Welcome to Fractal Forums

Hi all, i am trying to implement a HDR mapping technique, but i run into some problems

is there a way to apply high dynamic range mappings without involving big numbers ?

ist there any web resource concerning with that ? i was trying to normalize my results,
but the high values destroy my small values, i have found a description of tone mapping,
but with this algorithm i needed to calc exp(100) or even bigger, and thus, i could not get
a result, how are you doing High Dynamic Range mappings ? ??? ???

here is a link to a method i tried to implement, but i failed ( due to too large numbers )
http://www.gamedev.net/reference/articles/article2208.asp

Your question got me exploring the web for references, and reading some of the many topics that surround "High Dynamic Range Processing".  The first two sites had many references (which I am still checking into):

High Dynamic Range Imaging
    http://en.wikipedia.org/wiki/High_dynamic_range_imaging (http://en.wikipedia.org/wiki/High_dynamic_range_imaging)

High Dynamic Range Rendering
    http://en.wikipedia.org/wiki/High_dynamic_range_rendering (http://en.wikipedia.org/wiki/High_dynamic_range_rendering)

I also found something from "Afrigraph '06" (sponsored by SIGGRAPH)
http://portal.acm.org/citation.cfm?id=1108590.1108612&coll=portal&dl=ACM&type=series&idx=1108590&part=Proceedings&WantType=Proceedings&title=Computer%20graphics%2C%20virtual%20reality%2C%20visualisation%20and%20interaction%20in%20Africa
It is a PDF file on "Volumetric High Dynamic Range Windowing for Better Data Representation"
    http://portal.acm.org/ft_gateway.cfm?id=1108612&type=pdf&coll=portal&dl=ACM&CFID=10758382&CFTOKEN=51701067 (http://portal.acm.org/ft_gateway.cfm?id=1108612&type=pdf&coll=portal&dl=ACM&CFID=10758382&CFTOKEN=51701067)
    http://w210.ub.uni-tuebingen.de/dbt/volltexte/2005/1753/pdf/wsi-2005-03.pdf (http://w210.ub.uni-tuebingen.de/dbt/volltexte/2005/1753/pdf/wsi-2005-03.pdf)

thank you for searching, i was exploring that wikipedia site also,

all in all, i do not need a HDR algorithm, i need rather a Tone Mapping method, i have implemented the most simple method
to map big values down to 0..1

in the german wikipedia they stated that the most easy formula one could use is

x/(x+1)

if x is the big number, it will converge against one, slowly, so, i simply apply that to my results, and after
that i scale them up with normalizing ... works quite well

watch http://www.fractalforums.com/index.php?topic=518.0 for first results

afrigraph is the south african siggraph chapter; it's the only graphics conference i've attended (unfortunately), and i've attended it twice :) anyway i don't think "sponsored" is really how it works, just wanted to point out!


regarding tonemapping, i have no idea why you'd want to compute exp(100), tonemapping is supposed to be dynamic range compression and would more likely make use of a logarithm than an exponential. judging from the renders i've seen posted, mutatorichhabedasdeutschenamevergessen doesn't really need any of that postprocessing stuff though, what kind of ranges are in your source signal?

producing antialiased high dynamic range images is kinda tricky too btw.

the exp was needed for calculating the log average, but i do not need it anymore, i am just taking x/(x+1) from my values ...

Thomas Ludwig (lycium) wrote:
>
>    >
>    .....i don't think "sponsored" is really how it works, just wanted to point out!

Then may I suggest you tell the people that created the web page in which I gave the link above, for that is what it is saying on their site.       :D   ;D   ;)
 
 

*giggle* well if it's so then it's so :) i was fortunate enough to meet the very nice lady (whose name i've forgotten, even though she gave me fridge magnets for "quaternion" etc!) who headed siggraph at the time of the first afrigraph, so i really should have known better!

i just took a look around that page and noticed that they're accepting both papers and art submissions this year... most interesting :)

You have quaternion fridge magnets?  Cool.  I want some.  From now on, when someone asks me what fractals are good for, I'll say they keep shopping lists stuck to the refrigerator.

Eric Bigas wrote:
>
>    You have quaternion fridge magnets?  Cool.  I want some.
>    From now on, when someone asks me what fractals are
>    good for, I'll say they keep shopping lists stuck to the refrigerator.

Then you should see my collection of fractal coffee mugs, mouse pads, stationary, etc....    ;)   :D   ;D


 

hmm this is getting really off-topic now ;)

@paul: i for one would love to see your collection of fractal-related (physical) stuff!

@eric: technically they're rendering words, like "teapot" "ray trace" "radiosity" "hierarchical" and of course "quaternion" :) they mix interestingly with those sex-oriented magnets :P