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

Re: Grainy image from horsehead nebula

Expand Messages
  • Mike Dodd
    ... Are you sure about denser, Joe? True, NB filters transmit somewhat less light than RGB filters, but I don t think they are 2-3 times denser. The Astrodon
    Message 1 of 20 , Jan 22, 2013
    • 0 Attachment
      --- In ccd-newastro@yahoogroups.com, "JoeMize" wrote:
      > Mark, H-a is 2-3 times denser than regular LRGB filters, OIII and SII are even more dense. Because of that you should be using much longer H-a exposures, try 15-20 minutes.

      Are you sure about "denser," Joe? True, NB filters transmit somewhat less light than RGB filters, but I don't think they are 2-3 times denser. The Astrodon website <http://www.astrodon.com/products/filters/narrowband/> states, "[our] 5 and 3 nm FWHM filters guaranteed >90% transmittance (typically 92-97%). Measurement of three 7 nm H-a filters from a competitor showed 86%, 74% and 71% transmittance at the H-a emission wavelength compared to typical 95% transmittance for our 5 nm narrowband."

      Even the worst filter (71%) wouldn't require 2-3 times the exposure.

      NB filters _look_ dense because they reject most of the spectrum, but at the passband peak, they are fairly transparent (71% isn't bad, and 95% is darn-near clear).

      There is no argument that longer exposures improve S/N, but that's true no matter what filter you're using -- even no filter at all.

      The nice thing about NB imaging is that you _can_ make long exposures to improve S/N without suffering the consequences of light pollution or moon glow.

      Regards, Mike
      http://astronomy.mdodd.com
    • JoeMize
      Mike you are correct and I also understand both WB and NB AstroDon discrete filter transmittance throughputs are nearly identical. However I was speaking in
      Message 2 of 20 , Jan 23, 2013
      • 0 Attachment
        Mike you are correct and I also understand both WB and NB AstroDon discrete
        filter transmittance throughputs are nearly identical. However I was
        speaking in generalities to someone new to mono-CCD Cameras, "These are my
        first attempts with a CCD camera..." trying to help him understand why his
        short 5 minute H-a exposures were noisy compared to his first linked image
        below.

        > The images for M42 turned out to be pretty good - here is a 5 min
        > exposure:
        > https://picasaweb.google.com/lh/photo/S1NxZAawS_UyEjw82ED-JayoCSnGTeDMLN9VMnkvzBM?feat=directlink
        > But the images for the Horsehead nebula were VERY grainy:
        > https://picasaweb.google.com/lh/photo/LryYFclMN-I85gHJ0rBVi6yoCSnGTeDMLN9VMnkvzBM?feat=directlink
        > I didn't change anything in the imaging train or such. Both images are 5
        > min exposures. How did this happen?

        I was suggesting he expose his H-a filter much longer, 15-20mins, to smooth
        his images. I was not arguing about Stacking many short NB exposures to
        eventually achieve the same thing...joe :)

        "May You Go Among The Imperishable Stars"
        Joe Mize www.cav-sfo.com
        Chiefland Astronomy Village (CAV), Fla
        StarFields Observatory, (SFO).

        ----- Original Message -----
        From: "Mike Dodd" <mike@...>
        To: <ccd-newastro@yahoogroups.com>
        Sent: Tuesday, January 22, 2013 10:49 PM
        Subject: [ccd-newastro] Re: Grainy image from horsehead nebula


