Pete Ludé, Mission Rock Digital

Report
LIPA Laserama
Topics on Laser Illuminated Projectors
February 19, 2014
LIPA Membership
Contact LIPA at [email protected]
2/19/14
2
Today’s Agenda
• Regulatory Update:
•
•
Are they legal?
Radiance is the same – lamps and laser projectors
IEC standards updates
• Understanding speckle
•
…and how to measure it
• Laser Color Primary Selection
•
Impacts on Gamut, Image Quality and Efficiency
• Do you see what I see?
•
Color Matching and the Single Observer
• Any Questions?
Contact LIPA at [email protected]
2/19/14
3
Regulatory update
LIP light output = Lamp projector light output
Pete Ludé
Mission Rock Digital, LLC
[email protected]
Study conducted
 LIPA Commissioned Study: Tested optical
characteristics of




Lead Researcher: Dr. David Sliney




35mm film projector
Current Xenon short-arc digital cinema projectors
Prototype laser projectors
Casey Stack, Laser Compliance
Jay Parkinson, Phoenix Laser Safety
David Schnuelle, Dolby Laboratories
Eight projectors tested in various locations over 7
months.
Contact LIPA at [email protected]
2/19/14
5
Hot off the press!
 Published in Health Physics, March 2014




Radiation Safety Journal
Official Journal of the Health Physics Society
Peer review complete
Cover story!
Additional analysis presented at
Society of Motion Picture & Television Engineers
Conference – October 22, 2013.
Contact LIPA at [email protected]
2/19/14
6
Laser Brightness (Radiance)
LARGE FOCAL SPOT
(FILAMENT IMAGE)
LENS
MICROSCOPIC
FOCAL SPOT
(“DIFFRACTION LIMITED”)
LASER
From Sliney DH and Trokel, S, 1993
Contact LIPA at [email protected]
LENS
2/19/14
7
Comparison of Radiance Values
Light Source
Radiance Value
Units
5mW laser pointer
70
MW/m2 sr
The SUN
(visible λ)
7
MW/m2 sr
30,000 lumen
cinema projector
2
MW/m2 sr
Contact LIPA at [email protected]
2/19/14
8
Comparing Radiance: Lamp vs. Laser
40
30
(W • cm-2 • sr -1)
Normalized
Measured Radiance
35
25
20
15
10
5
0
Proj 6
Laser
Actual Luminance Power (Lumens): 5,000
Normalized Luminance Pwr (Lumens): 5,000
Contact LIPA at [email protected]
Proj 2
Proj 1
Proj 4
Proj 5
Xenon
17,000
5,000
Xenon
30,000
5,000
Laser
55,000
5,000
Laser
2,000
5,000
2/19/14
9
Conclusion
Traditional lamp projectors
and new laser-illuminated projectors,
when of equal luminance power,
emit almost identical radiance.
Contact LIPA at [email protected]
2/19/14
10
10
IEC Regulatory Changes
Laser Projector Regulation under IEC
• All laser product
requirements are defined in
60825
•
Medical
•
Industrial
•
Laboratory use
•
Laser Welding
•
Laser Illuminated Projectors
IEC 60825-1 Ed 2 (2007)
Safety of Laser Products
Part 1: Equipment classification & Requirements
Contact LIPA at [email protected]
2/19/14
12
Laser Projector Regulation under IEC
IEC 60825-1 Ed 3 (2014)
Safety of Laser Products
Part 1: Equipment classification & Requirements
IEC 62471 Ed 1 (2006)
Photobiological safety of lamps and lamp systems
Carve-out for devices with radiance < (1 MW•m-2 •sr -1)/α
Contact LIPA at [email protected]
2/19/14
13
Laser Projector Regulation under IEC
IEC 60825-1 Ed 3 (2014)
Safety of Laser Products
Part 1: Equipment classification & Requirements
IEC 62471-5 Ed 1 (2015?)
Photobiological safety of Lamp Systems
for Image Projectors
Contact LIPA at [email protected]
2/19/14
14
US State Laser Regulations
WA
ND
ME
MT
OR
MN
ID
VT
WI
SD
WY
CT
MI
NJ
OH
IN
IL
UT
MD
DE
WV
CO
CA
RI
PA
OA
NB
NV
NH
MA
NY
KS
VA
MO
KY
NC
TN
AZ
OK
NM
SC
AR
MS
GA
AL
No relevant laser regulations
LA
TX
Some relevant laser regulations
AK
FL
HI
0
500 Miles
0 500 Km
Contact LIPA at [email protected]
0
0
500 Miles
500 KM
Most involved & potentially
burdensome
0
100 Miles
0 100 Km
2/19/14
15
Speckle
What is Speckle?
•
Interference pattern that occurs when coherent light is
scattered off an optically rough surface (i.e. screen)
•
Visible noise on uniform areas of scene
•
•
Decreases perceived contrast
•
Most visible on uniform, bright scene elements (e.g. sky)
•
