Camera Architecture and You

Download Report

Transcript Camera Architecture and You

Camera Architecture and
Microscopy
Austin Blanco
How does a camera work?
► Light
is converted to electrical charge.
► Charge is stored in potential wells, these are
the “Pixels” on the camera.
► Energy collected in each pixel is digitized.
► Digital data is transferred to computer.
Photons Emitted from
Sample Fluorphore or
Transmitted Light
Pixels in CCD Array
Pixel Architecture
Active Area
Masked Area
Register
Output Node
A/D Converter
Register
Output Node
A/D Converter
• Most microscope manufacturer cameras use this array!
Zeiss Axiocam, Olympus DP Series, Nikon Qi/DS, Hamamatsu “Orca” Series,
Photometrics “CoolSNAP” Series, Q-Imaging “Retiga” Series.
Back Thinning
Barrier
Potential
Well
P
E
-Very Sensitive (<90% Quantum Efficiency)
-Expensive Process (Chemicals etch silicon)
-Can only be used on FT or Full Frame sensors (not Interlines)
N-Channel
Active CCD
Area
Active CCD
Area
Active Area
Masked Area
EM Register
Output Node
A/D Converter
EM Register – “Impact Ionization”
Color Methods
► Pixels
are not able to discern colors.
► This limitation lends to three methods of
“colorization” in order to build an image.
3 Shot Color
CCD
RGB Merge
3 CCD Color
Light
RGB Merge
Color Mosaic
-Enables fast acquisition
-Sacrifices intensity spatial resolution
-Lowers Sensitivity
-Only used on interline sensors
Pixel Size
Larger pixels can
hold more energy
– Providing more
dynamic range
Smaller pixels hold less
energy.
-Lower Spatial
Resolution
-Greater Dynamic
Range
-Faster
-More Sensitive
-Lower Spatial
Resolution
-Greater Dynamic
Range
-Faster
-More Sensitive
Resolution
Choosing the Best Camera
Sensitivity
Back Thinned
Frame Transfer
EM CCD
Interline
Monochrome
Color Mosaic
Speed
Resolution
CMOS vs. CCD
► What
is a “CMOS”
 Complimentary Metal Oxide Semiconductor
►Designates
an electrical conductor layout which is
used for many applications:




Sensors
Memory
Transceivers
Data Converters
CMOS Architecture
Active Pixels
A/D Converters
CMOS Readout
Column Select
Row Select
Positive Aspects
►
Both Sensor and electronics for control are manufactured
on chip
 Theoretically lower cost
 Smaller complete package (no external boards required)
 On chip processing is more powerful (color compositing)
►
A/D at each pixel
 Superior control of gain for color applications
 Far greater speed potential than CCD’s at lower readout rates
(lower A/D Noise per pixel)
 Sub Arrays have much more power due to non-serial nature of
architecture
►
Rugged Design
 Less soldier contacts means less potential for something to break
Negative Aspects
►
Shuttering Methods
 Nowhere for stored charge to go!
 Commonly use “Rolling Shutter” to avoid problem
►
Distorts moving objects
 Full frame shuttering
►
►
►
►
QE
requires multiple transistors per pixel
Solves rolling shutter drawbacks
Reduces fill factor
 Under 10um pixel size QE is lower for CMOS than CCD
►
►
►
More on-active area parts reduces fill factor
> 10um CCD & CMOS are equivalent in QE
Dynamic Range

multiple layers of construction in CMOS design
►
►
►
Offer photons the opportunity to bounce to neighboring pixels
Causes image softening
Reduces dynamic range (creates new noise source)
 A/D variations between pixels increases noise factor
Integration Limits
► CMOS
design changes
 Require more work for integration into camera
bodies
 Slow concept to market (~18 months)
► CCD
Design Changes
 “Drop In” new CCD’s can be used in old designs
 Faster turnaround means more profit now
(~ 8 months)
Future of CMOS
► Potential
for back thinning
 Could avoid current problems with softening
and lower QE
 With back thinning would compete and possibly
beat CCD’s for low light applications
► Specific
Manufacturing Facilities
 Originally CMOS were thought to be cheap in
production. Ultimately specific foundries will be
required for scientific grade chips.
►Cost
will go up to or above CCD manufacture.
Conclusions
► CCDs
and CMOS sensors fit our needs in
complimentary ways:






CCDs for low light quantitative apps
CMOS for high resolution brightfield apps
CMOS for rugged environments
CCDs for speed apps
CMOS for small form factor
CCDs for high dynamic range brightfield
► CMOS
sensors are a “disruptive technology”
 Someday CMOS will eclipse and possibly replace CCDs,
just like tape -> CD -> Mp3.
 Reaching the performance level of CCDs will require
CMOS prices to meet or exceed current CCD pricing,
meaning a stable market for imaging apps in the future
Thank You!
► Austin
Blanco
► Technical Instrument
 Training on Imaging / Software
 Custom Programming for Software
 Consultation on Existing and New Systems
► 510-708-2995
► [email protected]