A Brief on HDRI ( High Dynamic Range Imaging)

By: Brendan Fitzgerald Presented By The Digital Media Club 1/23/08 www.dmcscad.com

First and foremost I just want to say that if this subject interest you, you should definitely think about taking Professor Kirt Wittes HDR class. Even though I am sure some of you have experimented with HDR, it is complicated and to really understand what is going on it really takes some studying. I still feel very much like a newbie in this area. I do not consider myself an authority on HDR but more of an enthusiast. Please if you what to learn more there is a great book called The HDRI Handbook by Christian Bloch or speak with Professor Kirt Witte. Im sure that if I wasnt able to answer your questions they will.

What is HDRI? HDRI or High Dynamic Range Imaging is a photographic process that allows the photographer to capture an extreme range of light. This process creates an image that is not only closer to the dynamic range the human eye sees but also can supercede the dynamic range of light that a human eye is capable of seeing. Think you have seen an HDRI? Most likely not, an HDRI is a 32 bit or image or floating point image. This means that the information contained within an HDRI surpasses the viewable bit depth of all current viewing devices, save HDR Monitors. So how do we view an HDRI on current equipment? Well there are 2 ways really. One way would be to tone map the image in which you are compressing all the information from a 32bit scale down to an 8 bit scale. The second way is you open an HDR and cycle through the Exposure Values in one of the many programs capable of opening an HDR. Photography Here is Some basic photography information when thinking about shooting an HDRI. To shoot an HDRI all you need is to have several bracketed exposures of your object or environment. With that said obviously the more brackets in your HDRI the greater the dynamic range. Since my camera has an Auto bracketing feature I generally shoot 3 exposures for my HDRI, One

normal exposure and then 2EV+ / 2EV-. This way is pretty much the bare minimum and in certain lighting situations more EVs may be required. Also when shooting HDRI it is important that you lock in your color balance (white balance thought this can also be changed in post if shooting RAW), Depth of Field or Fstop and the Focus. It is a good Idea if you are using an DSLR to set your focus to infinity. ISO: The ISO setting on a camera determines how sensitive your CCD chip is to light. A higher ISO will make the CCD more sensitive to light but will produce more grain in the image. ISO100=low ISO800=High Exposure time (shutter): This setting determines how long your CCD is exposed to capture the image you are shooting. Faster shutter speeds are generally used for quick moving objects so that the photographer can cut out or down on motion blur. With that said the shutter is a great way to control the amount of light that you let into the camera. With HDRI, manipulating the shutter speed is the best way to bracket your exposure. F-Stop (Aperture): The Aperture on a camera acts in the same way as the iris of the eye. When the Aperture is closed very small (I.e. F22) very little amount of light may be let in (depending on your shutter speed). The Aperture also serves another function then controlling your light, it also controls your depth of field. An aperture setting of f22 will have a high depth of field in which everything will be in focus, were as an aperture setting of f2.8 will have a shallow depth of field and only a selected area will be in focus. For Shooting textures and HDRI environments an aperture of f11-f16 is desirable because it gives you a good depth of field and also a fair amount of light to pass. EV(Exposure Value): Exposure value is something that you might see on your consumer camera and also it shows up in the metadata of Raw files from high end cameras. Exposure Value is just a number that determines the amount of light hitting the sensor by combining the shutter speed and the aperture. HDR Applications Here is a list of some HDR applications: Adobe Photoshop good for generating HDRIs and Tone Mapping also since the advent of CS3 you can now use the stamp tool on 32bit images.