        > --- In ccd-newastro@yahoogroups.com, "JoeMize" wrote:
        >> Mark, H-a is 2-3 times denser than regular LRGB filters, OIII and SII are
        >> even more dense. Because of that you should be using much longer H-a
        >> exposures, try 15-20 minutes.
        >
        > Are you sure about "denser," Joe? True, NB filters transmit somewhat less
        > light than RGB filters, but I don't think they are 2-3 times denser. The
        > Astrodon website <http://www.astrodon.com/products/filters/narrowband/>
        > states, "[our] 5 and 3 nm FWHM filters guaranteed >90% transmittance
        > (typically 92-97%). Measurement of three 7 nm H-a filters from a
        > competitor showed 86%, 74% and 71% transmittance at the H-a emission
        > wavelength compared to typical 95% transmittance for our 5 nm narrowband."
        >
        > Even the worst filter (71%) wouldn't require 2-3 times the exposure.
        >
        > NB filters _look_ dense because they reject most of the spectrum, but at
        > the passband peak, they are fairly transparent (71% isn't bad, and 95% is
        > darn-near clear).
        >
        > There is no argument that longer exposures improve S/N, but that's true no
        > matter what filter you're using -- even no filter at all.
        >
        > The nice thing about NB imaging is that you _can_ make long exposures to
        > improve S/N without suffering the consequences of light pollution or moon
        > glow.
        >
        > Regards, Mike
        > http://astronomy.mdodd.com
        >
        >
        >
        >
        > ------------------------------------
        >
        >
        >
        >
        >
      • mark_manner_spot_obsrv
        Mike, regarding you comment to Joe below if you are saying that the same amount of light gets through a narrow band and broad band filter, that isn t correct.
        Message 3 of 20 , Jan 23, 2013
        • 0 Attachment
          Mike, regarding you comment to Joe below if you are saying that the same amount of light gets through a narrow band and broad band filter, that isn't correct. Perhaps I am missing your point, so pardon me if I am. Although the transmission percentage for narrow and wide band filters is close to the same, you are only transmitting that percentage of photons with the wavelength in the band of the filter. So, a 90% Ha filter and a 90% L or R will be quite different (assuming a source that emits light over the wavelengths in question). As a quick test, to get the same number of counts in an image taken with an STL11K just now pointing at an alnitak flat panel (that puts out light in rgb(and some nir), I had to use 30 sec for the Ha filter, 2.5 sec for the R, and 1 sec for the L.
          Mark

          --- In ccd-newastro@yahoogroups.com, "Mike Dodd" wrote:
          >

          > Are you sure about "denser," Joe? True, NB filters transmit somewhat less light than RGB filters, but I don't think they are 2-3 times denser. The Astrodon website states, "[our] 5 and 3 nm FWHM filters guaranteed >90% transmittance (typically 92-97%). Measurement of three 7 nm H-a filters from a competitor showed 86%, 74% and 71% transmittance at the H-a emission wavelength compared to typical 95% transmittance for our 5 nm narrowband."

          > Regards, Mike
          > http://astronomy.mdodd.com
          >
        • Mike Dodd
          ... Sure, because the EL flat panel probably puts out only 1/30 of its light in at 656nm. But we care about that only for flat frames. Imaging is different.
          Message 4 of 20 , Jan 23, 2013
          • 0 Attachment
            mark_manner_spot_obsrv wrote:
            > Mike, regarding you comment to Joe below if you are saying that the
            > same amount of light gets through a narrow band and broad band
            > filter, that isn't correct. Perhaps I am missing your point, so
            > pardon me if I am. Although the transmission percentage for narrow
            > and wide band filters is close to the same, you are only transmitting
            > that percentage of photons with the wavelength in the band of the
            > filter. So, a 90% Ha filter and a 90% L or R will be quite different
            > (assuming a source that emits light over the wavelengths in
            > question). As a quick test, to get the same number of counts in an
            > image taken with an STL11K just now pointing at an alnitak flat panel
            > (that puts out light in rgb(and some nir), I had to use 30 sec for
            > the Ha filter, 2.5 sec for the R, and 1 sec for the L. Mark

            Sure, because the EL flat panel probably puts out only 1/30 of its light
            in at 656nm.

            But we care about that only for flat frames. Imaging is different.

            When imaging in NB, I'm interested in the specific wavelength emitted by
            that target. As the Astrodon quotation shows, NB filters pass between
            71% and 97% of the light at their peak wavelength. Yes, stars and
            objects emitting other wavelengths will be greatly attenuated. I would
            expect a reflection nebula to be dim through an H-a filter.

            I was trying to make the point that the light from a NB target is not
            attenuated very much through a NB filter, and so does not require longer
            exposures. in other words, if I were to image the Horsehead at the same
            exposure time unfiltered and through an Astrodon H-a filter, both images
            would have approximately the same ADU for the HH areas, ignoring other
            possible light pollution.

