ClarkVision.com

The First Digital Camera: 1971

by Roger N. Clark

All images, text and data on this site are copyrighted.
They may not be used except by written permission from Roger N. Clark.
All rights reserved.

Introduction

When was the first digital camera made? 1971! If you read sites like wikipedia: Digital Camera, or this article on The First Digital Camera we would think it was in 1975. On sites like the above, we learn that Fairchild Semiconductor built the first commercial CCD in 1973, and then in 1975, Steve Sasson at Kodak built a 0.01 megapixel resolution camera, and it took 23 seconds to write the digital data to a magnetic tape.

The 1971 Digital Camera

But prior to 1975, Thomas B. McCord of MIT and James A Westphal of CalTech had already build a digital camera and published their results in the journal Applied Optics:

McCord, Thomas B. and Westphal, James A. (1972) Two-Dimensional Silicon Vidicon Astronomical Photometer. Applied Optics, 11 (3). pp. 522-526. ISSN 0003-6935. http://authors.library.caltech.edu/33192/

Their camera used a silicon vidicon 256 x 256 pixel array (0.065 megapixel) and wrote 8-bit image data to 9-track magnetic tape in about 4 seconds. They published isophote images of Jupiter and the globular star cluster 47 Tucanae obtained at the Cerro Tololo Interamerican Observatory, Chile, in 1971. Their paper was submitted to Applied Optics on October 12, 1971 and the paper was published in March 1972. This, I believe, is the first digital camera.

The McCord and Westphal digital camera weighed 10 kg and had approximate dimensions of 20 X 20 X 40 cm. The electronics and tape recorder were mounted in a 53-cm electronics rack and the camera was attached by a cable.

U.S. Patent 3951552

McCord and Westphal filed a patent on August 7, 1972 on their digital camera and the patent was issued on April 20, 1976. The patent includes the statement: "After a predetermined vidicon exposure interval the image electrically stored in the target is read out by a scanning electron beam, digitized and recorded to provide an immediately available digital record of the object."

Differences in Technology: Vidicon, CCD, and CMOS

The sensor in the tube in the McCord and Westphal camera is a silicon pixel array just like that in a CCD or CMOS sensor we have today. The main difference in the 3 systems (1: vidicon, 2: CCD, 3: CMOS) is the readout mechanism. All three use silicon to absorb photons to liberate electrons to be detected. All three make an analog signal that gets amplified and digitized. The main difference is how to get the signal off the light sensitive silicon array.

The first generation readout was the electron tube. An electron beam was used to probe the silicon array. It worked well, but the noise was high. Then came along the charge-coupled readout mechanism in the CCD. Early CCDs had poor charge transfer leading to lines in the image, but generally had lower noise than an electron beam readout. CMOS puts a transistor amplifier in each pixel and solves some of the problems with moving small charges around in the CCD readout mechanism. But in all three, the basic light sensitive silicon pixel array is the same. It has been refined over the years of course, and now CMOS sensors are surpassing CCDs in low noise readout.

The main advancements in digital camera sensors in the decade from 2001 to 2010 is 1) improved fill factor, 2) improved microlenses, 3) lower noise readout, and 4) lower fixed pattern noise. The quantum efficiency of the silicon pixel array has stayed about the same, around 40%, about half that of the McCord and Westphal camera. The CCD and CMOS readout mechanisms include support electronics around each pixel that lowers quantum efficiency, and the electron beam readout in the vidicon did not have that problem. By reversing the sensor and electronics, called thinned back side illuminated, the quantum efficiency of CCDs and CMOS can be made with quantum efficiencies above 90%, but as of this writing there are no back-side-illuminates CCDs or CMOS large sensors in consumer digital cameras, as production cost is high.

So not only was the McCord and Westphall digital camera ahead of its time in quantum efficiency, it had over 6 times the pixels, and wrote images faster than the Kodak 100x100 pixel camera made with a CCD 4 years after the McCord and Westphall digital camera.

The Second Generation Digital Camera, 1972

The digital camera was first presented publicly at the end of August 1971 at a conference in Santa Cruz, California (See Westphal and McCord, 1972). McCord's lab at MIT went on to produce improved digital cameras. J. Kunin, in his MIT 1972 masters thesis, describes the second generation McCord digital camera. It used an RCA 4532A vidicon tube, a 10-bit analog-to-digital converter, of which the top 8-bits were used, and data were recorded to a 9-track magnetic tape drive. The system wrote 256x256 pixel images to tape in 3.3 seconds and images were displayed on a 9-inch monitor.

