Monday, 9 June 2014

Sony NWZ-ZX1 vs iPhone 5S Measurements

As an addendum to part 2 of my Sony NWZ-ZX1 review, I'll quote here Soomal's RMAA for the ZX1 and iPhone 5S side by side so they are easier to compare.

Please note that these are only the figures for 16bit/44.1khz playback. The NWZ-ZX1 can playback higher bitrate and DSD files, and Soomal measured those as well, so check the original links below for details.

NWZ-ZX1 iPhone 5S
Noise level, dB (A): -92.5 -96.0
Dynamic range, dB (A): 91.7 96.1
THD,%: 0.0042 0.0016
Intermodulation distortion,%: 0.010 0.0050
Stereo Separation, dB: -93.2 -89.0

All measurements quoted from Soomal's reviews:

Please keep in mind that RMAA results are not comparable between different test beds. ie: you cannot compare Soomal's results against other people's results unless they are using the same testing equipment chain. However, it is possible to compare Soomal's results for one piece of equipment against another. 

A good article about the limitations of RMAA is Nwavguy's guide: 

However, the limitations that Nwavguy explains are largely related to figuring out absolute values for various performance characteristics. For the purpose of relative comparison, as in this case where we are trying to compare the ZX1 and 5s, the results are helpful. 

Here are my remarks on the measurements. Although they both appear to be high performance players, the 5S measures better in almost every respect except stereo separation. Stereo separation is of course important because you do not want one channel bleeding into the other and in this respect the ZX1 has something like a 4db edge.

However, most troubling for the ZX1 is its measured dynamic range of 91.7db. This is the result of the high noise floor on the ZX1. The noise on the ZX1 is not academic - it is something you can hear as soon as you plug an earphone or headphone in. The iPhone 5S has a noticeably blacker background. This is easy to determine through the use of an A/B switch.

The high noise floor on the ZX1 means that it is not capable of achieving the maximum dynamic range allowed by 16 bit audio.

The 16-bit compact disc has a theoretical dynamic range of about 96 dB... The perceived dynamic range of 16-bit audio can be as high as 120 dB with noise-shaped dither, taking advantage of the frequency response of the human ear. Digital audio with undithered 20-bit digitization is also theoretically capable of 120 dB dynamic range. Similarly, 24-bit digital audio calculates to 144 dB dynamic range.
The ZX1 is billed as a 'high resolution' audio player and a suitable device for playing back 24 bit audio, and yet the high noise floor makes this somewhat pointless. The ZX1 is being sold on the premise that it can do something that it can't.

Something which isn't measured here is output impedance. The iPhone 5S has an output impedance of something like 2.3ohms. The ZX1's output impedance has proven to be difficult to measure, but Tom from CYMBACAVUM writes:
Class D designs inherently have very low output impedance because they usually run feedback in a full-bridge design across the LPF. Essentially, OI is effectively "zero" with properly implemented Class D designs, except for out-of-band frequencies, such as >20 kHz, where it'll increase significantly, because that's where the switching voltage fluctuates a lot. Full-bridge Class D designs will dampen essentially any load it can detect effectively. Power/voltage are entirely different questions, and that goes into the switch efficiency of its caps.
Note that both the iPhone 5S and the ZX1 are Class D designs, though it appears that the 5S may have extra resistance added in it's output path.