11/4/2004 Tone Reproduction Luminance levels A media based - PDF document

11/4/2004 Tone Reproduction • Luminance levels A media based framework for color and tone reproduction in Computer Graphics Joe Geigel Range of Sky = 12400 nits Department of Computer Science lighting in Rochester Institute of Technology display Trees = 64 nits environment jmg@cs.rit.edu Range of lighting in real world 11/ 04/2004 A media based framework for color and tone reproduction in 2 Computer Graphics High Dynamic Range (HDR) I maging I mage Synthesis in Computer Graphics real photo Photographic Photography: camera scene processing print • high dynamic range imaging is a set of techniques that allow a far greater dynamic range of exposures than normal digital processing imaging techniques. • The intention is to accurately represent the Computer 3D camera tone synthetic wide range of intensity levels found in real Graphics: models model reproduction image scenes, ranging from direct sunlight to the deepest shadows. Reinterpretation of scene Scene luminance Wikipedia optimized for viewing (24 bit RGB) 11/ 04/2004 A media based framework for color and tone reproduction in 3 11/ 04/2004 A media based framework for color and tone reproduction in 4 Computer Graphics Computer Graphics The need for Tone Reproduction Tone Reproduction in Computer Graphics • Simple Linear tone reproduction Light source = Searchlight Light source = firefly [Ferwerda 1998] [Tumblin 1993] 11/ 04/2004 A media based framework for color and tone reproduction in 5 11/ 04/2004 A media based framework for color and tone reproduction in 6 Computer Graphics Computer Graphics 1

11/4/2004 Approaches to Tone Reproduction in CG Media Based Tone Reproduction • Perceptual Based Operators – Based on Perceptual Studies and Models of Human Perception Simulated response of – Goal: Tone Mapping to reproduce how humans perceive scene that media • Media Based Operator – Observation 1: Tone Reproduction problem is not unique to CG As viewed • Over a century of advanced study on a • Scientists and artists have devised processes and techniques given • Techniques optimized for particular display devices piece of – Observation 2: We may not always want what we see! media • Artistic photography (Ansel Adams) • Compositing with content on existing media • Parameter space of practioners – Our Goal: • Achieve tone reproduction via simulation of existing processes • Perception not ignored, rather already considered 11/ 04/2004 A media based framework for color and tone reproduction in 7 11/ 04/2004 A media based framework for color and tone reproduction in 8 Computer Graphics Computer Graphics Media Based Tone Reproduction Media Based Tone Reproduction Framework • Simulation HDR – Model and simulate response of light entering the system Image • Capture (Camera) • Media Response – Brightness – Spectral Camera Media Presentation – Acuity – Noise / Grain • Media Presentation simulation • Color Management Output – Definition: A system for ensuring that colors remain the same Image device regardless of the device or medium used to display the colors. Appearance Device – Model the device on which the image will be displayed • Brightness Capabilities color management • Color Gamut • Viewing Environment 11/ 04/2004 A media based framework for color and tone reproduction in 9 11/ 04/2004 A media based framework for color and tone reproduction in 10 Computer Graphics Computer Graphics Example: Print Photography Example: Print Photography • Follow the path of light from scene to photo to viewer! • Print photography process Radiance / luminance enlarger film Camera Film Process Negative print Printer Paper Print Process viewer camera scene exposure exposure Processed Photographic Radiance / luminance Optics Photographic Material Material 11/ 04/2004 A media based framework for color and tone reproduction in 11 11/ 04/2004 A media based framework for color and tone reproduction in 12 Computer Graphics Computer Graphics 2

