2TheLighthouse
New Member
Hello! Inspired by @NTTY's excellent recent reviews of CD players, I decided to run my Sony Discman D-12 through its paces and see how one of these early units stacks up to the modern competition.
Tons of thanks to @NTTY for providing their Test CD and for all the work they've done documenting its use!
Background:
The Discman D-12 (also known as the D-22 in other markets) was introduced in 1987-88. It was one of the first Sony portables to be able to run off of standard batteries (4x AA). It is very similar to the Discman D-20, the only real difference is the inclusion of MEGABASS on this model
This particular unit was manufactured in Japan in February of 1991, making it older than me! It's in good shape other than an extremely finicky Line Out port. All measurements shown in this review were therefore done through the headphone out. This did allow me to measure the MEGABASS however, and the few tests I was able to do on the Line Out seemed near-identical to the headphone out. Measurements were done with the included power adapter as I didn't see much significant noise from it appearing in the measurements.
Measurements (Headphone Out into MOTU UltraLite mk5):
I used NTTY's review of the FiiO DM13 as a baseline for what measurements to include. I can take more though if anyone has any specific requests! Also, this was my first time conducting measurements like this or using REW, so if there are any errors you notice please forgive me and let me know so I can correct them
I'll start with the bandwidth measurement, including measurements of the MEGABASS at both the Mid and Max positions.
These lines from top to bottom are MEGABASS Off, MEGABASS Max, and MEGABASS Mid. These measurements were taken with the headphone volume set to about 6 which corresponds to a comfortable listening volume on my Focal Clears. I actually really like the filtering on offer here; the Mid setting provides a huge +12dB bass boost but stays nicely focused on the sub-bass without bleeding much into the mids. The 6 dB or so of treble boost is also welcome and helps to make the MEGABASS a "fun mode", especially for cheap headphones where extreme lows and highs are lacking. The Max setting for me starts to get too muddy with the boost being significant up to around 200 Hz. Interestingly, the MEGABASS curves flattened out a lot when I cranked the headphone volume to max, which I reckon is due to clipping as we'll see more about next...
Now I'll try a 999.91 Hz Sine wave at 0dbFS to test THD and noise:
Yikes! It was clear to me that with the headphone output set to max, some serious clipping was happening here. I did the same test but at -6dBFS and things shaped up a lot:
Much better! 0.11% THD and a SNR of 73.5 dB might not seem like much to celebrate, but for a player of this age I personally find it passable (and certainly an improvement over the 0dbFS test). This puts it at around 11 bits of resolution which while significantly short of the CD standard 16 bits still manages to beat vinyl and cassette (and even a lot of master tapes...).
A key issue here is obviously that maxing the headphone volume causes massive distortion on peaks near 0dBFS. I was curious if the headphone out was still being clipped significantly even with this -6dBFS test, but re-running the test at a lower headphone volume returned similar results. This makes me think that there is some inherent non-linearity happening here. Let me know if you have any insight on this!
Distortion aside, the noise level actually looks quite good to me! I was surprised by this especially given the age of the unit and the wall adapter powering it. It looks to me like there are some power hum components creeping in, but they remain below -100dbFS.
Next, I'll run a multitone test:
This looks alright to me, though noise seems to be pretty significant, peaking around -80dBFS.
Here's a few IMD tests (AES17 DSD and AES17 MD):
The DSD result of -37.1 dB seems quite bad, maybe slow attenuation on the DAC filter is causing some serious aliasing problems? The MD result is much better however at -61.1 dB.
Finally, I'll finish off with the Stereophile 3DC test which should look like a wave oscillating between 3 distinct sample levels:
Annnnnd nope! There is some subtle oscillation happening, but it is buried amongst a bunch of noise, not surprising given our earlier results.
Some final measurements are as follows:
Conclusion and Subjective Impressions:
All in all, this is (for its age) a pretty decent little player! Sure, pretty much anything you could buy new these days will run laps around it, but for a groovy little piece of history it's perfectly functional with a flat frequency response and low enough noise and distortion to not be a problem for me personally. I can't say I've heard any of the negatives presented on these graphs in actual listening, though when testing I could faintly hear those nasty IMD peaks in the AES17 DSD test, so maybe avoid music with lots of near-nyquist content
Subjectively I really dig this thing. It's built like a brick with a solid metal bottom, has simple and easy to use controls, and the sound and volume on offer suit my needs just fine. I get pretty solid battery life out of it too using modern NiMH rechargeables. My main beef with it is the difficult to read display. There is also of course no anti-skip on it, but whatever passive shock absorption they built into it works surprisingly well (not for walking around with it mind you, but well enough that it won't skip when you press the buttons or bump into your desk).
I hope you enjoyed this somewhat pointless review of an old piece of gear! Let me know if you have any questions or insights to share, or if there are any measurements I skipped that you would like to see. Also let me know if you'd be interested in more reviews of random old portable CD players, I have a late 90's Panasonic and a terrible early 2000's Durabrand (my boyhood player) ready to throw on the bench if there's interest!
