understood your fix correctly, it was not a simply expsoure adjustment applied to the whole image, but rather a gradient?


                    I actually used the Camera Raw adjustment brush when converting them.
                    However a gradient on the converted tifs might work just as good.

                    It takes some experiments to find the correct amount.

                    Hans



                    > -Matt
                    >
                    >
                    >
                    > --- In PanoToolsNG@yahoogroups.com, "Hans" <hans@> wrote:
                    > >
                    > >
                    > >
                    > > --- In PanoToolsNG@yahoogroups.com, "Hans" <hans@> wrote:
                    > > >
                    > > >
                    > > >
                    > > > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
                    > > > >
                    > > > > Thanks John. The images were shot in portrait orientation, which I think would orient the shutter movement up and down. I guess the most important thing I've taken away so far is that this is not a common problem, so worth doing some testing on my end.
                    > > >
                    > > > Well this is a quite common problem and its old as focal plain shutters has existed. Which means it is soon a 100 anniversary as the first Leica was in 1913.
                    > > >
                    > > > Almost all shutters today are vertical which means horizontal when you take the images in portrait.
                    > > >
                    > > > As I said the difference in exposure to make this is minimal only 10% is enough as no automatic exposure or vignetting correction helps.
                    > > >
                    > > > I can see you panorama may even have been taken in minus degrees which makes another trigger for it.
                    > > > Remember that even in todays digital world the shutter is still a mechanical thing.
                    > > >
                    > > > I had a lot of this problem on my Pentax 67 back in the 80s.
                    > > >
                    > > > Hans
                    > >
                    > > Here is an example. These are images from a brand new Nikon D7000
                    > > http://www.panoramas.dk/technics/bad-shutter-befor-after.jpg
                    > >
                    > > The after shows it after applying graduation adjustment in photoshop.
                    > >
                    > > Here is how the stitch looks from the not adjusted.
                    > > http://www.panoramas.dk/technics/bad-shutter-result.jpg
                    > >
                    > > As you can see the difference is very small just 3% but the result on the stitched is large.
                    > >
                    > > You can not see that there is a small exposure difference without placing the images side by side.
                    > >
                    > > Hans
                    > >
                    > >
                    > >
                    > >
                    > >
                    > > >
                    > > >
                    > > >
                    > > >
                    > > > >
                    > > > > --- In PanoToolsNG@yahoogroups.com, "John Houghton" <j.houghton@> wrote:
                    > > > > >
                    > > > > > --- In PanoToolsNG@yahoogroups.com, web@ wrote:
                    > > > > > > The images were shot with a D300 in manual mode with nothing left to chance (that I know of... but definitely not shutter speed, aperture or iso... not sure what else is left when shooting raw) and processed identically in ACR to tiff. Putting the tiffs side by side I dont notice an exposure difference,
                    > > > > >
                    > > > > > It's difficult to tell whether the images are in portrait or landscape orientation, so the shutter curtain explanation suggested by Hans may or may not be a possible explanation. Setting the images side-by-side won't reveal anything. They need to be overlaid as in the PSB file. I would take a small section of 3 or 4 overlapping sky images and limit the investigation to those. You could try a curves adjustment with a L-R graduated selection or mask - in 16 bit mode and with a curves adjustment layer on top to enhance the contrast so that the levels differences are magnified. And you could try different blenders: the Photoshop Auto-Blend Layers often works well.
                    > > > > >
                    > > > > > John
                    > > > > >
                    > > > >
                    > > >
                    > >
                    >
                  • web@drmattnolan.org
                    Mark, For a while I had been using Prophoto too, based on some books and articles I read. However, I found it in the end to be more trouble than it was worth.
                    Message 9 of 19 , Jul 4, 2011
                    • 0 Attachment
                      Mark,

                      For a while I had been using Prophoto too, based on some books and articles I read. However, I found it in the end to be more trouble than it was worth. In particular, I was always getting color space mismatches for web stuff, which gave images a greenish cast when trying to display as sRGB. Often I wouldnt catch it, then I'd have to reupload something. It was also causing problems printing, trying to get hardware and software to honor color space choices and also hoping the printer could print all of the colors without swapping some. Plus my eye isnt critical enough, I'm pretty happy if what comes out on the printer looks reasonably close to the screen. And since my photography is primarily in glacier regions, there's not a lot of color there. So a long winded answer, but the short version is just convenience and one less thing to screw up or screw things up. But once I start testing the solution to the zebra pattern, I will test color space for completeness.

