"chung" wrote in message
...
Steven Sullivan wrote:
the usual stuff for the first page or so, but then Nika Aldrich chimes in
and it gets interesting.

http://gearslutz.com/board/showthrea...6&page=4&pp=30

Therer's also a link to *another* student's CD vs DVD-A vs SACD study

http://gearslutz.com/board/attachmen...achmentid=6491





Thanks for the link; it is very interesting.

I especially find it illuminating when someone brought out some marketing
promo that touted the impulse response of DSD and how that looked "better"
than CD and other PCM systems (post #282). As Nika said, that is doctored
data and completely bogus. He gave a good explanation of it he

http://gearslutz.com/board/showthrea...&page=10&pp=30

See posts #289 and #290. In particular, this is germane:

"The (DSS) system can be said to have EITHER 120dB of dynamic range OR it
can be said to have a high frequency response and great impulse response,
but not both. If we only look at the region up to 20kHz, filtering off all
of the HF noise above 20kHz then the impulse response of DSD will look
just like the impulse response of 44.1kS/s in your plot above, and the
dynamic range is about the same. If, however, we want to give DSD credit
for having a very high frequency response and therefore a very steep
transient response then we also have to factor in all of that excess noise
that accumulates up in that range.

"The graphic you provided is an attempt to say that DSD is the best of all
worlds - it has huge dynamic range and great transient response (frequency
response). You know this not to be the case. If the graph truly showed the
unfiltered transient response the graph would have a lot of noise in it as
I stated above. This graph attempts to say, "look, if we filtered out all
of that noise (can't we just forget about the noise?), but managed somehow
not to filter out the signal, then it would have a transient response like
this." Unfortunately this is completely bogus."

This should be required reading for those who use impulse response of DSD
as the technical reason why it sounds "better". BTW, vinyl has a terrible
impulse response. The "Analog" impulse response on some of those diagrams
is *NOT* the response from vinyl or tape, but rather some theoretical,
computer generated, model.


Your prejudice is showing.

By no stretch of the imagination is this the same graph. Just examine them
carefully....no analog, different order. It is only the writers opinion
(not stated) that this is the same graph doctored Photoshop. All one can
say for sure is that both are a picture of an impulse response. I have seen
similar charts reproduced elsewhere...for example in the student paper
recently cited here that are yet different measurements but show much the
same thing.

Moreover, to the graph "with noise".....which is worse (aurally)..... a near
perfect impulse response with some low level noise, or an impulse response
that doesn't appear in nature, and which falls short of accurately capturing
the dynamics of the wavefront. Noise is natural...pre-echo and truncated
dynamics are not.

Finally, to say that if you filter out the noise you get the same impulse
response as 44.1/16 or 48/24 is totally misleading...of course a 20k cutoff
is a 20k cutoff. Fortunately for us, SACD machines do reproduce frequencies
above 20k. CD machines do not. Simple as that. So the better transient
response *is* a factor in SACD reproduction. And IMO it is audible and
probably along with a lower noise floor in the critical upper bass and
midrange is the main reason people's reaction tends to be "it sounds natural
and unstrained".