More visible when you move your head back and forth
(“subjective” speckle)
Figure of merit: Speckle contrast ratio
SCR= standard deviation / mean intensity in %
Source: Goodman (8), Curtis (7)
Contact LIPA at [email protected]
Source: K.O. Apeland (5)
•
•
•
Between 0 and 1
0 means “no speckle”
Can be expressed as percentage
2/19/14
17
Methods to reduce speckle
In Theory:
In Practice:
•
• Polarization diversity
• Temporal averaging
Array of multiple emitters
•
Slightly different frequencies
(wavelength diversity)
•
Spatially separated (angle diversity)
• Wavelength diversity
•
Rotating diffusers
• Angle diversity
•
Vibrating diffusers
• Temporal coherence reduction
•
Hadamard matrices
• Spatial coherence reduction
•
Vibrating screen
•
Other methods…
Source: Goodman (8)
Contact LIPA at [email protected]
2/19/14
18
Speckle Metrology Considerations
•
Source (Laser)
•
Projector Focal plane (≠ screen?)
•
Reference light source (coherent)
•
Luminance power (brightness)
Contact LIPA at [email protected]
2/19/14
19
Speckle Metrology Considerations
•
•
Source (Laser)
•
Projector Focal plane (≠ screen?)
•
Reference light source (coherent)
•
Luminance power (brightness)
Camera
•
Clear aperture / f-number
•
Pixel size (relative to speckle size)
•
Focal length (related to distance)
•
Shutter speed / Integration time
•
Focus point (= screen?)
•
Spectral filtering (high/low-pass)
Contact LIPA at [email protected]
2/19/14
20
Speckle Metrology Considerations
•
•
•
Source (Laser)
•
Projector Focal plane (≠ screen?)
•
Reference light source (coherent)
•
Luminance power (brightness)
Camera
•
Clear aperture / f-number
•
Pixel size (relative to speckle size)
•
Focal length (related to distance)
•
Shutter speed / Integration time
•
Focus point (= screen?)
•
Spectral filtering (high/low-pass)
Image Processing
•
Gamma (Optical-Electrical transfer curve)
•
Exposure
•
Compression algorithm
•
Bit depth / dynamic range
•
Contact
LIPA at [email protected]
2/19/14
21
Speckle Metrology Considerations
•
•
•
Source (Laser)
•
Projector Focal plane (≠ screen?)
•
Reference light source (coherent)
•
Luminance power (brightness)
Camera
•
Clear aperture / f-number
•
Pixel size (relative to speckle size)
•
Focal length (related to distance)
•
Shutter speed / Integration time
•
Focus point (= screen?)
•
Spectral filtering (high/low-pass)
Image Processing
•
Gamma (Optical-Electrical transfer curve)
•
Exposure
•
Compression algorithm
•
Bit depth / dynamic range
•
Contact
LIPA at [email protected]
•
Screen
•
•
•
Screen gain
Total Integrated Scatter
Objective (second) screen
2/19/14
22
Speckle Metrology Considerations
•
•
•
Source (Laser)
•
Projector Focal plane (≠ screen?)
•
Reference light source (coherent)
•
Luminance power (brightness)
Camera
•
Clear aperture / f-number
•
Pixel size (relative to speckle size)
•
Focal length (related to distance)
•
Shutter speed / Integration time
•
Focus point (= screen?)
•
Spectral filtering (high/low-pass)
Image Processing
•
Gamma (Optical-Electrical transfer curve)
•
Exposure
•
Compression algorithm
•
Bit depth / dynamic range
•
Contact
LIPA at [email protected]
•
•
•
Screen
•
Screen gain
•
Total Integrated Scatter
•
Objective (second) screen
Room Geometry and Environment
•
Projection and camera capture angles
•
Viewing distance / Ambient light
•
Ratio of image area to average speckle size
2/19/14
23
To learn more…
LIPA
Speckle Metrology Working Group
Update report at:
Technology Summit on Cinema at NAB
April 5-6, 2014
Las Vegas Convention Center
https://www.smpte.org/tsc2014
Contact LIPA at [email protected]
2/19/14
24
Laser Color Primary Selection
Options and Tradeoffs
Impacts on Gamut, Image Quality and Efficiency
Bill Beck
BTM Consulting, LLC
[email protected]
+1 617.290.3861
Primary Selection: Lumens vs. Watts
545 nm, 669 lm/W
532 nm, 603 lm/W
462 nm, 45 lm/W
445 nm, 20 lm/W
618 nm, 277 lm/W
640 nm, 120 lm/W
System A - “Native DCI” (P3)
nm
lm/W
lm/Color
Req’d W
277
17,880
65
618
669
