HDR: background  

 

  HDRI - a brief overview   
     


Well, most of you will know about the principal advantages of HDR Images. But what exactly means HDR? Whatīs the difference between HDRI and common RGB Images?
 

 

 

  What is HDRI ?  
     



As you might know each Pixel of a digitized Image normally is represented by three color-values known as RGB. RGB stands for Red, Green and Blue, the three base colors of the additive color space used by computers. With these three colors it is possible to mix any color within this colorspace. Each of these colors is represented by an integer number with
a range of 0-255. That means, it is possible to use 256 different brightnesses for each of these colors respectively for the Image. Image-formats based on this RGB range are known as LDRIs, Low Dynamic Ranges Images because they are bound to the 0-255 integer range.

Now, when you take a look at the real world, mother nature isnīt bound to a range when presenting itīs beauty to the human eye. There you have the darkness of a cave with almost no light (at least in the visible spectrum) but also the brightness of an object like the sun which is millions of times brighter than normal daylight, filtered through the atmosphere.

In 1985 Greg Ward invented the RGBE format to be able to work with real world ligthing in his rendering system RADIANCE. The RGBE format basically uses a 4th component known as the shared exponent (E). With the help of this exponent it was possible to describe the actuall brightness of a pixel in a real world brightness range (High Dynamic Range). 

In the years after that invention there where developed several different formats to represent the real world brightness information inside of a digital picture. E.g. Bill Reeves of Pixar created a 33-bit long RGB format , Dan McCoy added it in 1996 to his free TIFF library (libtiff). Greg Ward later added the so called LogLuv format to this library which captures 5 orders of magnitude and the full visible gammut in 24 bits using a perceptual color encoding (There is also a 32-bit version f this format which s able to hold up to 38 orders of magnitude and often results in smaller files due to run-length encoding).

The HDR format of the pictures in this database uses the RGBE format explained above.

Up
 

 

 

  Comparision of ligthing information in RGB and HDR images  
     



What happens if you take an image of a desklamp and try to darken it so that you can see more information inside of the bulb?

F-Stop range within a RGB-image
The RGB picture suffers from having not enough ligthing information.

F-Stop range within a HDR-image
In the HDR-image you can  change the fstop from the bulb without sacrificing contrast and detail..


In the example above you can see that HDR definitely contains much more information about the lighting of your scene than a standard RGB image. With this information it is possible to create much more realistic Lighting not only in 3D scenes but in every artwork which is affected by lighting situations.

The next big advantage especially for the 3D artist is the ability to create much more convincing Motion Blur effects and Reflections on your objects.

Up
 

 

 

  Comparision of lightness and shadow detail in RGB and HDR images  
     
 

Watch the detail in the different f-stop steps:

Detail in lightness areas Detail in shadow areas
 
     
  In the daker f-stop steps you can clearly see the detail in the lightness areas. When pushing the f-stop higher you can see the firstly missing detail in the shadow areas.

 
  Up

 

 

  HDR-images and Reflections  
     
 

This reflecting Sphere was rendered with the
help of a RGB image as reflecting environment.

As you can see, the reflections of the
brighter parts are also very dull and the
reflection doesnīt look real.

 
     
 

This reflecting Sphere was rendered with the help
of a HDR image as reflecting environment. ->


Now the brighter parts differ nicely from the other parts. The image has much more life in it than the dull standard version. 

Up

 
 
     

 

  HDR-images and Motion Blur  
     
 

This reflecting Sphere was rendered with the help of a RGB image as reflecting environment. 

As you can see, the reflections of the brighter parts look very dull when motion blurred. It is not possible to e.g. render really bright reflections on top of car laquer in a city environment by night without HDR Images.

 
     
 

This reflecting Sphere was rendered with the help
 of a HDR image as reflecting environment. ->

Now the brighter parts differ nicely from the other parts. The image has much more life in it than the dull standard version.

Up  
 
     

 

  OpenEXR, a new HDR image format  
     
 
OpenEXR is a High Dynamic Range image format just like the HDR format. Nethertheless there are big differencies between these two formats:

Althought the OpenEXR format also was developed in response to the demand for higher color fidelity, it is much more sophisticated than the HDR format which more originates from academical needs.

With OpenEXR, the goal was to develop a format which is more efficient than the standard 32-bit floating point formats. Because the traditional formats do store more dynamic information than needed in the daily work, the files tend to explode in size. When working with motion pictures, you got huge amounts of data that way without really needing all the stored information. OpenEXR uses clever sorting of needed data and also several lossless compression methods, some of which can achieve compression ratios of about 2:1 for images with film grain.

Further on OpenEXR images can store an arbitrary amount of channels, each with different datatypes and they are able to store different attributes like color balance information of digital cameras, etc.

 
  Up

 

 

  Why do we support OpenEXR?  
     



We want to enable our customers to always work with State of the art technology. We are pretty sure that OpenEXR will be THE standard for HighDynamicRange images in the near future.
Because of that, the HDRIbase is available also as OpenEXR-Version.

For the case that you are not sure, if your favourite application will recognize the OpenEXR format in the near future, the bundle of both versions (HDR+OpenEXR) may be of interest for you. Because of that, we decided to sell a bundle of them also to give you the unique chance to profit from the momentary standard in conjunction with the future standard.

Up
 

 

 

  Where to get more information?  
     


Please visit Sachform Technology to get more information about the technology and our spherical mapping products.
 

 
     
 

copyright by  SACHFORM Technology (c)2003-2005 - All Rights Reserved.
All trademarks on this website are owned by their respective holders.