Tons of thanks to @NTTY for providing their Test CD and for all the work they've done documenting its use!
Background:
The Discman D-12 (also known as the D-22 in other markets) was introduced in 1987-88. It was one of the first Sony portables to be able to run off of standard batteries (4x AA). It is very similar to the Discman D-20, the only real difference is the inclusion of MEGABASS on this model
This particular unit was manufactured in Japan in February of 1991, making it older than me! It's in good shape other than an extremely finicky Line Out port. All measurements shown in this review were therefore done through the headphone out. This did allow me to measure the MEGABASS however, and the few tests I was able to do on the Line Out seemed near-identical to the headphone out. Measurements were done with the included power adapter as I didn't see much significant noise from it appearing in the measurements.
Measurements (Headphone Out into MOTU UltraLite mk5):
I used NTTY's review of the FiiO DM13 as a baseline for what measurements to include. I can take more though if anyone has any specific requests! Also, this was my first time conducting measurements like this or using REW, so if there are any errors you notice please forgive me and let me know so I can correct them
I'll start with the bandwidth measurement, including measurements of the MEGABASS at both the Mid and Max positions.
These lines from top to bottom are MEGABASS Off, MEGABASS Max, and MEGABASS Mid. These measurements were taken with the headphone volume set to about 6 which corresponds to a comfortable listening volume on my Focal Clears. I actually really like the filtering on offer here; the Mid setting provides a huge +12dB bass boost but stays nicely focused on the sub-bass without bleeding much into the mids. The 6 dB or so of treble boost is also welcome and helps to make the MEGABASS a "fun mode", especially for cheap headphones where extreme lows and highs are lacking. The Max setting for me starts to get too muddy with the boost being significant up to around 200 Hz. Interestingly, the MEGABASS curves flattened out a lot when I cranked the headphone volume to max, which I reckon is due to clipping as we'll see more about next...
Now I'll try a 999.91 Hz Sine wave at 0dbFS to test THD and noise:
Yikes! It was clear to me that with the headphone output set to max, some serious clipping was happening here. I did the same test but at -6dBFS and things shaped up a lot:
Much better! 0.11% THD and a SNR of 73.5 dB might not seem like much to celebrate, but for a player of this age I personally find it passable (and certainly an improvement over the 0dbFS test). This puts it at around 11 bits of resolution which while significantly short of the CD standard 16 bits still manages to beat vinyl and cassette (and even a lot of master tapes...).
A key issue here is obviously that maxing the headphone volume causes massive distortion on peaks near 0dBFS. I was curious if the headphone out was still being clipped significantly even with this -6dBFS test, but re-running the test at a lower headphone volume returned similar results. This makes me think that there is some inherent non-linearity happening here. Let me know if you have any insight on this!
Distortion aside, the noise level actually looks quite good to me! I was surprised by this especially given the age of the unit and the wall adapter powering it. It looks to me like there are some power hum components creeping in, but they remain below -100dbFS.
Next, I'll run a multitone test:
This looks alright to me, though noise seems to be pretty significant, peaking around -80dBFS.
Here's a few IMD tests (AES17 DSD and AES17 MD):
The DSD result of -37.1 dB seems quite bad, maybe slow attenuation on the DAC filter is causing some serious aliasing problems? The MD result is much better however at -61.1 dB.
Finally, I'll finish off with the Stereophile 3DC test which should look like a wave oscillating between 3 distinct sample levels:
Annnnnd nope! There is some subtle oscillation happening, but it is buried amongst a bunch of noise, not surprising given our earlier results.
Some final measurements are as follows:
| Crosstalk | -44.0 dBFS (okay but not great, better than vinyl!) |
| Pitch Error | 0.019% (190 ppm, not great but not horrible for a 34-year-old clock chip) |
Conclusion and Subjective Impressions:
All in all, this is (for its age) a pretty decent little player! Sure, pretty much anything you could buy new these days will run laps around it, but for a groovy little piece of history it's perfectly functional with a flat frequency response and low enough noise and distortion to not be a problem for me personally. I can't say I've heard any of the negatives presented on these graphs in actual listening, though when testing I could faintly hear those nasty IMD peaks in the AES17 DSD test, so maybe avoid music with lots of near-nyquist content
Subjectively I really dig this thing. It's built like a brick with a solid metal bottom, has simple and easy to use controls, and the sound and volume on offer suit my needs just fine. I get pretty solid battery life out of it too using modern NiMH rechargeables. My main beef with it is the difficult to read display. There is also of course no anti-skip on it, but whatever passive shock absorption they built into it works surprisingly well (not for walking around with it mind you, but well enough that it won't skip when you press the buttons or bump into your desk).
I hope you enjoyed this somewhat pointless review of an old piece of gear! Let me know if you have any questions or insights to share, or if there are any measurements I skipped that you would like to see. Also let me know if you'd be interested in more reviews of random old portable CD players, I have a late 90's Panasonic and a terrible early 2000's Durabrand (my boyhood player) ready to throw on the bench if there's interest!
Last edited:
