Sony CDP-X559ES Review (CD Player)

[NTTY]

Great reviews.

Is there a reason for leaving the odd CD player, such as the Yamaha CD-2000, out of the comparative performance tables?

Thanks, and no reason except I did not copy/paste from the latest review I added it (with the Pioneer PD-D9).

 

[Scytales]

No side bands in that case

Thanks! So, there is indeed something to measure. Is it relevant or discriminatory? Only time and more samples will tell.

 

[NTTY]

Well, I like the test because it also provides a good view of inter channel crosstalk.

The test I use forces the second channel to stay up and running by sending a small level dithered signal into it, and that’s why I saw the same result with this test, meaning it is a good crosstalk test too.

About the IMD crosstalk, it obviously provides some data too, and I would have never thought about it. I’ll check with other CD players to confirm.

 

[SigmaR1]

What a good surprise to see a CDP X559ES in the hands of NTTY! Thank you Florent for this fantastic review!

It seems that the only improvement upon the previous generation of players (CDP X333ES/X555ES) is the efficiency of the digital filter presenting a slightly better attenuation, correct?

Otherwise, both players have similar performance, excepted the attenuation of the PS leakage... absent in the X333ES, but visible in the X559 despite the 2 transformers

Note that the analogic stage of the X333ES uses uncoupling capacitors whereas the X559ES provides a a dedicated servo circuit to remove the DC component of the signal. Obvioulsly, it doesn't make so much differences in SNR between the 2 players.

 

[SigmaR1]

Wow, the circuit board looks like something I designed in my student days: single sided, no ground plane, lots of through hole components and more than a few carbon composition resistors. It's amazing that it achieves this kind of performance.

And it is important to remember how quickly CD player technology progressed within about a decade.

Is it really that poorly designed/built?

 

[AnalogSteph]

Kinda par for the course for a consumer product of the time. (The X779ES got a slightly fancier board with green solder mask.) We tend to forget how much less flashy circuit boards used to be 35-40 years ago. Yes, single-sided PCBs are not ideal for digital circuits, and mid-'80s Philips players actually sported double-sided affairs already, but they were not without their issues either. It took a while for via problems to be worked out for one. You can see that this board was designed by a company with RF experience, the "make the traces as big as you can as long as clearance is adequate" approach is typical. (I imagine you won't get many issues with lifted pads.) After all any board starts out as solid copper, then the space between the traces gets etched away, so you might as well have them take up as much as possible and it won't cost you anything.

The PCB may be the least impressive component in this case. Otherwise, the wiring is neatly twisted, I also appreciate the plastic cover for the "High Voltage Area", and they got this copper-clad chassis and two transformers. At the end of the day, the proof of the pudding is in the eating, and this unit clearly delivers on the measurement front. Can't be too bad of a design then, can it?

 

[SigmaR1]

Kinda par for the course for a consumer product of the time. (The X779ES got a slightly fancier board with green solder mask.) We tend to forget how much less flashy circuit boards used to be 35-40 years ago. Yes, single-sided PCBs are not ideal for digital circuits, and mid-'80s Philips players actually sported double-sided affairs already, but they were not without their issues either. It took a while for via problems to be worked out for one. You can see that this board was designed by a company with RF experience, the "make the traces as big as you can as long as clearance is adequate" approach is typical. (I imagine you won't get many issues with lifted pads.) After all any board starts out as solid copper, then the space between the traces gets etched away, so you might as well have them take up as much as possible and it won't cost you anything.

The PCB may be the least impressive component in this case. Otherwise, the wiring is neatly twisted, I also appreciate the plastic cover for the "High Voltage Area", and they got this copper-clad chassis and two transformers. At the end of the day, the proof of the pudding is in the eating, and this unit clearly delivers on the measurement front. Can't be too bad of a design then, can it?

Thank you for this detailed answer

 

[capslock]

Kinda par for the course for a consumer product of the time. (The X779ES got a slightly fancier board with green solder mask.) We tend to forget how much less flashy circuit boards used to be 35-40 years ago. Yes, single-sided PCBs are not ideal for digital circuits, and mid-'80s Philips players actually sported double-sided affairs already, but they were not without their issues either. It took a while for via problems to be worked out for one. You can see that this board was designed by a company with RF experience, the "make the traces as big as you can as long as clearance is adequate" approach is typical. (I imagine you won't get many issues with lifted pads.) After all any board starts out as solid copper, then the space between the traces gets etched away, so you might as well have them take up as much as possible and it won't cost you anything.