64,529
96
545
45
3,289
74
462
366
85,697
235
Bill Beck
BTM Consulting, LLC
System B - “Available Lasers”
nm
lm/W
lm/Color
Req’d W
120
18,391
154
640
603
65,985
109
532
20
1,321
65
445
261
85,697
328
February 19, 2014
26
First Pass Observations…
• “Infinite” number of RGB combinations and “Spectral Power
Distributions” (SPD) to achieve desired gamut, white-point and
primaries - requires design TRADEOFFS
• Desired color-space can be produced with native RGB
wavelengths and balance delivered from the laser engine…
• …or via color correction in the projector, which always reduces
overall brightness and sometimes bit depth
• Likely ideal solution will be a bit of both
Contact LIPA at [email protected]
2/19/14
27
Single line vs. Multi/Wide-band Primaries
Narrow band RGB laser “lines” FWHM ≤ 1 nm
•
•
•
Simple modeling and supply chain … but
Massive Speckle
Potential for “Observer Metameric Failure” (OMF)
Multiple RGB lines per primary - n x FWHM ≤ 1 nm
•
•
•
Wavelength options depend on physics and availability
Little impact on speckle if narrowband
Unknown impact on OMF
Spectrally broadened RGB bands FWHM 10 - 40 nm
•
•
•
Contact LIPA at [email protected]
Replicates incoherent white light
Low speckle and OMF
Hard to achieve with available lasers
2/19/14
28
Single line vs. Wide-band Primaries
• Wide, “filled in” primary bands are ideal but…
• Very difficult to procure laser sources
•
At the right wavelengths
•
Fill in the bands of interest
•
Exhibit the same good beam quality, i.e., low étendue
•
Have similar lifetimes
• …all, at a reasonable cost
Let’s look at the tradeoffs
Contact LIPA at [email protected]
2/19/14
29
Primary Selection vs. Gamut
Rec 709
DCI P3
Rec 2020
•
•
•
•
Narrowband primaries “on locus”
Wider gamut and more saturated
But higher speckle and OMF
Longer Reds and shorter Blues are
commercially available
• Shorter Green adds Magenta but
cuts Yellow saturation
• Wider gamut primaries reduce
luminous efficacy (lm/watt)
Contact LIPA at [email protected]
2/19/14
30
Primary selection vs. Speckle Contrast Ratio (SCR)
•
•
Benchmark is Xenon illumination – Incoherent and Lambertian
•
RGB pass bands for DCinema installed base ~60 nm wide
•
System f# ~2.4 (fast) to maximize angle and usable lamp output
•
SCR for Xenon ~ 1% - hard to measure
Single wavelength, narrow line (≤1 nm) RGB primaries SCR ~20%
•
UNWATCHABLE in Green and Red; speckle noticeable even in Blue
•
Multiple emitters of the same wavelength – little reduction in SCR
•
Multiple beamlines that “fill” 10 - 40 nm reduce speckle to Xenon levels
***Each Laser Primary should fill 10 – 40 nm band***
Contact LIPA at [email protected]
2/19/14
31
Primary Selection vs. Observer Metameric Failure (OMF)
•
Three factors to consider:
•
•
•
•
Bandwidth is first order – wider is better for OMF and Speckle
•
•
•
Smooth SPD is better than peaky
Wide band primaries reduces saturation and gamut slightly
Wavelength is important, especially for narrow band primaries
•
•
•
Spectral Bandwidth of each primary
Spectral Power Distribution (SPD) i.e., flat vs. peaky
Color point of primary (wavelength or x,y)
Intersection with the tri-stimulus curves determines impact
More work is needed here – computational and observational
See: Wiley Periodicals Vol. 34, Number 5, October 2009 Rajeev Ramananth
Contact LIPA at [email protected]
2/19/14
32
Primary Selection vs. Luminous Efficacy
• Luminous Efficacy = White balanced lumens / RGB watt
• Ideal is to use “native” laser primaries:
•
Rec 709 :
613/550/463 nm = 362 lm/W
•
DCI P3 :
618/545/462 nm = 366 lm/W
•
Rec 2020 :
630/532/467 nm = 288 lm/W
• Readily available lasers:
640/532/445 nm
•
Rec 709 :
Raw 249 lm/W Correction reduces lm/W
•
DCI P3 :
Raw 261 lm/W Correction reduces lm/W
•
Rec 2020 :
Raw 261 lm/W Very slight reduction in lm/W
Contact LIPA at [email protected]
2/19/14
33
Primary Selection vs. Wall Plug Efficiency (WPE)
Projector + Engine WPE is a very complex function of:
