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

Re: RGB background ADU counts

Expand Messages
  • Randy Nulman
    Hi Adam, I think we all agree that background adu is not relevant in terms of color balance (except for tweaking the black points in each channel so that the
    Message 1 of 43 , Aug 2, 2006
    • 0 Attachment
      Hi Adam,
      I think we all agree that background adu is not relevant in terms of
      color balance (except for tweaking the black points in each channel
      so that the background remains "neutral")

      I do want to chime in and say that many beginning and intermediate
      imagers do not know how to do a G2V calibration....and when done, it
      is often done incorrectly <G>)

      My thought is that the first/best option is for one to look at the
      combine ratios (or exposures) of RGB images for similar
      setups...camera, filters, scope, etc and use that as a starting
      point...then tweak as necessary for one's own equip. (I think this
      is what Ron was alluding to when he mentioned "trial and error"?)

      Yes, if one truly knows how to do a G2V calibration, this is also a
      good starting point. However, varying altitudes, sky conditions and
      other factors still will require "tweaking". Nothing is as simple
      is one would hope it to be <VBG>....

      Regards,
      Randy Nulman


      --- In ccd-newastro@yahoogroups.com, "Adam Block" <ngc1535@...>
      wrote:
      >
      > Frank,
      >
      > As an example, if you were imaging a red nebula you certainly
      would not want to
      > match the background counts in the other filters. The background
      is truly non-neutral
      > which is the implied assumption I think you are making with your
      question.
      >
      > A better way is to use a reference to determine your system color
      normalization
      > coefficients.
      >
      > However, this brings me to Ron. In your verbose answer... you did
      not mention the
      > G2V solar analogue as a quick way to do this. It would be
      independant of other
      > references if you agree to the premise of the technique. How come
      no mention?
      >
      > Adam
      >
      >
      >
      > ------------------------------------------------------------
      >
      > Adam Block
      > ngc1535@...
      > http://caelumobservatory.com/
      > Caelum videre iussit, et erectos ad sidera tollere vultus.
      > He bid them look at the sky and lift their faces to the stars.
      > --Ovid, Metamorphoses 1. 85-8
      > ------------------------------------------------------------
      >
      >
      >
      > -----Original Message-----
      > From: Wodaski Yahoo [mailto:yahoo@...]
      > Sent: Wednesday, August 2, 2006 09:01 AM
      > To: ccd-newastro@yahoogroups.com
      > Subject: Re: [ccd-newastro] RGB background ADU counts
      >
      > Not exactly.
      >
      > The actual goal is equal signal to noise ratio (S/N). This does
      not
      > always equate to equal signal, since the noise can vary.
      >
      > For example, if your have light pollution, it is not going to be
      equally
      > bright through each filter. It will typically be brightest through
      your
      > green filter. So the green ADU needs to be higher to get equal S/N.
      >
      > It can be quite challenging to compute exposures required for
      equal S/N.
      > So an empirical approach to the problem can be a good approach -
      adjust
      > your exposure times until the color is accurate. I've used this
      with
      > many camera/scope/filter wheel combinations with good success. A
      bit
      > time consuming, but you can't beat empirical data for reaching a
      working
      > conclusion. <G>
      >
      > The way I do it is to start out guessing at exposure ratios, based
      on
      > whatever information I have at hand. I shoot a target with well-
      known
      > and balanced color, typically a face-on galaxy. M81 has too subtle
      a
      > color palette to be a good example; you need a galaxy with a large
      red
      > core, substantial starbirth in the arms, and enough dust to give
      you an
      > accurate read on those colors. Nebulae are often domination by
      emission,
      > and are not useful for this type of exercise.
      >
      > You then adjust your exposure ratios as needed. You can use MaxIm
      DL or
      > Photoshop to quantify the color imbalance, and make a more
      educated
      > guess at your next set of exposure ratios. For example, if the red
      is
      > 110% of what it should be, then your next red exposure should be
      90% of
      > the last one.
      >
      > Noise is what creates the color imbalance, so what you are doing
      by
      > finding exposure times that provide color balance is making the
      S/N
      > levels in the color images similar enough to be able to tweak the
      image
      > into final balance. Balance will vary with elevation, sky
      conditions,
      > sky brightness, etc., but this method will get you close.
      >
      > This gets you what I would call a practical color balance.
      Subjects with
      > very low signal in one color (e.g., a galaxy with very little
      starbirth)
      > will require longer total exposure times to get accurate color.
      But
      > that's true anyway, due to the inherent lower signal (less signal,
      lower
      > S/N, less color accuracy).
      >
      > Ron Wodaski
      >
      > Photon Collector wrote:
      > > When imaging an object using RGB filters, is the goal to achieve
      similar background ADU counts for each color filter?
      > >
      > > ---Frank Rocketman Uroda
      > >
      > > [Non-text portions of this message have been removed]
      > >
      > >
      > >
      > >
      > >
      > > Yahoo! Groups Links
      > >
      > >
      > >
      > >
      > >
      > >
      > >
      >
      >
      >
      >
      >
      > [Non-text portions of this message have been removed]
      >
    • Terrence R. Redding, Ph.D.
      Ron, you are a true gentleman to take so much time to write such a complete response. As I read, I started the google search as you suggested, and indeed
      Message 43 of 43 , Aug 13, 2006
      • 0 Attachment
        Ron, you are a true gentleman to take so much time to write such a
        complete response. As I read, I started the google search as you
        suggested, and indeed started to find a great deal of information
        about the nature of M31.

