It is currently Wed, 19-12-18, 1:39 GMT

All times are UTC




Post new topic Reply to topic  [ 11 posts ] 
Author Message
PostPosted: Fri, 13-02-15, 19:27 GMT 
Offline
User avatar

Joined: Mon, 07-01-08, 13:30 GMT
Posts: 350
Location: Switzerland
Research paper detailing the rendering of the black hole in the movie Interstellar. Maybe of interest to a few people around here?

Oliver James, Eugénie von Tunzelmann, Paul Franklin and Kip S Thorne (2015) "Gravitational lensing by spinning black holes in astrophysics, and in the movie Interstellar "

Figure 15 shows what it would have looked like if they had included the effects of colour/intensity shifting (they decided not to incorporate this in the film's rendering on the grounds that it would confuse people... I am not particularly convinced of the merits of this argument but there you go)


Top
 Profile  
 
PostPosted: Fri, 13-02-15, 20:10 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4593
Location: Hamburg, Germany
ajtribick wrote:
Research paper detailing the rendering of the black hole in the movie Interstellar. Maybe of interest to a few people around here?

Oliver James, Eugénie von Tunzelmann, Paul Franklin and Kip S Thorne (2015) "Gravitational lensing by spinning black holes in astrophysics, and in the movie Interstellar "

Figure 15 shows what it would have looked like if they had included the effects of colour/intensity shifting (they decided not to incorporate this in the film's rendering on the grounds that it would confuse people... I am not particularly convinced of the merits of this argument but there you go)


Great, Andrew!

This is indeed very interesting. Gravitational lensing for Black Holes is indeed planned for celestia.Sci.

While the math is availble in various research papers, one needs to estimate before starting, whether ray tracing would be required or whether things could be rendered sufficiently well in real time. The above project will be very useful in this respect, besides offering just a beautiful visualization along with further inspiration...

Many thanks for sharing!
Fridger

_________________
Image


Top
 Profile  
 
PostPosted: Sat, 14-02-15, 5:47 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 431
Location: South Korea
Kip Thorne's results appear to conflict with the original 1979 paper on visualizing rotating black holes by Jean-Pierre Luminet.
This stackexchange post explains the current situation: http://physics.stackexchange.com/a/151311

Fig. 15(c) in the Interstellar paper, which was claimed to be the most realistic:
Attachment:
cqg508751f15_online.jpg
cqg508751f15_online.jpg [ 45.03 KiB | Viewed 3911 times ]


Version by JA Marck based on Jean-Pierre Luminet's paper:
Image

Here is the explanation by Jean-Pierre Luminet:
Quote:
[The Interstellar black hole accretion disk] does not take into account the intrinsic flux properties of the thin accretion disk (the luminosity is not uniform throughout the rings) and the strong Doppler and gravitational spectral shifts induced by the rotation of the disk at relativistic speed. Due to those effects, the apparent flux must be enhanced on the side where the photons are blueshifted and decreased on the opposite side.


Here is a powerpoint by JP Luminet which has some more graphical results by JA Marck: http://www.brijuni-conference.irb.hr/luminet06.ppt
And here is a paper by JA Marck: http://arxiv.org/abs/gr-qc/9505010


Top
 Profile  
 
PostPosted: Sat, 14-02-15, 10:36 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4593
Location: Hamburg, Germany
dirkpitt wrote:
Kip Thorne's results appear to conflict with the original 1979 paper on visualizing rotating black holes by Jean-Pierre Luminet.
This stackexchange post explains the current situation: http://physics.stackexchange.com/a/151311

Thanks, DW, for contributing more info to this topical subject.

Yet, if two such BH lensing visualizations look partially different, it seems too early to me to call this a conflict.

My little plea below is to try to conceptionally separate possible calculational errors from various inequivalent modelling assumptions ...

Did you check carefully the inherent simplifications e.g. about the intricate BH angular momentum structure, and various other possible modelling assumptions in the respective codes?

