Technical description

OVERSAMPLING

TheaudiodataonCDsandDVDsisstoredatasampli					ng
rate of 44.1 or 48kHz i.e. for each second of m					usic
44.100or48.000sampledvaluesareavailablefore					ach
channel. In the SADV 1245 R the audio data read from
the DVD or CD is „multiplied“to a higher sampling					rate
(384kHz)beforeitisconvertedbackintoanalogue			music
signals.Thisprocessdeliversasignificantlybett			er,more
finelygraduatedsignalto theconverter, whichcan				then
beconvertedwithcorrespondinglyhigherprecision.					The
raised sampling rate is a calculating process for w					hich
there are many different mathematical methods. In
almost all digital audio devices which exploit the
advantagesofincreaseddigitalsamplingrateapro					cess
known as a FIR (Finite Impulse Response) filter is
employed for this purpose. At	 we have been
carryingoutresearchformorethantenyears,aime					dat
improving the oversampling process, because the
standardFIRmethodhasonedrawbacktosetagainst					its
indisputable advantages: it adds small pre and pos					t
echoestothemusicsignals.At	wehavedeveloped
mathematical processes (known as Bezier polynomial
interpolators)whichdonotsharethisdisadvantage				.For
this reason they should sound better and more natur					al
than the usual standard process. Since the calculat					ing
procedure employed by us is considerably more
complex than the standard method, the		SADV 1245 R
features a highperformance digital signal processo					r
(DSP) which carriesouttheoversamplingprocess w					ith
immense precision (56bit) using special algorithms
developedby .

ThefreelyprogrammableDSPwhichweuseiscapable
of carrying out the oversampling process using any
method of calculation. For this reason we have im
plemented two slightly modified Bezier processes
(OVS2) in the SADV 1245 R in addition to the pure
Bezierprocess(OVS3),togetherwithtwovariants	ofthe
standard process (standard OVS FIR filter and OVS1)		.
For more information on the different processes ple		ase
refer to the next section. You can switch between t		he
differentalgorithmsusingtheFILTERbutton,andd		ecide
foryourselfwhichofthefiltersyouprefer.

Standard OVS FIR filter

ThelongFIRfilteristhestandardoversamplingpr	ocess
in digital technology, offeringextremelylinear fr	equency
response, very high damping, linear phase charac
teristicsandconstantgroupdelays.Thedisadvanta	geis
thepreandpostechoeswhichareaddedtothesig	nal.
These „time range errors“ tend to affect the music
signal’sdynamics,precisionandnaturalness,andr	educe
spatialorientation.

Frequency response and transient characteristics of the long FIR filter

OVS 1(short FIRfilter)

Shorteningthefilter(lowercoefficient)reducesth etime rangeerrors,albeitcombinedwithaslightlossof linearity inthefrequencyrangeanddampingperformance.

Frequency response and transient characteristics of the short FIR filter

OVS 2( Bezier interpolatorplus IIRfilter)

InthisprocessanidealBezierinterpolatoriscomb	ined
with what is known as an IIR filter. This eliminate	s the
problematic preecho of the FIR method. This proces	s
produceshighly„analogue“systemcharacteristics,	witha
soundqualityandmeasuredperformancesimilartot	hose
ofgoodanaloguediscplayers.

Frequency response and transient characteristics of the Bezier interpolator plus IIR filter

OVS 3(pure	Bezierinterpolator)
This process delivers a perfect reconstruction of t			he
originalmusicsignal.Itexhibitsnopreorpost		echoesof
anykind,anddoesnotaddcolorationortimingerr			orsto
theoriginalsignal. Insonictermsthismethodoff			ersan
impressive blend of naturalness, good dynamics and
accuracy.Thisisourpreferredprocessduetoits			advan
tages in respect of sound, and is the basic (defaul			t)
settingofthe	SADV 1245 R.

Frequency response and transient characteristics of the Bezier interpolator

Note:

For optimum performance of the  oversampling algorithmspleaseselect CD-Upsample= OFFinthemenu 'PREFERENCES / SOUND'. Wesuggesttousethissettingatalltimes.