The PCB may be the least impressive component in this case. Otherwise, the wiring is neatly twisted, I also appreciate the plastic cover for the "High Voltage Area", and they got this copper-clad chassis and two transformers. At the end of the day, the proof of the pudding is in the eating, and this unit clearly delivers on the measurement front. Can't be too bad of a design then, can it?

There are also examples of terrible layout from the 90s. Philips CD931 and Teac CDP-3000 come to mind. I owned both back in the day, and while I reverse engineering the schematic, I got he impression that a student intern had done the layout. Single sides boards, ground and power lines strewn all over the board with little attention to local supply decoupling or where noisy ground currents flow. Digital lines with no attempt to curb the worst ringing, and plenty of HC logic, which was particularly bad in this respect.

 

[AnalogSteph]

If I remember Bruno Putzeys right who started out at Philips in the late '90s, it does seem to have been common practice to task junior engineers with designing the lowly consumer goods (mini stereos and whatnot). So possibly that's what happened with the CD 931, which after all was a midrange model released years after CD players were a halo product for the company and while build quality was on a steep decline (didn't the 900 series have all-plastic faceplates? plus the notorious stripped gears in the CDM-9 mechanism). Production for this generation had even been moved to Malaysia. Not long after, Philips-branded hi-fi was history altogether.

The PCB inside the CD 951 looks more confidence-inspiring, with plenty of via stitching in sight... mind you, for twice the asking price it better be.

 

[srkbear]

Would love to see how my Sony UBP-x1100ES Blu-ray/SACD player measures up these days. This was the last “ES” SACD edition they released, in 2019, and was inexplicably discontinued in 2021–their current TOTL offering is their lower price point UBP-x800M2, which is not marketed as part of their alleged “ES” or “elevated standard” tier of audiovisual products. From my research, they apparently used an AKM4452 192khz/32bit DAC chip, and the unit is capable of decoding up to DSD 11.2MHz DSF, DVD Audio, SACD, Blu-ray audio, and as expected with a Sony player, just about every imaginable audio codec via USB or a DNLA server.

It also includes their “DSEE Extreme” audio upscaling feature, which I was finally able to sort out by feeding the separate HDMI audio signal to my Topping d90se, via an HDMI to I2S audio de-embedder—it basically upsamples the signal one fold (44.1khz—>88.2khz, DSD 2.8MHz to 5.6MHz, etc). Not entirely clear what purpose that serves, at least for PCM. I’m currently using it with the de-embedder to decode my Blu-ray audio and SACD discs to my Topping, and can flip back and forth between it and my Roon-ready audio streamer by switching between the USB and I2S inputs.

It certainly sounds great to me, but considering that Tidal and Qobuz often offer 96 or 192khz/24 bit versions of the majority of the albums I own on SACD, I’m fairly convinced that these versions (which have often been remastered more recently) offer higher resolution and dynamic range than the SACD discs, and avoid the potential timing issues with the I2S input—I trust the clock in my Topping considerably more than the Sony’s, despite the fact that I’m aware that at 55 I’m unlikely to hear any differences either way

 

[SigmaR1]

Which [CDP X707ES], btw, should have more headroom for overs than the unit reviewed here [CDP X559ES].

I just noticed this remarks, could you explain why? Thanks

 

[AnalogSteph]

I must have gotten this confused with the X779ES (measurements), actually. (Though one would hope that there would be no regression in the newer model.)

Hmm, that makes me wonder - why doesn't the X559ES also have some headroom for overs then? The digital filter is the same CXD2560M, and I don't think they would have been so stingy to only give it +/-5 V analog supplies like so many lesser Sony players have... nope, +/-15 V for the entire output stage. Weird. Some detailed comparison of schematics may be in order.

 

[SigmaR1]

I must have gotten this confused with the X779ES (measurements), actually. (Though one would hope that there would be no regression in the newer model.)

Hmm, that makes me wonder - why doesn't the X559ES also have some headroom for overs then? The digital filter is the same CXD2560M, and I don't think they would have been so stingy to only give it +/-5 V analog supplies like so many lesser Sony players have... nope, +/-15 V for the entire output stage. Weird. Some detailed comparison of schematics may be in order.

Thanks you AnalogSteph, however I can't see where the measurements of the X779 show a better protection for overs. Moreover, both x559 and x779 were from the same generation.

I have to check previuos results provided by Scytales showing the (lack of) headroom for overs od the SCD 555ES, which was launched a few years after the x559/x779 players, with even more advanced IC deidcated to digital filtering.