As young theoretical physicists my wife and I once calculated analytically various implications of the BH angular momentum barrier that is quite unlike to what we are used to in standard quantum mechanics (text book) scenarios. The BH case tends to lead to "Regge like poles" in the complex angular momentum plane and correspondingly to various orbiting phenomena of "test particles" (photons!...) approaching the BH etc...

Given the substantial complexity of BH lensing, it would not be surprising to me if various simplifications had been tacitly implemented here and there...Such steps are often unavoidable (notably in computer codes), be it for reasons of speed or otherwise.

Just to name a familiar, much simpler setting: consider strong lensing of galaxies from lenses that correspond to single galaxies or galactic clusters. In the literature, one finds plenty of inequivalent assumptions about such spacially extended lenses (as the two of us know very well ;-) ).

  • Simplification to spherically or axially symmetric lenses;
  • simplification to single point lenses with the deflection angle alpha_hat matched to the total mass of the lensing galactic cluster/ galaxy under consideration;
  • modelling the galactic cluster as a sum of point lenses;
  • making various assumptions about the radial mass distribution of extended lenses;
  • working strictly in the optical limit i.e. neglecting next order corrections from General Relativity. The latter have all been worked out in the literature...;
  • ...and much more...

Despite significant diagreements in resulting visualizations using these simplifying scenarios, one would not speak of conflicts here. Each of the above scenarios have certain regions of approximate validity the limitations of which should of course be pointed out clearly.

Fridger

_________________
Image


Top
 Profile  
 
PostPosted: Sat, 14-02-15, 15:23 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 431
Location: South Korea
Sure, in fact Thorne does acknowledge the works by JA Marck and JP Luminet and admits that creative decisions led to the less accurate depiction in the movie.

But it's also interesting to note that Fig. 15(c) does in fact take into account Doppler and intensity shifts according to the paper so it's not entirely clear why Marck's and Thorne's sims look so different. For example, Marck's version has an accretion disk that is much more opaque and seemingly more luminous compared to Thorne's.

Thorne has stated in his paper that the BH code does not aim to be the fastest ("fast throughput has been only a secondary concern"). The main "adjustments" claimed in their paper have been to use a slower spin, neglect spectral and flux shifting, and to add glow (lens flare). The other difference is the use of light bundles (rays of finite thickness) instead of zero-thickness rays. So there does not seem to be much simplification going on here.

Theory wise, they use Boyer-Lindquist coordinates instead of Marck's use of Eddington coordinates. It's not clear if this matters though, as both authors reproduce similar high-level effects such as an inner structure with multiple images due to photons near the event horizon making several orbits (!) around the BH before finally escaping. There is a kind of intuitive structure to the images if you know what to look for, so I have some hope that one doesn't have to resort to expensive curved ray tracing... BTW, this should not be confused with multiple images in non-BH lensing, in which a light ray from a distant source follows two (or more) trajectories due to a lensing mass.


Top
 Profile  
 
PostPosted: Sun, 15-02-15, 10:25 GMT 
Offline
User avatar

Joined: Mon, 07-01-08, 13:30 GMT
Posts: 350
Location: Switzerland
The properties of the accretion disc used in the Interstellar simulation were also the result of the creative compromises:
James et al. 2015 wrote:
This artists’ Interstellar disk was chosen to be very anemic compared to the disks that astronomers see around black holes and that astrophysicists model—so the humans who travel near it will not get fried by x-rays and gamma-rays. It is physically thin and marginally optically thick and lies in the black hole’s equatorial plane. It is not currently accreting onto the black hole, and it has cooled to a position-independent temperature T = 4500 K, at which it emits a black-body spectrum.

So this may well have something to do with the observed differences.

From what I can tell, the earlier simulations were for a non-rotating black hole, rather than the spinning black hole used for Interstellar. That particular image from Marck is also aligned exactly with the disc rather than the Interstellar rendering which is from outside the disc plane.