            As I wrote last night, longer exposures always improve S/N, so images
            benefit from using a NB filter that _allows_ very long exposures. It's
            great to be able to expose for 20 or 30 minutes without the image being
            ruined by sky glow.

            --
            Mike

            Mike Dodd
            http://astronomy.mdodd.com
            Louisa County, Virginia USA N37.58.23 W77.56.24
          • mark_manner_spot_obsrv
            Mike the ratio of light through the Ha filter to the L and RGB filters won t be as significant with a real sky source as compared to the flat panel, but my
            Message 5 of 20 , Jan 23, 2013
            • 0 Attachment
              Mike the ratio of light through the Ha filter to the L and RGB filters won't be as significant with a real sky source as compared to the flat panel, but my point in response to yours is still valid.
              Best,
              Mark


              --- In ccd-newastro@yahoogroups.com, Mike Dodd wrote:
              >
              > mark_manner_spot_obsrv wrote:
              > > Mike, regarding you comment to Joe below if you are saying that the
              > > same amount of light gets through a narrow band and broad band
              > > filter, that isn't correct. Perhaps I am missing your point, so
              > > pardon me if I am. Although the transmission percentage for narrow
              > > and wide band filters is close to the same, you are only transmitting
              > > that percentage of photons with the wavelength in the band of the
              > > filter. So, a 90% Ha filter and a 90% L or R will be quite different
              > > (assuming a source that emits light over the wavelengths in
              > > question). As a quick test, to get the same number of counts in an
              > > image taken with an STL11K just now pointing at an alnitak flat panel
              > > (that puts out light in rgb(and some nir), I had to use 30 sec for
              > > the Ha filter, 2.5 sec for the R, and 1 sec for the L. Mark
              >
              > Sure, because the EL flat panel probably puts out only 1/30 of its light
              > in at 656nm.
              >
              > But we care about that only for flat frames. Imaging is different.
              >
              > When imaging in NB, I'm interested in the specific wavelength emitted by
              > that target. As the Astrodon quotation shows, NB filters pass between
              > 71% and 97% of the light at their peak wavelength. Yes, stars and
              > objects emitting other wavelengths will be greatly attenuated. I would
              > expect a reflection nebula to be dim through an H-a filter.
              >
              > I was trying to make the point that the light from a NB target is not
              > attenuated very much through a NB filter, and so does not require longer
              > exposures. in other words, if I were to image the Horsehead at the same
              > exposure time unfiltered and through an Astrodon H-a filter, both images
              > would have approximately the same ADU for the HH areas, ignoring other
              > possible light pollution.
              >
              > As I wrote last night, longer exposures always improve S/N, so images
              > benefit from using a NB filter that _allows_ very long exposures. It's
              > great to be able to expose for 20 or 30 minutes without the image being
              > ruined by sky glow.
              >
              > --
              > Mike
              >
              > Mike Dodd
              > http://astronomy.mdodd.com
              > Louisa County, Virginia USA N37.58.23 W77.56.24
              >
            • Don Goldman
              Hi, Guys, Interesting thread. Yes, our 5 nm H-a transmits 90% at the emission wavelength at 656 nm but only over that ~ 5nm window. The luminance filter
              Message 6 of 20 , Jan 24, 2013
              • 0 Attachment
                Hi, Guys,

                Interesting thread.

                Yes, our 5 nm H-a transmits >90% at the emission wavelength at 656 nm but only over that ~ 5nm window. The luminance filter transmits >90% from 400 - 700 nm, or let's say over 300 nm. So, ignoring QE differences for the moment, the H-a filter would provide a signal 5/300 or about 1.6% as strong as a luminance, or maybe 5% of a Blue filter. Also, as pointed out, the fluorescence-shifted white LED flat panels are dimmer at 656 nm. That, combined with the <5% H-a relative signal means that a 4s exposure for 30,000ADU for a blue filter on an LED flat panel may take 2 min for an H-a filter on an LED flat panel. This is why 5 min NB exposures are generally way too short, and 15-45 min are typical. Unless you are imaging from a city light dome, you generally don't reach the sky noise limit in that time. IF you do, you could use a 3 nm filter that still has >90%T but will decrease the background by 40-50%. Lastly, I take automated sky flats with CCDAutoPilot. The Lum, and RGB are taken first when the sun is further down below the horizon, and the NB filters next as it gets brighter, when there is sufficient light to take the same short exposures to get all flats ~~30k ADU for my camera system.

