Home
AudioAsylum Trader
Digital Drive: Re: Domain Dementia... by Todd Krieger

Upsamplers, DACs, jitter, shakes and analogue withdrawals, this is it.

For Sale Ads

FAQ / News / Events

 

Re: Domain Dementia...

208.176.171.144


[ Follow Ups ] Thread:  [ Display   All   Email ] [ Digital Drive ]
[ Alert Moderator ]

"Todd, the response is abs(Sin(x)/x). It doesn't go negative, you can't have a negative response. We're talking about amplitude here which is the modulus."

I guess so. Although I'd cringe at the the notion of how a DAC that had a frequency response curve like that would *sound* like... [-;

"What you've been thinking about up till now is the sum of fourier coefficients in the time domain summing up to create the impulse response."

I see what you're saying. The sum of the FIR coefficients convolved with a pulse signal in real time. Which indeed would yield the impulse response. (I've never heard these called "Fourier coefficients.") This response shows the function of the filter.

"That is indeed the function A(sin(n.x)/x), a scaled/stretched sinc function - here it does go negative since we are describing the time-domain signal here."

This is the *only* way in digital audio the "sinc" function is applied. It goes negative simply because an audio signal can go negative.

"But you don't get it for free - there is still a sinc rolloff in the Bode plot (amplitude/frequency)."

There is a rolloff, but a non-OS or ZOH DAC does not use a "sinc" filter, since it would require a digital (oversampling/upsampling) filter to implement it. The rolloff is caused by modulation in the HF (the samples do not always occur at the peaks of HF signals), causing the *average* amplitude to decrease, hence the measured rolloff. (The *peak* amplitude would be about the same.)

"That is why non-OS DACs that do not employ corrective analog post-filters are not flat across the band."

The post filter is needed *more* with non-OS because no prior (digital) filtering has been applied. Otherwise, the ultrasonic images of the signal would pass through the output at *full* intensity if no such filter was used. (Some have claimed that zero filtering causes havoc with downstream electronics due the excessive HF content.)

"Before I forget - I wanted to ask - have you gotten your trial dAck! yet?"

Still waiting for it. Thanks.


This post is made possible by the generous support of people like you and our sponsors:
  Atma-Sphere Music Systems, Inc.  



Topic - upsampling article - solidgore 08:08:38 08/22/03 ( 47)