Top
 Profile  
 
PostPosted: Mon, 16-02-15, 3:09 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 431
Location: South Korea
Good points.
I don't know if these parameters are supposed to emerge from other properties of the BH such as rotation speed, mass, but it seems that the optical thickness and temperature of the disk have been arbitrarily chosen in both papers. I wonder if that's a valid thing to do?


Top
 Profile  
 
PostPosted: Tue, 17-02-15, 12:34 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 431
Location: South Korea
I just discovered that Thomas Muller at U. Stuttgart has, back in 2012, already made a very interesting "realtime" implementation of lensing with source code(!) around Schwarzschild (non-rotating) black holes.

http://www.visus.uni-stuttgart.de/index.php?id=2037

Image

The code uses pure OpenGL with a fragment shader and a Qt GUI (which required some tweaking to compile under Qt5).
The innovation is that Muller does not do slow, iterative ray tracing! He computes analytic solutions to the geodesic equation in his shader to determine without integrating entire photon trajectories numerically, where photons hit the accretion disk.

It's very promising, although unfortunately dragging is quite laggy on my GPU (Intel HD 3000).
Also I had to comment out the use of a switch-case statement in the shader since it wouldn't compile under glsl 1.2.
My accretion disk looks incorrect too:

Attachment:
Screen Shot 2015-02-17 at 21.12.00.png
Screen Shot 2015-02-17 at 21.12.00.png [ 112.47 KiB | Viewed 3842 times ]


The author claims 40 fps on a Geforce 8400M GT, which is a GPU from 2007 so other GPUs and/or drivers will certainly fare better.


Top
 Profile  
 
PostPosted: Tue, 17-02-15, 14:06 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4593
Location: Hamburg, Germany
I know a number of interesting BH-related visualization papers by Thomas Mueller & collaborators:

Also in 2011-2012: a Qt-based interactive visualization of the Distortion of the stellar sky by a Schwarzschild Black Hole (together with Daniel Weiskopf)

http://www.vis.uni-stuttgart.de/~muelleta/IntBH/
(with sources!)

I provided more respective info earlier in this CM thread of mine: viewtopic.php?f=7&t=402#p8202

+++++++++++++++++++
Most importantly, you need to first download the 3 videos at the end to your harddisk and play them from there. Use e.g. the VLC player to display the .mp4 format.
+++++++++++++++++++

Download by selecting "Save Link (NOT image!)" after clicking on each of the 3 images...

I also knew the above BH- lensing visualization by him and Joerg Frauendiener:
Here is the publicly accessible arXiv link to the original paper that is quite interesting, as DW already pointed out.

http://arxiv.org/abs/1206.4259

Fridger

_________________
Image


Top
 Profile  
 
PostPosted: Wed, 18-02-15, 3:06 GMT 
Offline
User avatar

Joined: Tue, 04-09-07, 2:32 GMT
Posts: 431
Location: South Korea
Yes! I don't know how I missed these before (although I did reference one of Weiskopf's papers in my ISU lensing work).

Here is another very interesting paper by Eric Agol on fast "sem"-analytic computation of Kerr black hole geodesics:

http://iopscience.iop.org/0004-637X/696/2/1616/article

Image

Fortran source code is available: http://www.astro.washington.edu/users/agol/geokerr/

I don't know Fortran, but the comments suggest that the code is grid-based and thus might translate well to a fragment shader for GPU-accelerated rendering.


Top
 Profile  
 
PostPosted: Wed, 18-02-15, 7:58 GMT 
Offline
Site Admin
User avatar

Joined: Fri, 31-08-07, 7:01 GMT
Posts: 4593
Location: Hamburg, Germany
Quote:
I don't know Fortran,...


I do .. since childhood ;-)

Fridger

_________________
Image


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 11 posts ] 

All times are UTC


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group