                Don Goldman
                Astrodon

                --- In ccd-newastro@yahoogroups.com, "mark_manner_spot_obsrv" wrote:
                >
                > Mike the ratio of light through the Ha filter to the L and RGB filters won't be as significant with a real sky source as compared to the flat panel, but my point in response to yours is still valid.
                > Best,
                > Mark
                >
                >
                > --- In ccd-newastro@yahoogroups.com, Mike Dodd wrote:
                > >
                > > mark_manner_spot_obsrv wrote:
                > > > Mike, regarding you comment to Joe below if you are saying that the
                > > > same amount of light gets through a narrow band and broad band
                > > > filter, that isn't correct. Perhaps I am missing your point, so
                > > > pardon me if I am. Although the transmission percentage for narrow
                > > > and wide band filters is close to the same, you are only transmitting
                > > > that percentage of photons with the wavelength in the band of the
                > > > filter. So, a 90% Ha filter and a 90% L or R will be quite different
                > > > (assuming a source that emits light over the wavelengths in
                > > > question). As a quick test, to get the same number of counts in an
                > > > image taken with an STL11K just now pointing at an alnitak flat panel
                > > > (that puts out light in rgb(and some nir), I had to use 30 sec for
                > > > the Ha filter, 2.5 sec for the R, and 1 sec for the L. Mark
                > >
                > > Sure, because the EL flat panel probably puts out only 1/30 of its light
                > > in at 656nm.
                > >
                > > But we care about that only for flat frames. Imaging is different.
                > >
                > > When imaging in NB, I'm interested in the specific wavelength emitted by
                > > that target. As the Astrodon quotation shows, NB filters pass between
                > > 71% and 97% of the light at their peak wavelength. Yes, stars and
                > > objects emitting other wavelengths will be greatly attenuated. I would
                > > expect a reflection nebula to be dim through an H-a filter.
                > >
                > > I was trying to make the point that the light from a NB target is not
                > > attenuated very much through a NB filter, and so does not require longer
                > > exposures. in other words, if I were to image the Horsehead at the same
                > > exposure time unfiltered and through an Astrodon H-a filter, both images
                > > would have approximately the same ADU for the HH areas, ignoring other
                > > possible light pollution.
                > >
                > > As I wrote last night, longer exposures always improve S/N, so images
                > > benefit from using a NB filter that _allows_ very long exposures. It's
                > > great to be able to expose for 20 or 30 minutes without the image being
                > > ruined by sky glow.
                > >
                > > --
                > > Mike
                > >
                > > Mike Dodd
                > > http://astronomy.mdodd.com
                > > Louisa County, Virginia USA N37.58.23 W77.56.24
                > >
                >
              • Mark Striebeck
                I m hesitant to say thanks for all the replies - because that might this very educational thread. But seriously: thanks everybody for your replies. I just
                Message 7 of 20 , Jan 24, 2013
                • 0 Attachment
                  I'm hesitant to say thanks for all the replies - because that might this
                  very educational thread. But seriously: thanks everybody for your replies.
                  I just totally underestimated how big the difference between DSLR and CCD
                  is. And because I went with a monochrome CCD, the additional complexity of
                  using filters (I just wish that my filter wheel would come in soon :-(

                  MarkS


                  On Thu, Jan 24, 2013 at 11:50 AM, Don Goldman <donclearview@...>wrote:

