"Arny Krueger" wrote in message
...
"Robert Morein" wrote in message

"Arny Krueger" wrote in message
...
"ScottW" wrote in message
news:AmUvb.4608$ML6.2599@fed1read01
"Erik Squires" wrote in
message
ervers.com...



Since the sample rate has been stated to be 44.1 KHz, the reconstruction
filter needs to cut off slightly below 22.05 KHz, not 15 KHz. In the case
of
a 15 KHz signal, this gives a 7.05 KHz range in which the filter needs to
cut off.

Unfortunately, I misread the question.
Your mathematical analysis is basically correct.
However, your statement that the reconstruction filter problem has been
solved is not correct, and probably accounts for the small variations in
perceived quality among DACs.

A digital reconstruction filter cannot be perfect unless it has infinite run
length and is noncausal. While digital filters are typically noncausal, they
have limited run length.

Likewise for the analog filter. However, the run length of an analog filter
is even more severly restricted.

The reconstruction filter remains the weak link in DAC design.