Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
NASA Space

Pioneer Anomaly Solved By 1970s Computer Graphics 169

Posted by CmdrTaco
from the whats-your-vector-victor dept.
Frans Faase updated us on a Pioneer Mystery we've been following for many years: something is tugging Pioneer 10 & 11. A few years ago a theory surfaced but now "A new computer model of the way heat is emitted by various parts of the Pioneer spacecraft, and reflected off others, finally solves one of the biggest mysteries in astrophysics. Previous calculations have only estimated the effect of reflections. A computer modeling technique called Phong shading was used to work out exactly how the the emitted heat is reflected (PDF) and in which direction it ends up traveling. Taking into account the reflections on the antenna seem to make the anomaly disappear."
This discussion has been archived. No new comments can be posted.

Pioneer Anomaly Solved By 1970s Computer Graphics

Comments Filter:
  • To be fair... (Score:5, Informative)

    by cobrausn (1915176) on Thursday March 31, 2011 @10:02AM (#35678116)
    The technique for Phong Shading was introduced in 1973 as an improvement to Gouraud Shading, but was too computationally intensive to be used for graphics back then. This is no longer the case [valvesoftware.com].
    • Re:To be fair... (Score:5, Interesting)

      by BitterKraut (820348) on Thursday March 31, 2011 @10:16AM (#35678270)
      It is also interesting to note that Phong shading is based on an empirical formula. That means it has not been derived from any known (i.e. accepted) "laws of nature". It is used in Computer Graphics because it can be calculated efficiently and approximates what we see or measure closely enough. Strictly speaking, it is not possible to scientifically explain any phenomenon by showing that Phong shading explains it. But as it seems, the whole scenario is so complex that showing its compatibility with the Phong model must already be regarded as a remarkable achievement.
      • the whole scenario is so complex that showing its compatibility with the Phong model must already be regarded as a remarkable achievement.

        I think what is remarkable is that the monolith is able to make its effect on the probes look identical to phong shading.
    • by mooingyak (720677)

      And in 1973 a disappointed Sisqo sang "Let me see that Phong..."

    • The whole "1970s computer graphics technique" thing is sort of silly anyway. If anything, it's surprising that the technique was invented so recently. "Area under curve in Second Life calculated by 1600s mathematics technique..."
    • by dachshund (300733)

      The technique for Phong Shading was introduced in 1973 as an improvement to Gouraud Shading, but was too computationally intensive to be used for graphics back then. This is no longer the case.

      And the saddest thing is that Phong died shortly after completing his dissertation. So he never knew the impact his techniques had on the field.

    • by six (1673)

      The technique for Phong Shading was introduced in 1973 as an improvement to Gouraud Shading, but was too computationally intensive to be used for graphics back then. This is no longer the case [valvesoftware.com].

      It was too computationally intensive for *realtime* rendering in 1973, but clearly not out of reach for the kind of modeling software NASA people were using ...

      Also, it should be noted that realtime phong shading was already common in demos/intros running on 33 MHz 386 CPUs back in the 90s

    • In many domains it was no longer the case at least 20 years ago.
  • Oh, well. It was fun while it lasted.

  • by pyalot (1197273) on Thursday March 31, 2011 @10:08AM (#35678200)
    everybody knows that. A much better aproximation to real life surfaces are the Oren–Nayar or Cook-Torrance models of the family of BRDFs.
    • by cobrausn (1915176)
      Phong is only 'wrong' in graphics if it produces something other than what you expect as the result. Ah, computer graphics. It should be added to that old saying, 'Close only counts with horseshoes, hang grenades, and now, computer graphics.'
      • by pyalot (1197273)
        So Phong is "right" for the probe, because it incidentally matches what they're seeing better? Alright, I propose a better solution, how about we invent some imaginary matter with exotic properties permeating the space, but that can't be seen, which incidentally has exactly the right properties to fit the measured data?

        Saying Phong is right after fitting the calculated data to the measured data just suspect.
        • by cobrausn (1915176)
          I should point out that I was only referring to the 'computer graphics' part of Phong Shading, not the application mentioned in the article. I also find the result suspect - the Phong method is 'right' here only because it fits the data, which is interesting, but proves nothing until they can explain why.
          • by tnk1 (899206) on Thursday March 31, 2011 @12:27PM (#35679720)

            They did explain why. Thermal effects. The only thing that Phong shading did was to remove an obstacle to that hypothesis that was merely due to built-in inaccuracies in how they accounted for those effects initially.

            In other words, the shading increased the accuracy of the calculations and it was found that when it was applied the most likely solution became even more likely.

            While it is true that the Phong solution is still likely "wrong" due to being not perfectly accurate, it's still a lot less wrong than thermal effects uncorrected, and much, much less wrong than assuming aliens or an entirely new discovery in gravitation.

