"Apodizing" dacs: Cambridge Audio Azur DacMagic vs Meridian 808.2

[sharpi31]

Two recent reviews in Stereophile, two very different price points, similar d/a technology(?):

http://www.stereophile.com/digitalpr...er/index2.html

http://www.stereophile.com/cdplayers...er/index5.html

Based on the measured results, how much difference in sound quality would you expect to hear?

(Obviously this is entirely hypothetical and listening to both sources in a controlled setting is the only way to make a worthwhile comparison between these two products)

[tritosine]

1. Learn how to generate FIR coefficients
2. Take an opensource 'Sample Rate Converter', and mess with it
3. Compile it with your own minimum phase coefficients
4. Use it with your favourite media player app.

wow!

Btw I saw that new one does appetizing by itself? How is it called? SoX ? http://sox.sourceforge.net/

beware, using such filters in series is counterproductive because of phase shift, but with standard DAC (linear phase) its just what you want.

[Werner]

Originally Posted by sharpi31

Two recent reviews in Stereophile, two very different price points, similar d/a technology(?):

The technology is only superficially similar.

The Meridian is 'apodising' (a lame term probably coined because
at some time someone had need for a term describing his new filter...)
in the sense that it starts rolloff before Fs/2, and attains full attenuation at Fs/2. The Cambridge is not, it is only 12dB down at Fs/2.

While both are minimum phase and thus lack pre-ringing in their reconstruction filter impulse, only the Meridian will suppress the pre-ringing embedded in the signal by the commonly-used half-band linear-phase filtered ADC. So the system output of the Meridian will be without pre-ringing, and with a lot of post-ringing. The system output of the Cambridge will still have near-equal amounts of pre- and post-ringing.

Not that the audibility of 20kHz+ filter ringing has been proven ...

[idlehands]

There is probably a signifcant difference between the products, much of which will be caused by superior PSUs and output stage(s) that a massive price tag can fund, rather than by filter design. Any low power amp/output stage can be massively improved with less than $10 worth of parts.

Its a shame that the cheap Chinese DAC board makers don't do models with either Wolfson chips or good PSUs on the opamps.