        The math is beyond me - being a social scientist - but I am
        considering taking several courses in astronomy with an effort to
        learn the math. Indeed, this past week I had a chance to visit the
        University of North Florida bookstore and purchased a text on an
        introduction to astronomy. I have also reviewed the courses
        available here in the Palm Beaches and hope to take a few at the
        local community college, which surprisingly has a credible astronomy
        program.

        Thank you again.

        Terry
        On Aug 13, 2006, at 12:41 PM, Wodaski Yahoo wrote:

        > More accurately: color perception is arbitrary.
        >
        > One reason you may not see a discussion of the meaning of that
        > particular color change in M31's structure is that you are looking at
        > very incomplete data. In order to analyze structure, astro-physicists
        > take much more data into consideration. Simply put, visual images of
        > galaxies do show some small structures accurately, but they are often
        > quite misleading with respect to larger structures.
        >
        > In order to assess large-scale structures (and some small structures),
        > additional information is required. This typically comes from:
        >
        > * Radio data
        > * X-ray data
        > * UV data
        > * IR data
        >
        > Each of these wavelength regions has special characteristics that
        > allow
        > the observer to learn more. For example, long-IR photons penetrate
        > dust
        > (which no visible wavelengths do). Our cameras can't even record this
        > wavelength because they are not sensitive to it (though one can do a
        > small amount of dust penetration imaging with shorter IR wavelengths).
        > Deep IR images of galaxies have revealed hidden structure that has
        > been
        > most revealing. Such images exist for M31; you can search for them
        > with
        > Google. There is one image that I recall; it is a combination of IR
        > and
        > visual data which illustrates how different they are.
        >
        > Just as it is best not to get too excited if one comes up with a
        > "common-sense" refutation or expansion of relativity theory, it is
        > best
        > not to assume that something you see visually implies something about
        > the large-scale structure of a galaxy. Visual data can be misleading
        > because it is so incomplete. If you want to confirm what you see, or
        > think you see, your best bet is to start digging through the
        > literature
        > and see what research has been done in other wavelengths to confirm or
        > reject your hypothesis. Few amateurs work on this type of stuff, so
        > your
        > best bet would be the astro-physical journals. Be prepared to do a lot
        > of homework on this kind of stuff; it can take days to track down the
        > relevant literature, and when you do you may find that the article was
        > published in a journal that charges even for the Internet version
        > of the
        > paper (or it may require an annual subscription, which may run from
        > $100
        > to $3000 or more).
        >
        > Not trying to discourage you; just trying to paint a picture of the
        > way
        > things work. I occasionally pursue areas of investigation, and I now
        > know that I need to have a lot of time available to do it. It can be
        > fun, but be prepared to meet new concepts and sometimes-daunting math.
        > That is how you can spend a lot of time on this - you may, for
        > example,
        > discover a paper that shows the math for density waves in a disk-
        > shaped
        > galaxy, but you may not have the math background to understand what
        > the
        > paper is saying. You can then spend several days learning enough to
        > grasp what the paper has to tell you - and it may or may not be in
        > your
        > area of interest! Still, you can learn a lot this way. Just be
        > prepared
        > to invest the time and mental energy required.
        >
        > Ron
        >
        > Terrence R. Redding, Ph.D. wrote:
        > > Ron, good morning.
        > >
        > > Granted, all color to some degree is arbitrary. However, as I
        > > understand it, we can use color to classify emissions and thus the
        > > same color to identify the same emissions. not something I know
        > > anything about, but something that fascinates me none the less. For
        > > example, this image http://www.budguinn.com/gallery/main.php?
        > > g2_itemId=174&g2_imageViewsIndex=1 has a ring of red to pink hues
        > > encircling M31. I have yet to see it discussed, but the pattern
        > > suggests to me some attribute of the galaxy's formation.
        > >
        > > The repetative and consistent informational value associated with
        > > color caused me to balk briefly at your reference to the arbitrary
        > > nature of color.
        > > On Aug 8, 2006, at 3:40 PM, Wodaski Yahoo wrote:
        > >
        > >> All color is, to some degree arbitrary. For example, the human to
        > >> human
        > >> variations are quite large - and that doesn't even include the
        > various
        > >> forms of color blindness.
        > >>
        > >> And of course we are sampling millions of colors, some of them
        > >> emission
        > >> lines, with three pretty wide (and overlapping) filters.
        > >>
        > >> So, IMO, we hardly need to even tip our hat to objectivity. Ok,
        > >> maybe a
        > >> little. <G>
        > >>
        > >> Humor aside, we know the apparent color of many structures in
        > >> galaxies.