                      Thanks,
                      Matt


                      --- In PanoToolsNG@yahoogroups.com, "Mark Fink" <markdfink@...> wrote:
                      >
                      > Hi Matt,
                      >
                      > I don't know if this is contributing to the problem, but why are you
                      > converting to sRGB TIFF? I convert to ProPhoto TIFF based on something I
                      > heard this guy say a couple years ago:
                      >
                      > http://schewephoto.com/sRGB-VS-PPRGB/
                      >
                      > It would be interesting to redo a section of your pano in ProPhoto and
                      > compare via overlay with your sRGB.
                      >
                      > Mark
                      >
                      > -----Original Message-----
                      > From: PanoToolsNG@yahoogroups.com [mailto:PanoToolsNG@yahoogroups.com] On
                      > Behalf Of web@...
                      > Sent: Sunday, July 03, 2011 6:46 AM
                      > To: PanoToolsNG@yahoogroups.com
                      > Subject: [PanoToolsNG] zebra sky help
                      >
                      > I've noticed on several of my gigapixel panoramas that there is a vertical
                      > banding problem, something like zebra stripes. It seems like every other
                      > image or so is brighter than the other, so that from left-to-right there is
                      > not a seamless sky but rather alternating light and dark. It is subtle, but
                      > clear, especially when you zoom out. It shows up in snow fields too.
                      >
                      > Here is an example (not sure whether this works on a Mac):
                      > http://www.drmattnolan.org/photography/2011/ahab_gigatest/ahabridge_20k.htm
                      >
                      > This is a 20k wide version of a 120k pixel wide original; the zebra pattern
                      > is present in the full size original in Photoshop, so its not a resizing or
                      > HDview problem. Workflow was 16 bit throughout, from NEF to sRGB tiffs to
                      > PSB, then output. The images were shot with a D300 in manual mode with
                      > nothing left to chance (that I know of... but definitely not shutter speed,
                      > aperture or iso... not sure what else is left when shooting raw) and
                      > processed identically in ACR to tiff. Putting the tiffs side by side I dont
                      > notice an exposure difference, but I havent tested it well either. In
                      > PTgui, I tried with and without the exposure optimization tool, with not
                      > much difference. I guess I'm thinking I'm either missing some setting in
                      > PTgui or some setting on the camera, but before I spend much time with it I
                      > was hoping someone might already be familiar with this issue and could let
                      > me know the cause?
                      >
                      > Thanks,
                      > Matt
                      >
                      >
                      >
                      >
                      >
                      > ------------------------------------
                      >
                      > --
                      >
                      >
                      >
                      >
                      > -----
                      > No virus found in this message.
                      > Checked by AVG - www.avg.com
                      > Version: 10.0.1388 / Virus Database: 1516/3741 - Release Date: 07/03/11
                      >
                    • Erik Krause
                      ... In Photoshop: Use a levels or brightness adjustment layer and paint a gradient in the associated mask with the gradient tool. You can use the color sampler
                      Message 10 of 19 , Jul 4, 2011
                      • 0 Attachment
                        Am 04.07.2011 20:13, schrieb web@...:
                        > If I understood your fix correctly, it was not a simply expsoure
                        > adjustment applied to the whole image, but rather a gradient?

                        In Photoshop: Use a levels or brightness adjustment layer and paint a
                        gradient in the associated mask with the gradient tool. You can use the
                        color sampler tool to measure the brightness values in the same point in
                        the neighboring image. Once you found good values record the whole
                        process as an action and create a droplet from it. You then can drop all
                        your 400 files on it and let the computer process them while you do
                        something different.

                        You can even use a flat field: shoot an evenly lit surface with same
                        aperture and exposure settings (you can rotate the camera, shoot several
                        and average later). Do an auto levels on the resulting image and use the
                        result as a mask. To find the correct adjustment try on the original
                        image until lighting is perfectly even. To see this better you can put
                        an auto levels layer above.

                        --
                        Erik Krause
                        http://www.erik-krause.de
                      Your message has been successfully submitted and would be delivered to recipients shortly.