EasyManua.ls Logo

DVC DigitEyes Series - Do I Need a Digital Output Camera to Benefit from the Higher Signal-To-Noise Ratio of DVC Cameras; What Is the Best Way to Benchmark the DVC Camera Vs. the Competition in a Side-By-Side Comparison; What Is Meant by Equivalent Number of Bits (Enobs)

Default Icon
66 pages
Print Icon
To Next Page IconTo Next Page
To Next Page IconTo Next Page
To Previous Page IconTo Previous Page
To Previous Page IconTo Previous Page
Loading...
39
shutters (with moving parts). Since they typically use CCD "staring" arrays, an electro-
mechanical shutter is required between integration periods ! Apart from the somewhat annoying
"ticking" sound, there is also a decreased system reliability by introducing a component with
moving parts in an otherwise reliable, solid-state design.
8.3.5 Do I need a digital output camera to benefit from the higher signal-to-noise
ratio of DVC cameras?
No; the higher signal-to-noise ratio is found on the DVC-0A analog RS-170 only camera. If you
hook it up to a monitor, you'll observe the reduced noise or "snow" on the screen. Compare this
against any other video camera that has approximately the same sensitivity.
8.3.6 What is the best way to "benchmark" the DVC camera vs. the “competition"
in a side-by-side comparison?
To do a true side-by-side test of the DVC camera vs. the "competition", point both cameras at
the same scene, with both lenses at the same f-stop number . This ensures that the same
amount of light is incident on both CCDs.
To do an accurate comparison, make sure that the AGC circuits in both cameras are disabled.
Use a monitor with an A/B switch, so that monitor settings do not corrupt the experiment !
Usually, you'll find that the DVC camera has a brighter picture (more sensitivity) for the same
illumination. Note that this is done without compromising the signal-to-noise ratio of the DVC
camera. In order to achieve the same sensitivity in the competitor's cameras, the gain would
have to be increased significantly - this would lead to a decreased signal-to-noise ratio (more
"snow" in the image).
Now, turn the scene illumination down (use Neutral Density or ND filters, if you can; keep in mind
that ND filters are calibrated on a logarithmic scale - so a ND 2.0 transmits 1/100th of the
incident light, a ND 3.0 transmits 1/1000th of the incident light).
Turn up the video gain on both cameras till you can see an equally bright picture from both
cameras with the same amount of ND filters and the same f-stop of the lens.
This is where you will see the DVC camera outperform the competition - compare the amount of
"snow" in both images under low-light, high-gain conditions.
8.3.7 What is meant by Equivalent Number of Bits (ENOBs) ?
In the digital video domain, the signal-to-noise ratio, takes on a more serious role !
When the analog video signal and its accompanying noise component are digitized, the noise
tends to corrupt the lower significant bits. In an 8 bit system, the noise may show up as a pixel
value with an uncertainty. This means that for a uniformly illuminated "flat" field, digitized pixel
values might vary from, say, 80 (or Hex 50) to 83 (or Hex 53). With a noisy signal, the same
range might apply for the same pixel location, sampled in successive fields.
Note: the lower two significant bits have been lost to noise in this example ! Obviously, if the
analog signal has more noise in it, the result will be that more bits will be lost to noise - lower
ENOBs !

Table of Contents