11/4/2004 Photographic Units Media Based Tone Reproduction Framework exposure density HDR Image E = It D = log (O) I = Illuminance (lux) O = opacity = 1 / T t = time (sec) T = transmission Media Presentation Camera E = exposure (lux-sec) = I t / I o I t = transmitted light simulation I o = incident light Output Device Image defines the amount of A logarithmic means for device Appearance light hitting the describing transparency or photographic material at reflectance once the material is each point developed color management 11/ 04/2004 A media based framework for color and tone reproduction in 13 11/ 04/2004 A media based framework for color and tone reproduction in 14 Computer Graphics Computer Graphics Camera Model Media Based Tone Reproduction Framework Scene luminance to exposure HDR Image luminance from scene Lens transmittance Camera Media Presentation π ′ = τ θ + η 4 Exposure ( ) ( cos ) x L I t f 2 simulation 4 n Output Device Image Aperture / f-stop flare shutter time device Appearance eff. color management set by photographer 11/ 04/2004 A media based framework for color and tone reproduction in 15 11/ 04/2004 A media based framework for color and tone reproduction in 16 Computer Graphics Computer Graphics Photographic Response: Brightness Photographic Response: Brightness • A typical brightness response / characteristic curve Effects of speed III I - toe II - straight line y t IV section i s III - shoulder n Original 100 Speed Film 400 Speed Film 800 Speed Film II e γ IV - area of D Effects of gamma solarization I S Log Exposure speed - indicates gamma - slope of region Original Low Contrast sensitivity to light Medium Contrast High Contrast II. gives contrast range 11/ 04/2004 A media based framework for color and tone reproduction in 17 11/ 04/2004 A media based framework for color and tone reproduction in 18 Computer Graphics Computer Graphics 3

11/4/2004 Photographic Response: Spectral Photographic Response: Acuity point spread function modulation transfer function 100 80 200 60 100 panchromatic (%) 0 40 20 100 0 blue sensitive 0 40 80 120 0 spatial freq. (cycles/mm) 300 400 500 600 11/ 04/2004 A media based framework for color and tone reproduction in 19 11/ 04/2004 A media based framework for color and tone reproduction in 20 Computer Graphics Computer Graphics Photographic Response: Grain Modeling Photographic Response input image Graininess increases with magnification spectral expose resolution sensitivity exposure convert to density granularity transmission/ response 4X zoom reflection No magnification 2X zoom negative density conversion or print 11/ 04/2004 A media based framework for color and tone reproduction in 21 11/ 04/2004 A media based framework for color and tone reproduction in 22 Computer Graphics Computer Graphics Color Photography Media Based Tone Reproduction Framework HDR Image Camera Media Presentation simulation Output Device Image device Appearance color management 11/ 04/2004 A media based framework for color and tone reproduction in 23 11/ 04/2004 A media based framework for color and tone reproduction in 24 Computer Graphics Computer Graphics 4

11/4/2004 Photography: Presentation Media Based Tone Reproduction Framework • Photographic Model provides: HDR – Reflective densities for each emulsion layer Image – Spectral qualities of dyes on each layer • Image isn’t viewed unless illuminated Camera Media Presentation – Illuminant • Brightness • Spectral Qualities – Viewing Environment simulation • Brightness Output • Spectral Qualities Image Device device Appearance • After presentation, we have specification of – Image as viewed by a given light in a given environment color management – Must drive display device to match perceptual response 11/ 04/2004 A media based framework for color and tone reproduction in 25 11/ 04/2004 A media based framework for color and tone reproduction in 26 Computer Graphics Computer Graphics Device Modeling Media Based Tone Reproduction Framework • Device Characterization and Calibration HDR – Non-linear response (gamma) Image – Device independent color specification – Gamut mapping Camera Media Presentation simulation Output Image device Device Appearance color management 11/ 04/2004 A media based framework for color and tone reproduction in 27 11/ 04/2004 A media based framework for color and tone reproduction in 28 Computer Graphics Computer Graphics Viewing Environment General Brightness Adaptation • Viewing environment can affect image perception • Note also…differences in acuity • Adaptation • The process by which the visual mechanism adjusts to the conditions under which the eyes are exposed to radiant energy. – General brightness adaptation • Adjustments in response to the overall level of stimulus exposed – Lateral brightness adaptation • Adjustments in response due to stimulus in adjacent areas of the retina – Chromatic adaptation • Adjustments in response to the average chromaticity in the stimulus. [Ferwerda 1996] 11/ 04/2004 A media based framework for color and tone reproduction in 29 11/ 04/2004 A media based framework for color and tone reproduction in 30 Computer Graphics Computer Graphics 5