• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Review and Measurements of Cambridge DacMagic Plus

folzag

Member
Patreon Donor
Joined
Jul 24, 2018
Messages
60
Likes
70
#81
I own two of these; one in black and one in silver. This test is proof to me that measurements only go so far. While the M51 test showed far more 'anomolies' than this [tho, with notably higher SINAD at least], I can strongly affirm that when source is a cable box to> 50 foot hi speed HDMI to> M51, it far out resolves the same cable box to> short TOSLINK to> DacMagic+ to> 50 foot Mogami cables to the same amp & headphones as previous. ;)
Could be the cable box. Seems probable cable box design would pay close engineering attention HDMI path, and TOSLINK could be an after-thought.
 

pma

Active Member
Joined
Feb 23, 2019
Messages
221
Likes
303
Location
Prague
#82
Here is Toslink comparison to USB:


Performance is tiny bit worse with Toslink. Noise floor is a bit higher, there are a couple of tones on each side of our main signal at 12 khz. And some power supply spikes down low:

View attachment 22437

Nothing audible so use whichever is needed/convenient.
I assume those power spectral lines, though very low in level, might be a result of different cable shield leak currents when you changed from USB to Toslink. I have got rid of them completely by using an isolation transformer in front of the DacMagic Plus power supply.
 

pma

Active Member
Joined
Feb 23, 2019
Messages
221
Likes
303
Location
Prague
#83
I think back in 2011, there was not as much awareness of output impedance. Or the serious headphone audiophile community that exists today. So people did what they used to do which is to put a resistor on the output and with it, gain output protection, ease of design, etc.
I do not agree with this. I was using output impedance of <= 5ohm in my preamp headphone outputs, class A biased, and able to give more than 250mA output current, back in 2002. Still not usual, today.
 

audio_tony

Active Member
Joined
Feb 24, 2019
Messages
120
Likes
101
Location
Leeds, UK
#84
EDIT: Added jitter graph.

Hi all, new member here - I have been following the forum for a few weeks now.

This review was of particular interest to me, as I have an older Dacmagic (Azur) the model that preceded the one under test here.

I have performed some measurements myself - however I don't have access to an AP (lucky you Amir!!) - I use Virtins and an Asus STX sound card.

The figures I get for my DacMagic:

THD: 0.0009% (L) 0.0007% (R)
THD+N: 0.0011% (L) 0.0010% (R)
SINAD: 98.82dB (L) 99.98dB (R)
SNR: 102.72dB (L) 102.76dB (R)

I have posted the graph from Virtins and also RMAA test results (should be attached to this post).

I sometimes question the numbers produced by RMAA.

My test procedure (Virtins) is: Input levels are -6dB - through to a home build sound card buffer (a simple box, 100k Input impedance, 2x gain) -> Asus STX.

Test frequency is: 1.000428 kHz

For RMAA, I use -1dB levels, rest of the signal path is the same.

All tests conducted at 48kHz / 24bit, using SPDIF into the DAC.

I suspect the noise levels could be better (Asus quote 124dB for the card) however I guess the noise from the PC is inevitable.

Jitter was measured with the Jitter function in Arta - 48kHz / FFT: 32768 / Kaiser7 / Linear averaging.

There was no difference between Optical and SPDIF. I am using a USB to SPDIF converter (CM6631A chipset) from a Chinese Vendor on Ebay (£32 or 32 GBP) - I guess that helps.
 

Attachments

Last edited:

Jimster480

Major Contributor
Patreon Donor
Joined
Jan 26, 2018
Messages
1,216
Likes
704
Location
Miami
#85
Well for something that is 10 years old.... it looks like it performs pretty well!
Better than many other devices which cost way more and are newer.
Sometimes "newer" isn't always better.
 
Joined
Nov 23, 2018
Messages
32
Likes
30
#86
is there actually a difference when using a balanced out but using an xlr to rca to feed a unbalanced amp ? Some people at other forum claims to hear improvements as the rca out has to go through 1 extra op amp
 

audio_tony

Active Member
Joined
Feb 24, 2019
Messages
120
Likes
101
Location
Leeds, UK
#87
I've read that people who own the Azur DAC claim that taking an unbalanced feed from the balanced output claim it sounds better, but I don't believe it myself.
I challenge anybody to hear the difference a single opamp (or even multiple opamps for that matter) inserted into a circuit would make, assuming the circuit design is competent of course.
After all, the music we listen to will have passed through several opamps (usually standard NE5532s) during the course of mixing / mastering.
An NE5532 configured for a gain of x2 will yield THD less than 0.0001% and s/n ratio of -110dB + so quite how someone would hear any difference...
 

