DCS dCS 904 Benutzerhandbuch
dCS 904 User Manual
Manual for Software Version 1.5x and 1.36
dCS Ltd
June 2000
Manual part no: DOC135904 iss 2B2
Page 32
135904ma2b2.pdf file available from website
Contact
dCS
on + 44 1799 531 999
email to: more@dcsltd.co.uk
(inside the UK replace + 44 with 0)
web site: www.dcsltd.co.uk
DSD
DSD Overview
-
DSD is a single bit very high sample rate (2.822 MS/s) format, where the single
bit words are heavily noise shaped to push noise energy above audio. The
frequency response is very high (well above 100 kHz) although at these ultrsonic
frequencies noise is also present.
bit words are heavily noise shaped to push noise energy above audio. The
frequency response is very high (well above 100 kHz) although at these ultrsonic
frequencies noise is also present.
The dCS 904 offers a number of different DSD modulators – as Filter options. All
the modulators in have the same signal frequency response. They differ in the
way they shape the out of band Q noise, and in how far they suppress the in
band Q noise. Filters 1 to 5 suppress Q noise at least 120 dB below the nominal
0dB DSD signal, which is one of the marketing specs for SACD.
the modulators in have the same signal frequency response. They differ in the
way they shape the out of band Q noise, and in how far they suppress the in
band Q noise. Filters 1 to 5 suppress Q noise at least 120 dB below the nominal
0dB DSD signal, which is one of the marketing specs for SACD.
Filter
Comments
SQNR
(20 kHz,
dB)
Stability
Description
1
High SQNR, high stability
126.14
1.7*10
10
Two complex
zeros
2
High SQNR
127.23
8.2*10
8
Two complex
zeros
3
High SQNR, very high
stability
stability
124.66
1.5*10
11
Two complex
zeros
4
Extremely high stability
122.07
3.0*10
12
Two complex
zeros
5
Reduced 100k noise
122.27
7.1*10
10
Two complex
zeros
6
Single complex zero
110.78
3.7*10
11
Single complex
zero
7
Real zeros
101.5
2.0*10
10
Real zeros only
Table 3 – DSD Filter Summary
Signal to Q Noise and SACD Specs
-
Although 120dB SNR over the 0-20 kHz band is a good target, it does not match
the ears response that well. The F weighted curve is currently accepted as a
good model for the ear, and we can use this to weight the noise produced by the
various filter choices. Such a weighting is shown in Figure 14
the ears response that well. The F weighted curve is currently accepted as a
good model for the ear, and we can use this to weight the noise produced by the
various filter choices. Such a weighting is shown in Figure 14
on page 35. DSD
gives very good performance using such a weighting (better than 23 bit pcm)
The figure shows that filter 6 and 7 give more F weighted Q noise suppression
in the audio band than filters 1 to 5. Filter 6 gives around 20 dB more
suppression than filters 1 to 5 under all circumstances and filter 7 gives 20 dB
more suppression below 10 kHz.
in the audio band than filters 1 to 5. Filter 6 gives around 20 dB more
suppression than filters 1 to 5 under all circumstances and filter 7 gives 20 dB
more suppression below 10 kHz.
SACD
specifies ultrasonic noise in two bands. These are specs informative
specs E2 and E3, and filter performance for a number of specs including these
are given below. The measurements given have been made using a Fourier
Transform based method.
are given below. The measurements given have been made using a Fourier
Transform based method.
5
Super Audio CD System Specifications, Part 2, Audio Specifications, available from Philips System
Standards and Licensing, Licensing Support, Building SFF-8, PO Box 80002, 5600 JB Eindhoven, The
Netherlands
Netherlands