Another D/A converter chip had the pleasure of spending some time with the Audio Precision. This time it was the ADAU1966, on the official evaluation board from Analog Devices.

FFT 1 kHz

Noise figure

THD vs. level

Frequency response

Linearity

5 dB of deviation is not state of the art, but it is way below the noise floor of the D/A converter so we might be measuring noise here.

IMD vs. level

And here we go again. IMD is noise limited all the way up to -2 dBFS but then shoots up.

FFT of IMD, -2 dBFS

FFT of IMD, -1 dBFS

FFT of IMD, 0 dBFS

IMD waveform at 0 dBFS

It seems this converter behaves quite similar to the AK4393, albeit with a slightly higher noise floor. Again the IMD behaviour is affected by the sigma/delta modulator capping out at near full modulation.

Listening test

The board only spent one evening in my home but it did a proper warming up whilst the measurements were taken.

I listened to both 48k and 96k sources using the same source material. In other words I used an external PC to do the first upsampling stage, and it seems the 96k was much better than 48k, indicating the first upsampling stage is limiting the converter. Of course I have much more horsepower when doing it externally, so that could explain the differences in quality.

• Needs 10 minutes run-in
• Sound stage is better than AKM and ESS, but behind Wolfson
• Rhythmically just below the AKM, above ESS
• Wolfson seems to play slower? <- some psycho-acoustic effect going on here.
• Wolfson seems louder although measured exactly the same output level. Could have to do with linear phase / minimum phase interpolation.

Overall a very good D/A converter, taking second place after the wolfson and before the AKM/ESS.