Quote:
In a November 2008 review of the Cary CD 306 SACD Professional JA added a footnote in the meansurements section which reads "Footnote 1: The sample of the original CD 306 I examined truncated external 24-bit data to 16 bits. By contrast, the Professional Version performed in exemplary fashion with external data.

Quote:
Cary keeps insisting the original Cary 306 CD/SACD player will not truncate 24-bit data to 16-bit that contains input signals of: 96KHz 24bit or 192KHz 24bit.

All I can do is report on my own experience.

Quote:
Is it possible your measurement tests did not provide one of those two specific 24-bit inputs?

My tests were unambiguous: 24-bit data fed to the player's digital input was truncated to 16-bit resolution. Without having your player to check, there's not much else I can say, I am afraid.

John Atkinson
Editor, Stereophile

Quote:
My tests were unambiguous: 24-bit data fed to the player's digital input was truncated to 16-bit resolution. Without having your player to check, there's not much else I can say, I am afraid.

I looked at my measurements of the original sample of this player at the weekend.

shows waveform of the original sample of the Cary CD-306's output while it played back a tone at -90dBFS from SACD. It is a good sinewave, indicating that the DSD encoding has better-than-CD resolution, as should 24-bit encoding.

shows the waveform of the Cary's output while it played back an undithered tone at -90.31dBFS from CD, which has 16-bit resolution. The waveform correctly shows the 3 DC voltage levels that describe this signal.

shows the waveform of an undithered 24-bit version of the tone at -90.31dBFS fed to the first sample of the Cary's external data input. It looks more like the CD waveform than the SACD waveform, and you can see that each of the voltage steps is the same size as it is in the CD waveform. In fact, it comprises just the 16 most significant bits of the original 24-bit data. This truncation wasn't affected by upsampling.

Finally,

shows the spectra of dithered tones at -90dBFS from SACD, 16-bit CD, and for external 24-bit data. The jagged curve is for 24-bit data and shows that the 1kHz tone is accompanied by high levels of 3rd, 5th, and 9th harmonics, which are the result of the truncation shown in the graph above.

These tests were all done 4 years ago. While it is always possible that my test-lab PC, which was generating the 24-bit data I was using for the tests, had a one-off hiccup at that time and was putting out 16-bit files rather than 24, it has never done that with any of the other D/A tests I have performed in the past 8 years.

Did you check your serial number against the ones I measured?

John Atkinson
Editor, Stereophile