Vincent CD-S3 CD player vs. DacMagicPlus DAC shootout (measurements)

[pma]

CD player vs. DAC shootout

I was curious to compare parameters of the nicely built CD player Vincent CD-S3 (2004) vs. a good but not the newest DAC, Cambridge Audio DacMagic Plus (2012). Cambridge DacMagic Plus was already tested here. Vincent CD-S3 is a player with HDCD decoding option and has both RCA single-ended and XLR balanced outputs. It is a nicely built, heavy unit, from times when there was not a dictate to miniaturize everything.

I connected CD-S3 and DM+ simultaneously to my analog audio interface unit, which I use for level matching of various components to the input of the measuring card. This interface unit has balanced inputs and outputs and gain selectable by DACT stepper attenuator with 2dB steps. It has output noise of -120dBV(A) and negligible distortion.

CD-S3 sent the digital data through TOSLINK optical cable to DM+. CD-S3 was playing various test discs and signals were simultaneously measured at CD-S3 and DM+ XLR outputs. Signals were analyzed by MI Multi Instrument software and plotted as dBR relative to base frequency amplitude, in most cases. Because CD was a signal source, all measurements apply for 44.1kHz/16bit input data.

THD at 1kHz and 0dBFS

DacMagic Plus THD 1kHz 0dBFS

Vincent CD-S3 THD 1kHz 0dBFS

THD measured at DM+ output was 0.002% and SINAD was 90.91dB (BW 20kHz linear). THD at CD-S3 output was 0.001% and SINAD was 90.54dB. CD-S3 has a bit lower distortion, a bit worse SINAD, but DM+ has cleaner spectrum.

This plot shows measurement of both units with 48kHz measuring bandwidth. Note noise shaping of CD-S3 and lower noise of DM+.

THD at 1kHz and -60dBFS

DM+ THD 1kHz -60dBFS

CD-S3 THD 1kHz -60dBFS

THD measured at DM+ output was 0.133% and SINAD was 34.04dB (BW 20kHz linear). THD at CD-S3 output was 0.165% and SINAD was 31.17dB. CD-S3 has a bit higher distortion and worse SINAD, and we can see some spurious frequencies in this test, that are audible if we get high volume setting.

CCIF IMD 19+20kHz

DM+ CCIF 19+20kHz

CD-S3 CCIF 19+20kHz

Though CD-S3 has lower H3 distortion, DM+ performs better with cleaner spectrum and lower H2 (1kHz difference tone).

Noise - digital silence

DM+ has cleaner spectrum and lower noise, CD-S3 has some spuriae in the spectrum. This was measured with Fs=96kHz to show noise shaping.

DIM test – square 3.18kHz + sine 15kHz

DM+, DIM square 3.18kHz + sine 15kHz test

CD-S3, DIM square 3.18kHz + sine 15kHz test

CD-S3 is better in this test, has less fake frequencies in audio band.

Multitone test

Both CD-S3 and DM+ are doing well. Measured with Fs=96kHz and BW=48kHz.

Conclusion

Both the CD player and the DAC have respectable parameters, with one exception, and this is spectral cleanliness for low level signals. DM+ is clearly better in this. However, CD-S3 was a good partner to the 8 years newer DM+. Built quality is much better for CD-S3. But it was more expensive, about $1300.

 

[pma]

Vincent CD-S3 inside. Really nice unit.

 

[restorer-john]

Are you sure the DAC magic plus isn't muting on digital zero? External D/A converters are renowned for doing that.

 

[pma]

Are you sure the DAC magic plus isn't muting on digital zero? External D/A converters are renowned for doing that.

Good question, at least output relay does not mute. Below are 3 noise measurements in high sensitivity setup. In all cases, CD-S3 is connected to DM+ and XLR output of DM+ is analyzed.
1) DM+ is OFF so output relay is in mute position and the system measures its own noise and interferences. You can see that noise measurement rms limit is -118dBV(A), means through A-weighted audio band.
2) DM+ is ON but CD-S3 is OFF. This measures DM+ output noise and you can see that it is -103dBV(A). Output relay thus not muted.
3) DM+ is ON and CD-S3 is ON and playing digital zero track. Noise rms is again -103dBV(A), Some small spuriae present.

Please also note that digital zero plot in #1 post of the thread is not calibrated, shown in dBFS of the measuring system only for the purpose of direct relative comparison between CD-S3 and DM+.

This shows noise floor of the DM+ output.

 

[pma]

Are you sure the DAC magic plus isn't muting on digital zero? External D/A converters are renowned for doing that.

One more, for DM+ ON and XLR OUT muted because of headphones inserted in 6.3 jack OUT. Noise -118dBV(A) - measuring system noise.

 

[restorer-john]

for DM+ ON and XLR OUT muted because of headphones inserted in 6.3 jack OUT.

Why does it mute the XLR when using H/Ps, is it using the balanced line driver for the headphone amp and switching it via the (DPDT) relay?

I think we can be sure that digital zero is not muting the DM+ (which is a good thing)

Cheers.

 

[pma]

Why does it mute the XLR when using H/Ps, is it using the balanced line driver for the headphone amp and switching it via the (DPDT) relay?

I think we can be sure that digital zero is not muting the DM+ (which is a good thing)

Cheers.

I think the main reason is that when you insert headphones the DAC goes to lower output volume not to make you deaf and it activates volume control knob.
For XLR link output, volume knob is activated only by a start-up procedure when you have to hold volume knob pushed during turning power ON. The DAC then has this setting in memory for next time, until you change it again.

 

[pma]

Another test is with CD-S3 playing 1kHz/-90dBFS sine through DM+.

1st plot is for the dithered sine. Please note stellar cleanliness of the spectrum. Please also note elevated noise of the RedBook 44.1/16 format, TPD dithered 16bit resolution.

The same for undithered signal. This shows importance of proper dithering for 16bit data. 24bit data does not need so much care.

Noise bottom floor is lower, but see the spuriae that are audible.

 

[pma]

Now I have taken another CD player into the shootout, Marantz SA7001, and was interesting in HF spectrum at the output, compared to DacMagicPlus.

First, let's see the spectrum of both units playing 1kHz/-60dBFS 16bit tone, spectrum ploted from 10kHz to 10MHz with linear frequency axis scale.

Marantz SA7001 HF noise - stunning result, very clean.

So it usually goes with DACs.

Further example is 1kHz/-60dBFS sine tone in time domain. It was amplified of +30dB at the output of both units, otherwise the outpu signal would be too small for the oscilloscope input. No low pass filter was used.

Marantz SA7001

DacMagicPlus - tons of HF noise. Who cares, it is "usual" with contemporary products.

 

[pma]

Some new measurements of Vincent CD-S3, measured by E1DA Cosmos ADC

In the level matched ABX test, this CD player is undistinguishable form either DacMagicPlus or Topping D10s.