        > >> If one makes the apparent color of a variety of such structures
        > >> correct,
        > >> then one has a good color balance.
        > >>
        > >> And remember that the point is to get color data that has
        > roughly the
        > >> same S/N, so that you can tweak it for variable conditions - be
        > >> that sky
        > >> glow, light pollution, moonlight, elevation of the object, etc.
        > >> Some of
        > >> those are even measurable and you can adjust your ratios
        > accordingly.
        > >>
        > >> The method is only subjective to the extent that you fail to
        > work hard
        > >> enough to find color exposure ratios that provide accurate apparent
        > >> color.
        > >>
        > >> Ron Wodaski
        > >>
        > >> Photon Collector wrote:
        > >>> "The way I do it is to start out guessing at exposure ratios,
        > >> based on
        > >>> whatever information I have at hand. I shoot a target with well-
        > >> known
        > >>> and balanced color, typically a face-on galaxy. "
        > >>>
        > >>> This sounds very subjective to me. It sounds like you're basing
        > >> your colors on someone else's possibly subjective rendition of a
        > >> galaxy. Am I just misunderstanding your point?
        > >>> ---Frank Rocketman Uroda
        > >>>
        > >>>
        > >>> ----- Original Message -----
        > >>> From: Wodaski Yahoo
        > >>> To: ccd-newastro@yahoogroups.com
        > >>> Sent: Wednesday, August 02, 2006 12:01 PM
        > >>> Subject: Re: [ccd-newastro] RGB background ADU counts
        > >>>
        > >>>
        > >>> Not exactly.
        > >>>
        > >>> The actual goal is equal signal to noise ratio (S/N). This does
        > not
        > >>> always equate to equal signal, since the noise can vary.
        > >>>
        > >>> For example, if your have light pollution, it is not going to be
        > >> equally
        > >>> bright through each filter. It will typically be brightest
        > >> through your
        > >>> green filter. So the green ADU needs to be higher to get equal
        > S/N.
        > >>>
        > >>> It can be quite challenging to compute exposures required for
        > >> equal S/N.
        > >>> So an empirical approach to the problem can be a good approach -
        > >> adjust
        > >>> your exposure times until the color is accurate. I've used this
        > with
        > >>> many camera/scope/filter wheel combinations with good success.
        > A bit
        > >>> time consuming, but you can't beat empirical data for reaching a
        > >> working
        > >>> conclusion. <G>
        > >>>
        > >>> The way I do it is to start out guessing at exposure ratios,
        > >> based on
        > >>> whatever information I have at hand. I shoot a target with well-
        > >> known
        > >>> and balanced color, typically a face-on galaxy. M81 has too
        > subtle a
        > >>> color palette to be a good example; you need a galaxy with a
        > >> large red
        > >>> core, substantial starbirth in the arms, and enough dust to give
        > >> you an
        > >>> accurate read on those colors. Nebulae are often domination by
        > >> emission,
        > >>> and are not useful for this type of exercise.
        > >>>
        > >>> You then adjust your exposure ratios as needed. You can use MaxIm
        > >> DL or
        > >>> Photoshop to quantify the color imbalance, and make a more
        > educated
        > >>> guess at your next set of exposure ratios. For example, if the
        > >> red is
        > >>> 110% of what it should be, then your next red exposure should be
        > >> 90% of
        > >>> the last one.
        > >>>
        > >>> Noise is what creates the color imbalance, so what you are
        > doing by
        > >>> finding exposure times that provide color balance is making the
        > S/N
        > >>> levels in the color images similar enough to be able to tweak the
        > >> image
        > >>> into final balance. Balance will vary with elevation, sky
        > >> conditions,
        > >>> sky brightness, etc., but this method will get you close.
        > >>>
        > >>> This gets you what I would call a practical color balance.
        > >> Subjects with
        > >>> very low signal in one color (e.g., a galaxy with very little
        > >> starbirth)
        > >>> will require longer total exposure times to get accurate color.
        > But
        > >>> that's true anyway, due to the inherent lower signal (less
        > >> signal, lower
        > >>> S/N, less color accuracy).
        > >>>
        > >>> Ron Wodaski
        > >>>
        > >>> Photon Collector wrote:
        > >>>> When imaging an object using RGB filters, is the goal to
        > >> achieve similar background ADU counts for each color filter?
        > >>>> ---Frank Rocketman Uroda
        > >>>>
        > >>>> [Non-text portions of this message have been removed]
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>> Yahoo! Groups Links
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>>
        > >>>
        > >>>
        > >>>
        > >>> [Non-text portions of this message have been removed]
        > >>>
        > >>>
        > >>>
        > >>>
        > >>>
        > >>> Yahoo! Groups Links
        > >>>
        > >>>
        > >>>
        > >>>
        > >>>
        > >>>
        > >>>
        > >>>
        > >>
        > >
        > >
        > >
        > > [Non-text portions of this message have been removed]
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > > Yahoo! Groups Links
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        >
        >



        [Non-text portions of this message have been removed]
      Your message has been successfully submitted and would be delivered to recipients shortly.