 

[AnalogSteph]

Thanks you AnalogSteph, however I can't see where the measurements of the X779 show a better protection for overs.

From previously linked article:

Using a full-scale (0dBFS) 1kHz squarewave to drive the Sony, we got a result typical of players using linear-phase digital filters, but with no clipping of the ringing on the top and bottom of the squarewave noticeable (fig.9).

Fig.9 Sony CDP-X779ES, 1kHz squarewave at 0dBFS.

As the text implies, performance like this was not necessarily a given. For example, here's what the JVC XL-Z1050TN did:

Fig.3 JVC XL-Z1050TN, waveform of 1kHz squarewave at 0dBFS (2ms time window).

Similar results came from an NAD 5425, Denon DCD-2560, Sony CDP-X555ES and a California Audio Labs Tercet Mk.III.

An Audio Research DAC1 even produced hard clipping throughout:

Fig.6 Audio Research DAC1, 1kHz squarewave at 0dBFS.

Likewise a Luxman D-105U. I would expect similar from lots of lower-end Sony players with +/-5 V analog rails.

Mind you, performing flawlessly in this test seemingly was not entirely exotic... a Kinergetics Research KCD-40 (CXD1144B + 4x AD1860N-K), Philips LHH1000, PS Audio Ultralink DAC and various others did fine.

Such differences in overflow handling should be reflected in 0 dBFS+ handling somewhere... right? A full-scale square wave technically has a level of +3 dBFS.

 

[Scytales]

Just to complete this discussion, I just took some screenshots of the 1002.27 Hz square @ 0 dBFS from NTTY test CD at the output of a Sony SCD-555ES. None of the 5 digital filters of the Sony shows any sign of clipping despite minor peak output voltage variations. Here is the standard filter (left and right channels, which overlap perfectly):

Yet, this player suffers from inter-sample over distortion according to the relevant tests from the same disc.

 

[SigmaR1]

Thanks to both of you for the answers! This suggests that there is no obvious correlation between the overshoot observed on a square signal and the intersample overs, right?

 

[WilliamfromParis]

Thank's for the job Florent. I put the Sony in the hands of my tech to check the optical output. I love that Sony ES series.

 

[NTTY]

Thanks to both of you for the answers! This suggests that there is no obvious correlation between the overshoot observed on a square signal and the intersample overs, right?

There should be. I’ll check if I measured the Sony clipping or if it was my ADC. Sometimes I forget to lower the gain at the input of the ADC for that test…

 

[Billy Budapest]

Would love to see how my Sony UBP-x1100ES Blu-ray/SACD player measures up these days. This was the last “ES” SACD edition they released, in 2019, and was inexplicably discontinued in 2021–their current TOTL offering is their lower price point UBP-x800M2, which is not marketed as part of their alleged “ES” or “elevated standard” tier of audiovisual products. From my research, they apparently used an AKM4452 192khz/32bit DAC chip, and the unit is capable of decoding up to DSD 11.2MHz DSF, DVD Audio, SACD, Blu-ray audio, and as expected with a Sony player, just about every imaginable audio codec via USB or a DNLA server.

It also includes their “DSEE Extreme” audio upscaling feature, which I was finally able to sort out by feeding the separate HDMI audio signal to my Topping d90se, via an HDMI to I2S audio de-embedder—it basically upsamples the signal one fold (44.1khz—>88.2khz, DSD 2.8MHz to 5.6MHz, etc). Not entirely clear what purpose that serves, at least for PCM. I’m currently using it with the de-embedder to decode my Blu-ray audio and SACD discs to my Topping, and can flip back and forth between it and my Roon-ready audio streamer by switching between the USB and I2S inputs.

It certainly sounds great to me, but considering that Tidal and Qobuz often offer 96 or 192khz/24 bit versions of the majority of the albums I own on SACD, I’m fairly convinced that these versions (which have often been remastered more recently) offer higher resolution and dynamic range than the SACD discs, and avoid the potential timing issues with the I2S input—I trust the clock in my Topping considerably more than the Sony’s, despite the fact that I’m aware that at 55 I’m unlikely to hear any differences either way

My x1100es sounds fine too. I’m curious about the measurements. Ever notice how tiny the analog audio board is in this player (including the D/A conversion section)? Not sure if that is good or bad. Also, I am a bit flummoxed as to how much attention the Sony gives to vibration/resonance control—ie, the “frame and beam” construction of the chassis and the extra “girder” mounted to the transport that is not present in the x800m2.

 

[Descartes]

Thank you for the review, I still have a Sony SCD-XA777ES and use it every day!