Photomatix- Great for tone mapping and HDR Generation (this is what I use for HDR generation and Tone Mapping. HDR Shop- HDR shop was created by the father of HDR Paul Debevec and is great for panoramic transformations. Smart IBL- I havent used this yet but the Author of The HDRI Handbook Christian Bloch likes it so, its on my list. Basically its an application for the 3d side of HDR and helps you generate your different Maps (Light Map, Reflection Map, Background plate) Sticher- Sticher is a great program and while I know it has some competitors it is the program that I used to stitch together High-Res HDR Environments. Low-res HDR Enviroment A Low-Res HDR Environment is one that is shot using the chrome ball or gazing ball. The idea behind the chrome ball is that even though you are looking at the sphere straight on you are actually seeing an image reflected in the sphere that is a full 360 degrees. So, this means that if you where to stick your hand behind the sphere and then photograph the front when you unwrapped this image you would see your hand. This type of environment generation is really only good for acquiring HDR lighting, not reflections or backgrounds. This is the Quick, down and dirty method of HDR Environments but still very effective for the right project. Execution- You want to be as far from the sphere as possible while still filling your frame and without cutting off any part of the sphere. Then start shooting the varying exposures, you may start wherever you like but I like to get the normal exposure first. After that you need to take all those images and generate an HDR, crop the image to the edges of the sphere and then save the file. Open up HDR shop and go to panoramic transformations and convert it into a lat long file. From there you can bring that into Maya. High-res HDR Environment A High-Res HDR Environment is an HDRI that is shot using a spherical VR Head. The short and the sweet of this is that a Spherical VR Head allows the photographer to center the Nodal point of the camera over the center of the tripod. The Nodal point is the point in the camera in which it Sees from by centering this over the tripod you are able to eliminate

parallax or perspective change. Thus (barring environment changes during your shoot) this is like taking a picture from one place in space in all directions at once. Though this process is much more labor intensive and requires a lot more Hard Drive space, the artist can generate all the information they would need for there scene from this file. Execution- first you need access to a Spherical VR Head and a DSLR camera. We have this equipment in the cage, but you have to have taken Kirts HDR class to use it. Once you have secured your equipment you will know how to use it and execute this process. Also, by that time Im sure there will be a better way then the workflow that I use now. Tone Mapping Tone mapping or converting an HDRI to an LDRI is a good way for an artist to select the image that he wants. This is a necessary evil right now as the technology that is widely available doesnt support the output of HDRI. So, for instance if you shoot a background plate as an HDR to comp your CG object into. This will allow you to go back later and almost choose the lighting that you want to better fit your composite. Tone Mapping is not so much a science as an art, each image you tone map will present different challenges. Whatever the case I always advise people to hang on the the HDRI. Maya, HDR, and other 3d apps Ok this will be the shortest part of document as this is the area that I am the weakest in. Basically though HDRI is not an end all answer you dont just apply it and get instant goodness just like any other lighting technique it requires finesse and sometimes more time because it is so CPU intensive. Though when executed right the results are amazing. With any use of an High-Res HDR environment you want to make sure that you have 3 different maps and they are: Light Map- this is going to be the file that you cast light from and generally speaking this should be very small (i.e. 64x32px) you will only use this for lighting your scene. Reflection Map-The Reflection map will be a medium resolution file that you will generate your reflections from. The size of this map will be determined by the final resolution of your project.

Background map- With a High-Res HDR Environment you can use it as a background. The resolution of this will also be determined by your final output. If you are shooting a chrome ball or Low-Res HDR Environment then you will only need one map, the Light Map. In Maya you will need to use Mental Ray as Maya software is incapable of rendering HDR lighting. In the Mental Ray Render Globals go to the very bottom where you will see environment. Expand that tab and under there you will see IBL or Image Based Lighting, click create. From these settings you want to load in your light map. After that.. have fun there is way to much for me to go over in this brief but I have attached to this Brief an information from Maya Help docs on what all the options do. I will tell you, that once you Emit Light the samples boxes have a lot to do with the quality of your shadows.

Links www.theothersavannah.com http://facultypages.scad.edu/~kwitte/ www.debevec.org www.hdrlabs.com www.hdri-handbook.com

info Brendan Fitzgerald Digital Media Club President www.crazy8studio.com [email protected] [email protected]