Both the McCord and Westphal and the Kunin described systems had a peak quantum efficiency in the visible spectral range of over 80%, higher than today's consumer digital cameras by about a factor of 2. Both systems were also used for astronomical applications, so they cooled the vidicon tube to dry-ice temperature, which reduced dark current allowing exposure of many minutes. The data system described in Kunin's thesis allowed exposure times from 0.01 second to 99.9 minutes.

The system described by Kunin included improvements made over the original McCord and Westphal camera. Those improvements included the monitor to display images a few seconds after the camera acquired an image, and improved cold housing of the vidicon tube. Kunin's thesis showed black and white images from the system, including dark frames and images of the Moon.

Kunin also describes development of computer digital image processing tools to handle their new data. Tools included addition, subtraction, and ratioing of digital images.

The Third Generation Digital Camera, 1975

I started graduate school at MIT in 1975 to work on my Ph.D. Professor Thomas McCord was my thesis advisor and I began work on a spectrometer in McCord's lab. Parallel to development of the spectrometer was development of the 3rd generation digital camera. In April, 1976, McCord, engineers, other students and I took the new spectrometer and digital camera to the University of Hawaii, 88-inch telescope on Mauna Kea, Hawaii.

Conclusions

So not only did McCord and Westphal make the first digital camera, with operations in 1971, 4 years before the Kodak system, it was upgraded to the second generation and put to significant scientific use before the Kodak camera referenced so often on the web was even built. Further, the MIT group was developing computer digital image processing routines by 1972 to analyse the images from their digital cameras.

References

Kunin, J. S., 1972, A Technique for Two-Dimensional Photoelectronic Astronomical Imaging, With an Application to Lunar Spectral Reflectivity Studies Masters Thesis, MIT, September, 1972. http://dspace.mit.edu/bitstream/handle/1721.1/59619/29883941.pdf

McCord, Thomas B. and Westphal, James A., 1972, Two-Dimensional Silicon Vidicon Astronomical Photometer. Applied Optics, 11 (3). pp. 522-526. ISSN 0003-6935. http://authors.library.caltech.edu/33192/

McCord, Thomas B. and Westphal, James A., 1972/1976, Photometer-digitizer system, U.S. Patent 3951552, filed August 7, 1972, issued April 20, 1976. http://www.google.com/patents/US3951552

McCord, T.B., and J. Bosel, 1973, "Silicon Vidicon Astronomy at MIT." Astronomical observations with television-type sensors (J.W. Glaspey and G-A.H. Walker, eds.), U. of B.C. Vancouver, 137-161.

Thomas B. McCord, Michael J. Frankston, 1975, Silicon diode array vidicons at the telescope: observational experience Applied Optics, Vol. 14, Issue 6, pp. 1437-1446 http://dx.doi.org/10.1364/AO.14.001437 http://www.opticsinfobase.org/ao/abstract.cfm?uri=ao-14-6-1437

McCord, T.B., P. F. Kinnucan, and G. Fawcett, Jr., 1975, "Image Process- ing Systems Developed at MIT to Handle Silicon Vidicon Images and Spectra of Astronomical Objects." Image Processing Techniques in Astronomy, (de Jager/Nieuwenhuijzen, eds.) 329-334.

McCord, T.B., J. P. Bosel, and M. J. Frankston, 1975, "Performance of the MIT Silicon Vidicon Imaging System at the Telescope." Image Processing Techniques in Astronomy, (de Jager/Nieuwenhuijzen, eds.) 91-96.

McCord, T.B., and J. P. Bosel, 1975, "Potential Usefulness of CCD Imagers in Astronomy." Proceedings of the Symposium on Charge-Coupled Device Technology , Jet Propulsion Laboratory, March, 1975 .

Schaller, Phillip Henry, Jr, 1973, Applications of Two-Dimensional Vidicon Photometry: Venus, Masters Thesis, MIT, June 1973. http://dspace.mit.edu/bitstream/handle/1721.1/59620/29906577.pdf

Westphal, J. A. & McCord, T. B., 1972, A Silicon Vidicon Photometer, Publications of the Astronomical Society of the Pacific, Vol. 84, No. 497, p.133 (Paper presented at the Symposium "Advanced Electronic Systems for Astronomy - 1971," Santa Cruz, August 31 - September 2, 1971. http://adsabs.harvard.edu/full/1972PASP...84..133W

Home Galleries Articles Reviews Best Gear Science New About Contact

http://www.clarkvision.com/articles/first.digital.camera

First Published December 19, 2012
Last updated March 11, 2013