In Reply to: Digital equalizer cause phase errors? posted by tantra on April 14, 2007 at 12:24:52:
Tantra:Analog EQ's typically use minimum phase filters. If the digital EQ is using IIR (infinite impulse response) filters, the transfer function (amplitude and phase) will mimic the analog equivalent of the IIR filter. FIR (finite impulse response) filters are linear-phase.
If you're doing equalisation via convolution, and doing it properly, phase response should be improved as well. (Actually, convolution boasts improving amplitude response and phase response resulting in improved impulse response.)
The cool thing about correcting phase errors digitally is the ability to do what is called "forward-reverse" processing. This enables one to correct phase response by itself while leaving amplitude response unchanged. In the analog domain, (in real time now) if you muck with one you will impact the other. This is because frequency response is a function of both amplitude AND phase response. The only penalty is a time delay (latency) between input and output. Who cares? Once the song starts (1/2 second later) you soon forget all about that short short little wait! ;)
In a unit of this price range, where I think you're going to suffer is not in the processing power side (that's cheap nowadays) but in the digital inputs and outputs. If this were a "real" pro device, it would be slave-able or have a master clock with wordclock out or both. All the digital I/O is asynchronous. Don't get me wrong, ASRC's like the CS8420 in a Phase Locked Loop sound pretty darned good, especially when you feed them a low jitter signal.
I just think it would be nice to be able to go synchronous or USB input direct to I2S.
Someday!! :o)
Cheers,
Presto
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: Digital equalizer cause phase errors? - Presto 16:36:07 04/16/07 (1)
- Re: Digital equalizer cause phase errors? - tantra 07:50:19 04/17/07 (0)