                  > **
                  >
                  >
                  > Hi, Guys,
                  >
                  > Interesting thread.
                  >
                  > Yes, our 5 nm H-a transmits >90% at the emission wavelength at 656 nm but
                  > only over that ~ 5nm window. The luminance filter transmits >90% from 400 -
                  > 700 nm, or let's say over 300 nm. So, ignoring QE differences for the
                  > moment, the H-a filter would provide a signal 5/300 or about 1.6% as strong
                  > as a luminance, or maybe 5% of a Blue filter. Also, as pointed out, the
                  > fluorescence-shifted white LED flat panels are dimmer at 656 nm. That,
                  > combined with the <5% H-a relative signal means that a 4s exposure for
                  > 30,000ADU for a blue filter on an LED flat panel may take 2 min for an H-a
                  > filter on an LED flat panel. This is why 5 min NB exposures are generally
                  > way too short, and 15-45 min are typical. Unless you are imaging from a
                  > city light dome, you generally don't reach the sky noise limit in that
                  > time. IF you do, you could use a 3 nm filter that still has >90%T but will
                  > decrease the background by 40-50%. Lastly, I take automated sky flats with
                  > CCDAutoPilot. The Lum, and RGB are taken first when the sun is further down
                  > below the horizon, and the NB filters next as it gets brighter, when there
                  > is sufficient light to take the same short exposures to get all flats ~~30k
                  > ADU for my camera system.
                  >
                  > Don Goldman
                  > Astrodon
                  >
                  > --- In ccd-newastro@yahoogroups.com, "mark_manner_spot_obsrv" wrote:
                  > >
                  > > Mike the ratio of light through the Ha filter to the L and RGB filters
                  > won't be as significant with a real sky source as compared to the flat
                  > panel, but my point in response to yours is still valid.
                  > > Best,
                  > > Mark
                  > >
                  > >
                  > > --- In ccd-newastro@yahoogroups.com, Mike Dodd wrote:
                  > > >
                  > > > mark_manner_spot_obsrv wrote:
                  > > > > Mike, regarding you comment to Joe below if you are saying that the
                  > > > > same amount of light gets through a narrow band and broad band
                  > > > > filter, that isn't correct. Perhaps I am missing your point, so
                  > > > > pardon me if I am. Although the transmission percentage for narrow
                  > > > > and wide band filters is close to the same, you are only transmitting
                  > > > > that percentage of photons with the wavelength in the band of the
                  > > > > filter. So, a 90% Ha filter and a 90% L or R will be quite different
                  > > > > (assuming a source that emits light over the wavelengths in
                  > > > > question). As a quick test, to get the same number of counts in an
                  > > > > image taken with an STL11K just now pointing at an alnitak flat panel
                  > > > > (that puts out light in rgb(and some nir), I had to use 30 sec for
                  > > > > the Ha filter, 2.5 sec for the R, and 1 sec for the L. Mark
                  > > >
                  > > > Sure, because the EL flat panel probably puts out only 1/30 of its
                  > light
                  > > > in at 656nm.
                  > > >
                  > > > But we care about that only for flat frames. Imaging is different.
                  > > >
                  > > > When imaging in NB, I'm interested in the specific wavelength emitted
                  > by
                  > > > that target. As the Astrodon quotation shows, NB filters pass between
                  > > > 71% and 97% of the light at their peak wavelength. Yes, stars and
                  > > > objects emitting other wavelengths will be greatly attenuated. I would
                  > > > expect a reflection nebula to be dim through an H-a filter.
                  > > >
                  > > > I was trying to make the point that the light from a NB target is not
                  > > > attenuated very much through a NB filter, and so does not require
                  > longer
                  > > > exposures. in other words, if I were to image the Horsehead at the
                  > same
                  > > > exposure time unfiltered and through an Astrodon H-a filter, both
                  > images
                  > > > would have approximately the same ADU for the HH areas, ignoring other
                  > > > possible light pollution.
                  > > >
                  > > > As I wrote last night, longer exposures always improve S/N, so images
                  > > > benefit from using a NB filter that _allows_ very long exposures. It's
                  > > > great to be able to expose for 20 or 30 minutes without the image
                  > being
                  > > > ruined by sky glow.
                  > > >
                  > > > --
                  > > > Mike
                  > > >
                  > > > Mike Dodd
                  > > > http://astronomy.mdodd.com
                  > > > Louisa County, Virginia USA N37.58.23 W77.56.24
                  > > >
                  > >
                  >
                  >
                  >


                  [Non-text portions of this message have been removed]
                • Mike Dodd
                  ... Yes, but I think you ll be extremely pleased with the results once you get a handle on NB imaging. My suggestion: Put the H-a filter on the camera and go
                  Message 8 of 20 , Jan 24, 2013
                  • 0 Attachment
                    Mark Striebeck wrote:
                    > I'm hesitant to say thanks for all the replies - because that might
                    > this very educational thread. But seriously: thanks everybody for
                    > your replies. I just totally underestimated how big the difference
                    > between DSLR and CCD is. And because I went with a monochrome CCD,
                    > the additional complexity of using filters....

                    Yes, but I think you'll be extremely pleased with the results once you
                    get a handle on NB imaging.

                    My suggestion: Put the H-a filter on the camera and go out and make a
                    dozen 15-minute exposures, dark subtract them, and stack them to see
                    what you get. I'll bet you'll really be excited.

                    Enjoy!
                    --
                    Mike

                    Mike Dodd
                    http://astronomy.mdodd.com
                    Louisa County, Virginia USA N37.58.23 W77.56.24
                  • Mark Striebeck
                    Thanks for the encouragement Mike! My challenge there is that so far I could only manage to guide successfully for 5 minutes. If I go longer I get too many
                    Message 9 of 20 , Jan 24, 2013
                    • 0 Attachment
                      Thanks for the encouragement Mike!

                      My challenge there is that so far I could only manage to guide successfully
                      for 5 minutes. If I go longer I get too many failures. One of the main
                      issues (I suspect) is that I guide my SCT with a guidescope and not an OAG.
                      As soon as the filter wheel comes in, I can start using the new OAG (it
                      attaches into the filter wheel).

                      MarkS


                      On Thu, Jan 24, 2013 at 4:49 PM, Mike Dodd <mike@...> wrote:

                      > **
                      >
                      >
                      > Mark Striebeck wrote:
                      > > I'm hesitant to say thanks for all the replies - because that might
                      > > this very educational thread. But seriously: thanks everybody for
                      > > your replies. I just totally underestimated how big the difference
                      > > between DSLR and CCD is. And because I went with a monochrome CCD,
                      > > the additional complexity of using filters....
                      >
                      > Yes, but I think you'll be extremely pleased with the results once you
                      > get a handle on NB imaging.
                      >
                      > My suggestion: Put the H-a filter on the camera and go out and make a
                      > dozen 15-minute exposures, dark subtract them, and stack them to see
                      > what you get. I'll bet you'll really be excited.
                      >
                      > Enjoy!
                      >
                      > --
                      > Mike
                      >
                      > Mike Dodd
                      > http://astronomy.mdodd.com
                      > Louisa County, Virginia USA N37.58.23 W77.56.24
                      >
                      >
                      >


                      [Non-text portions of this message have been removed]
                    • Mike Dodd
                      ... Yes, flexure from an external guide scope makes long exposures difficult. The OAG sounds like a solution. Let us know how it works. -- Mike Mike Dodd
                      Message 10 of 20 , Jan 24, 2013
                      • 0 Attachment
                        Mark Striebeck wrote:
                        > Thanks for the encouragement Mike!
                        >
                        > My challenge there is that so far I could only manage to guide successfully
                        > for 5 minutes. If I go longer I get too many failures. One of the main
                        > issues (I suspect) is that I guide my SCT with a guidescope and not an OAG.
                        > As soon as the filter wheel comes in, I can start using the new OAG (it
                        > attaches into the filter wheel).

                        Yes, flexure from an external guide scope makes long exposures
                        difficult. The OAG sounds like a solution. Let us know how it works.
                        --
                        Mike

                        Mike Dodd
                        http://astronomy.mdodd.com
                        Louisa County, Virginia USA N37.58.23 W77.56.24
                      Your message has been successfully submitted and would be delivered to recipients shortly.