JJB70

Major Contributor
Patreon Donor
Joined
Aug 17, 2018
Messages
1,793
Likes
3,656
Location
Milton Keynes, England
#89
Should i upgrade mine to a tier 1/2 Topping or SMSL DAC to enjoy better sound then?
Only if you want to buy something newer. In terms of sound quality the DAC Magic+ is audibly transparent and remains pretty well featured and I see no reason to replace one if it still works.
 

Sueekhy

New Member
Joined
Apr 27, 2019
Messages
4
Likes
0
#90
Only if you want to buy something newer. In terms of sound quality the DAC Magic+ is audibly transparent and remains pretty well featured and I see no reason to replace one if it still works.
Thanks JJ, I'm always tempted by the lure of DSD native though....
 

Veri

Major Contributor
Patreon Donor
Joined
Feb 6, 2018
Messages
2,963
Likes
2,899
#92
For anyone wondering how the jitter is so squeaky clean on this older device. They used advanced (ESS Sabre-like) ASRC to get it so. Found the details here (dutch article): https://www.hifi.nl/recensie/3628/Cambridge-Audio-DacMagic.html

Each channel of the DacMagic has its own Wolfson WM8740 DAC and analog circuits that are dual mono and balanced. The DacMagic therefore also has XLR outputs in addition to cinch/rca. Thanks to a unique digital filter circuit with DSP (Digital Signal Processing), extra Jitter suppression and upsampling to a higher sample rate became possible. The DSP IC is a Texas Instruments 32-bit type that works with ATF (Adaptive Time Filtering) asynchronous upsampling. That ATF is in fact one of the most important assets of the DacMagic because the resampling process has a higher precision than that of most common sample-rate converters. The Cambridge DacMagic ATF DSP works with an algorithm that is closer to the original signal than conventional interpolators and can upsample up to 192 kilo Hertz (24-bit), according to the manufacturer.

Thanks to the 32-bit DSP IC, Cambridge Audio has been able to develop their own digital filter algorithms with three different filter options. With a button on the front panel you can choose between linear phase, minimum phase and steep filtering.
 
Joined
Apr 23, 2019
Messages
75
Likes
23
#93
Has anyone tried to feed this DAC from a Sony DAP (I have a NW-ZX300) via the OTG cable?

I was going to buy a used one but a guy on Head-Fi said that this DAC would not pick up a flow from the OTG cable on different Sony DAP's. He has the two models above mine.

Help?

Thanks.

Shane D
 
Joined
Sep 26, 2019
Messages
61
Likes
57
#94
I just wanted to say thanks for posting these measurements of the DacMagic Plus. I was impressed enough by the 32-tone IMD measurement to pick up one of these for a very good value used yesterday, and it really is such a nice unit. Probably wouldn't have thought to put it on my list to consider if the measurements hadn't been available.
 
Joined
Sep 26, 2019
Messages
61
Likes
57
#95
For anyone wondering how the jitter is so squeaky clean on this older device. They used advanced (ESS Sabre-like) ASRC to get it so. Found the details here (dutch article): https://www.hifi.nl/recensie/3628/Cambridge-Audio-DacMagic.html
There's a bit more about the technical details behind the Anagram ASRC here: http://www.auditionveritable.com/840C/840C White paper.pdf

Basically they use a non-linear rather than linear interpolator to do the ASRC followed by a conventional Bessel digital filter. They don't disclose exactly what order of non-linear interpolator they use (although I'd imagine you could find out by digging through old Anagram patent filings). There's a graph on page 10 comparing the DNR to an unidentified ASRC IC, probably the AD1896. I believe the AD1896 by comparison uses a linear interpolator.

Edit: I was curious... for those who want the gory details, here's their method: https://patents.google.com/patent/US7259700B2/
 
Last edited:
Top Bottom