            • by Josef Meixner (1020161) on Thursday March 31, 2011 @02:40PM (#35681608) Homepage

              While it is true that the Phong solution is still likely "wrong" due to being not perfectly accurate, it's still a lot less wrong than thermal effects uncorrected

              The main problem with Phong is, that it can create energy depending on the parameters. Meaning the emitted light can be stronger than the incident light and so in that calculation create thrust out of nowhere.

              Additionally the Pioneer probes are made out of metal, Phong is derived from a model of plastic. The properties of those two materials are quite different, one being a conductor the other an insulator, so the Fresnel equation gives quite different values for the reflective properties, additionally metals are often anisotropic in their reflection capabilities. This has influence on the direction and form of the lobe for the first order effects. I also don't understand, why they didn't use one of the established BRDFs which are at least physically correct.

              I looked through the paper and I see no prove, that the parameters they assume for diffuse and specular reflection don't violate the laws of physics.

              • I actually saw a conference with the authors of the paper this week and they had into account the fact that it was metal and even modeled it from the clone in the Smithsonian. And yes they treated it like an antenna so the radiative energy balance is taken care of. The Phong idea came from a student which i think is great, and the public acknowledgment from the professor was very nice. They even had a live webcast of the whole conference.I will ask if they will publish the recordings for download. http://w [aerospace.ubi.pt]
        • by russotto (537200)

          Alright, I propose a better solution, how about we invent some imaginary matter with exotic properties permeating the space, but that can't be seen, which incidentally has exactly the right properties to fit the measured data?

          You have a Ph.D in Astrophysics, don't you?

          • by gambit3 (463693)

            What does it matter if he does?

            He can't be heard if he DOESN'T have a Ph.D in Astrophysics?

            Conversely, if he DOES have one, does that make his proposal magically valid?

        • by pclminion (145572)

          Whatever the real model is, Phong is an approximation of it. We know it's a decent approximation, because when we use Phong to model the effects of illumination, the results look realistic. Nobody said it's a perfect model. It is apparently quite a bit better than the approximations used previously.

          Even when we do have an exact model of something, we A) still don't know if it's correct, just that it matches observations made so far, and B) when we use a computer to analyze a situation based on that model, w

        • by Chris Burke (6130)

          So Phong is "right" for the probe, because it incidentally matches what they're seeing better?

          You should not treat this question as though it is a rhetorical. You should treat it as though it is an actual question to which you do not know the answer. And then go find out the answer, if it bothers you so much.

    • Re: (Score:3, Insightful)

      by fatphil (181876)
      And the earth isn't round either. It's just a closer approximation to reality than
      saying the earth is flat. Or saying that Pioneer is a spherical cow. Scientists aren't looking for something that is right rather than wrong, they are looking for something that bounds the error term in a significantly tighter way. Phong apparently does this. Presumably any ad-hoc model that approximated reality closer than what was done before would have also decreased the error bounds.

      And Oren-Nayar? Have you mistaken Pionee
    • Everybody? That's a huge and unfounded assumption which is quite likely very wrong both for the general population and for the /. sub-population.
    • When you say "Phong shading is wrong", isn't that just with respect to visible light? Are you sure "Phong is wrong" for thermal radiation?

  • by Intron (870560) on Thursday March 31, 2011 @10:21AM (#35678334)

    ... from Slava Turyshev which describes what they did to model the craft and show that heat could be the culprit.

    http://www.planetary.org/programs/projects/pioneer_anomaly/update_20080519.html [planetary.org]

  • Radiosity being a 90s computer graphics term for calculation how radiation (heat and light) hit surfaces and are absorbed or re-emitted by them. It came from earlier studies on this not relating to computer graphics.

    You can render your radiosity results using phong shading or other shading techniques.

    • You can render your radiosity results using phong shading or other shading techniques.

      You could, but that would be a very stupid thing to do. Why go to the bother of computing an accurate colour for a point on a surface, to then modulate it with an in-accurate plastic surface approximation? That makes absolutely no sense.

  • Imagine how well you'd model this using monte-carlo techniques / ray tracing.

  • by Moof123 (1292134) on Thursday March 31, 2011 @11:47AM (#35679258)

    I RTFA, but didn't find the results of their calculation. The old method yielded 67% of the effect, but they didn't say what the new method resulted in (other than get the "right" answer). Also I'd want to know error bars. Does the new answer +/- error bars overlap with the detected phenomena within the error bars of it's value?

    • by Alsee (515537)

      Does the new answer +/- error bars overlap with the detected phenomena within the error bars of it's value?

      If you read the paper itself [arxiv.org], there's a graph on page 10 showing exactly that. The error bars on the measured size of the effect overlap almost completely with the error bars on this new calculation.

      Direct thermal radiation off the front of the craft explains most of the effect. The "Pioneer anomaly" vanishes completely once you factor in radiation reflected off the antenna dish. It still warrants some more investigation and more papers pinning it down better, but in my opinion this issue can now be pretty

1 Sagan = Billions & Billions

Working...