• RGB wavelengths – sets luminous efficacy (200 - 350 lm/RGB watt)
• Étendue at the PJ input – determines PJ throughput
• Aggregation and delivery efficiency – set gross RGB watts required
• Laser Device WPE – drives engine efficiency and cooling required
•
Ranges from 3% for some Greens to >30% short Blue
• Laser Source Speckle Contrast Ratio – if low, no additional losses in
projector for downstream speckle reduction
Contact LIPA at [email protected]
2/19/14
34
Current Laser Primary Options
Color
Wavelength
(nm – FWHM)
Device
Type
Watts per
Device
Lumens
Per watt
Lumens per
Device
étendue
650 - 1
Diode
~1
73
73
med
638 - 1
Diode; Bar
≤8
131
1,048
high
615 - 8
DPSS + OPO
10
301
3010
low
550 – 0.1
VCSEL SHG
2
679
1358
med
546 - 12
DPSS wide
spectrum
20-40
671
>20K
low
532 – 0.1
DPSS; VCSEL;
FL SHG
2-100
603
>60K
range
525 - 2
Diode
1
542
542
med
462 - 2
Diode
1
50
50
med
445 - 2
Diode
3
20
60
med
For reference ~ 85,000 RGB lm input to the projector for 30,000 lm output
VCSEL=Vertical Cavity Surface Emitting Laser SHG=Second Harmonic Generation DPSS=Diode Pumped Solid State FL=Fiber Laser
Contact LIPA at [email protected]
2/19/14
35
A few words on Optical Fiber Delivery
•
Watts / beamline and beam quality determine the number and
size of fibers required
•
Best case: high power per color - with some redundancy
•
•
•
Worst case: lots of low power devices with bad beam quality
•
•
•
Fewest fibers per kilo-lumen on screen
Smallest diameter (cheapest) fibers
Requires large number of large diameter fibers
Cable ends up too big, too stiff and too expensive
Don’t worry about the fibers
•
•
Contact LIPA at [email protected]
Single fiber cables can deliver kilowatts of laser power
Attenuation is very low - up to 100 meters or more
2/19/14
36
Summary and Conclusions
•
Primary wavelengths + BW impact:
•
•
Gamut, Speckle, Observer Metameric Failure (OMF), Luminous Efficacy (LE),
Wall Plug Efficiency (WPE)
Wide band primaries, where possible, reduce speckle and OMF
•
•
Difficult to achieve in practice
Slight tradeoff with saturation and gamut (smaller triangle)
•
Wide Gamut laser options are available, but less efficient than DCI P3
•
Optimum primary wavelengths and bandwidths do no coincide with
mature, low cost laser offerings, especially for Green and Red
•
•
•
RED: too long and narrow; high speckle and low lm/W
GREEN: is too narrow; high speckle and low electrical efficiency
BLUE: can fill the band at low cost but power per device is still low
Contact LIPA at [email protected]
2/19/14
37
Do you see what I see?
Color Matching and the Single Observer
Matt Cowan
Entertainment Technology Canada Ltd.
[email protected]
Metamerism
Metamerism is the matching of apparent colour of objects with different spectral
power distributions. Colors that match this way are called metamers. (wikipedia)
Observer metameric failure can occur because of differences in colour
vision between observers. …….. In all cases, the proportion of long-wavelengthsensitive cones to medium-wavelength-sensitive cones in the retina, the profile of
light sensitivity in each type of cone, and the amount of yellowing in the lens and
macular pigment of the eye, differs from one person to the next. This alters the
relative importance of different wavelengths in a spectral power distribution to
each observer's colour perception. As a result, two spectrally dissimilar lights or
surfaces may produce a colour match for one observer but fail to match when
viewed by a second observer.
(Wikipedia)
Contact LIPA at [email protected]
2/19/14
39
Raises 2 Issues
1.
With color science we should be able to calculate different
spectral distributions that give an exact “average” color
match. (Metamers)
2.
The population of observers will have differing sensitivity to
the degree of the average match. (Observer Metameric
Failure)
Contact LIPA at [email protected]
2/19/14
40
What we see, What we measure
(100 years of color science in 1 slide)
Metrics established through:

Deriving observer’s sensitivity to color through Cone Sensitivity Functions

Choosing a representative observer as the “standard observer”

Transforming cone functions to “color matching functions” (CMF)

Determining spectral power distribution (SPD) of stimulus

Integrating the SPD across the CMF to achieve 3 numbers (X,Y,Z) to
describe the stimulus color

Normalize the X,Y,Z values to achieve the familiar x,y,L coordinates
XYZ
SPD
Contact LIPA at [email protected]
CMF
x,y,L
2/19/14
41
Color Matching Functions
Cone functions are basic HVS characteristic
CMF is linear transform of cone functions
CIE 1931 Color matching functions
Contact LIPA at [email protected]
2/19/14
42
The Real World – we are all different
Standard – singular response
Figure 3: Cone spectral responses for 1000 simulated individual
observers randomly sampled from the Tl, Tm, L, M, and S values
of Equation 1 (Fairchild et al 2013). (Plot is 1000 narrow lines on same plot)
Contact LIPA at [email protected]
2/19/14
43
Standard Observer – did we get it right
in 1931?
Contact LIPA at [email protected]
2/19/14
44
Try a Different CMF – fix offset
Offset is failure
of 1931 CMF.
Scatter is
observer
metamerism
From Sony white paper “Color Matching between OLED and CRT” v1.0 Feb 15, 2013
Contact LIPA at [email protected]
2/19/14
45
Observer Metamerism failure
How significant is differences in observers?
Occurs with all illuminations – even daylight
Contact LIPA at [email protected]
2/19/14
46
Figure 7: The metameric pairs for each of the 24 XRite Color Checker patches as seen by the standard
observer on the left and the 95th percentile simulated observer on the right. (Fairchild et al 2013)
Contact LIPA at [email protected]
2/19/14
47
Conclusions
Color matching using instruments will be better if we use CMF’s
updated from 1931
Observer Metamerism failure is a fact of nature, we live with it
every day
Contact LIPA at [email protected]
2/19/14
48
www.LIPAinfo.org
LIPA Laserama
Questions??
Pete Ludé
Bill Beck
Matt Cowan
Mission Rock Digital, LLC
[email protected]
BTM Consulting, LLC
[email protected]
Entertainment Technology Canada Ltd.
[email protected]
+1 617.290.3861

similar documents