quote:

Originally posted by las3rjock Sorry jusware, but that's not correct either. Specifically, it is not true that "Any value in analog can be converted to digital as long as the bit register is big enough." An obvious counterexample is 1/3 -- this value cannot be represented exactly in any finite number of decimal or binary digits. Any real digital system (i.e. one that can be built and that actually works) uses some finite number of bits to represent values over some finite dynamic range. If the system has a dynamic range of R and uses N bits to represent each sample, then it has an accuracy of +/- ( R / 2^N ). For example, suppose a turntable has a pitch range of 16% and uses a 8-bit digital controller. Then the turntable is accurate to ( 16 / 2^8 ) = 1/16 % = 0.0625 % pitch. This may or may not be good enough for the application -- if it's not, then one can gain accuracy by using more bits, but at the cost of controller speed and more expensive components. I imagine that turntable control probably doesn't push the envelope for digital control, but I imagine there are applications (scientific measurement, for example) where the state-of-the-art in digital cannot match the accuracy and performance of the state-of-the-art in analog.

Woahhhhhhhhhhhhhhhhhhhhhhh

This guy knows his stuff.

quote:

Originally posted by las3rjock Sorry jusware, but that's not correct either. Specifically, it is not true that "Any value in analog can be converted to digital as long as the bit register is big enough." An obvious counterexample is 1/3 -- this value cannot be represented exactly in any finite number of decimal or binary digits. Any real digital system (i.e. one that can be built and that actually works) uses some finite number of bits to represent values over some finite dynamic range. If the system has a dynamic range of R and uses N bits to represent each sample, then it has an accuracy of +/- ( R / 2^N ). For example, suppose a turntable has a pitch range of 16% and uses a 8-bit digital controller. Then the turntable is accurate to ( 16 / 2^8 ) = 1/16 % = 0.0625 % pitch. This may or may not be good enough for the application -- if it's not, then one can gain accuracy by using more bits, but at the cost of controller speed and more expensive components. I imagine that turntable control probably doesn't push the envelope for digital control, but I imagine there are applications (scientific measurement, for example) where the state-of-the-art in digital cannot match the accuracy and performance of the state-of-the-art in analog.

Impressive bro...
Always said vinyl pwns you cd junkies

___________________
|||CHRISTOPHER THOMAS|||

quote:

Originally posted by beats and beeps Woahhhhhhhhhhhhhhhhhhhhhhh This guy knows his stuff.

Well, I'd always hoped a MSc in electrical engineering would be useful someday... Anyway, the lesson for the day is that, in general, the state-of-the-art in digital cannot be more accurate than the state-of-the-art in analog. Your digital signal has to come from somewhere, and with the exception of computer-generated data, that somewhere is usually an analog signal. However, it may happen that digital is good enough for a specific application, and has other advantages (durability, portability, etc.) that make it preferable to analog.

___________________
The original las3rjock! Accept no substitutes!

CURRENT TUNES

• Nickelson - Yin (Solid Globe Remix)
  
• Deep Dish - Say Hello
  
• Syntax - Bliss (Wrecked Angle Mix)
  
• The Chemical Brothers - Hold Tight London
  

I'd like to add that one major advantage of digital is it's tolerance to noise. In many digital circuits a "0" is reprented by any voltage less than 0.7 volts for example, and a "1" is represented by any voltage greater than 1.4 volts. So if there's interference that causes a signal of 0.4 volts to be changed to 0.5 volts or 0.2 volts or whatever there's no problem since it's still considered a "0" (same for "1"s)

An analogy would be carrying a bucket of marbles where each marble represents one unit versus carying a bucket of water where you have to measure the amount of water to get the value..."noise" would be something like rain (which would increase he value) or spillage (which would decrease the value) would easily affect the value, whereas marbles are easily counted (and it wouldn't rain marbles and would be harder to spill marbles). However, assuming there was no noise you could have an infinite number of values in the bucket of water where with the bucket of marbles you could only have as many values as the number marbles that could fit in the bucket. This is why analog is better for turntables.

Digital is great for comminications (less resistance to noise) and making copies of data (since it will be an exact copy) but useless in a turntable. It's much better to have the infinite range since there would be very little noise in the short distance from the pitch fader to the circuit that controls the motor.

If it's digital then say you needed to change the pitch of a record from 130.01 BPM to match another record at 130.05 BPM, but since it's digital it jumps to 130.09 BPM (those are random numbers). With an analog (assuming you have a delicate enough touch to get the fader in the perfect spot) you can get nearly exactly 130.05 BPM.

THat was a little longer than I thought it was going to be...sorry...hopefully it made sense.

___________________
-tom