[opendtv] Re: Precision

dan.grimes@xxxxxxxx wrote:


In video, how much precision is enough?


<snip>
First of all, bit depth: 8 bit, 10 bit, 12 bit, higher? What is the graduation difference the eye can see at a particular wavelength? Obviously, it is going to be different at different wavelengths. So what is our target maximum? Or should it be different for different wavelengths?
I think there is a fundamental misunderstanding of digitization in your question. When analog signals are properly digitized, the only difference between 8-bit, 10-bit, or 12-bit coding is signal-to-noise ratio, not "graduations." 10-bit systems have about 12 dB better SNR than 8-bit, and 12-bit systems have about 12 dB more than 10-bit.
To properly digitize an analog signal, there must be at least one-half of the least-significant bit of level uncertainty (call it noise or dither). If there is, then the necessary quantization error will be uncorrelated noise; if there isn't, the necessary quantization error will be correlated distortion.
Consider a one-bit PCM audio system. Without at least 1/2-LSB dither, it functions as a gate: Anything above the threshold gets through; anything below doesn't. It sounds unintelligible. With 1/2-LSB of dither, there will be a fairly loud, constant hiss (or some other noise, depending on the spectral shape of the dither), but the audio can be heard reasonably clearly through the noise. Add a bit, and the SNR increases by 6 dB.
So your question should be how much SNR we want, not how many gradations we need. 

TTFN,
Mark




----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org 
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: