Marantz SA-10 Review (SACD Player & DAC)

[SuicideSquid]

Then why if a DAC doesn’t have a USB input a converter is required. Why do computers bother with a USB output if it’s compatible with a standard SPDIF signal, when a coaxial cable would do!

This seems to be getting into "intellectually dishonest and deliberately obtuse" territory.

 

[Suffolkhifinut]

The laser reads 1s and 0s, the pits and lands represent the binary info.

Yes but it’s still an optical signal not electrical.

 

[DonR]

The CD player’s laser reads optically this must then be converted to digital 1s and 0s is that right? The USB signal isn’t optical and therefore they are different so any processing taking place is different? Doesn’t matter whether it’s a Marantz product or anyone else’s.

The CD uses a photodetector to read the laser's reflection off of the disc. At that point and thereafter it's electrical. There is some work to get that electrical signal converted into a bitstream but all that is well established technology. Unless the medium is damaged in some way, a cheap transport should be just as capable of reading the bitstream as an expensive one.

 

[Suffolkhifinut]

This seems to be getting into "intellectually dishonest and deliberately obtuse" territory.

At least you’re admitting it now.
Thanks,

 

[PeteL]

Most likely this is not happening. The USB data is delivered to the DAC and a free running clock is used most likely over an I2S interface. No SPDIF conversion needed, and it would only add jitter if it were there. Further you likely could never hear such a conversion were it to take place. Which again there is no such conversion in this case. You are imagining a difference which isn't there.

Not to be splitting air, but I don't know any DAC chip that will take a USB signal directly, SPDIF yes, altough not always desirable.. Let's just say that there is more similarities between a IIS signal and SPDIF, than a USB Packet stream... Yes it's still just pushing samples, but USB needs hand shaking, needs reclocking to work... You'll need some sort of USB receiver in between, typically a XMOS chip.

 

[Blumlein 88]

Then why if a DAC doesn’t have a USB input a converter is required. Why do computers bother with a USB output if it’s compatible with a standard SPDIF signal, when a coaxial cable would do!

A USB interface is asynchronous. It has the data for each sample in order, but no clock information. The clock is in the DAC. So you feed it with a buffer to keep the data flowing without running out or filling up and the DAC clocks it out. Generally internally an I2S interface does this. If a DAC only has SPDIF, then the DAC has to synch the clock to the incoming data rate. There are various ways to do this, and most increase jitter even if only slightly. The result of synching to that SPDIF clocked data stream then goes over an I2S interface to the DAC chip itself. It is the SPDIF connection which is more complex. You have it backwards in regards to conversions and complexity.

Data from a CD is fed usually into a buffer with the speed of the spinning disc modulated to maintain a more or less constant data rate. From memory the CD will spin between 200 and 500 rpm speeding up as the disc plays.

 

[Blumlein 88]

Not to be splitting air, but I don't know any DAC chip that will take a USB signal directly, SPDIF yes, altough not always desirable.. Let's just say that there is more similarities between a IIS signal and SPD, than a USB Packet stream... Yes it's still just pushing samples, but USB needs hand shaking, needs reclocking to work... You'll need some sort of USB receiver in between, tipically a XMOS chip.

USB signals have no clock related to the playback of digital audio. SPDIF does. DAC chips don't take SPDIF directly, they have to synch clocks to it before using the signal. The I2S interface has separate clock and data signals. You have to peel the data from the clock with SPDIF to feed it to the I2s connection. With USB you get the data and let the DAC clock be the master and clock it out without regard to outside timing sources.

 

[xaviescacs]

Why would it? Both send a digital signal to the same DAC. But I'm an electronics amateur, so I'll let you tell me.

How do you know is the same DAC? For this price/size there could be 10 DACs inside it. I mean, do we assume something about how the device is build and works internally or not? The manufacturer sells this device mainly to play CD/SACD, and I believe users than can be seriously interested in buying this kind of device will be interested in that too. Besides, so there is no need for any assumption if the device is tested with a CD. Testing only the USB input and not the other functionalities, as digital output or headphone jack is the result of a biased approach, ASR biased I mean. Not the end of the world of course, and it can be justified by saying that there is not enough time or that such details are of no importance here, but it makes no sense, it's simply wrong, to say is not necessary because we know the result in advance. No big deal, but let's not fool ourselves.

 

[DonR]

Not to be splitting air, but I don't know any DAC chip that will take a USB signal directly, SPDIF yes, altough not always desirable.. Let's just say that there is more similarities between a IIS signal and SPDIF, than a USB Packet stream... Yes it's still just pushing samples, but USB needs hand shaking, needs reclocking to work... You'll need some sort of USB receiver in between, typically a XMOS chip.

https://www.ti.com/lit/ds/symlink/pcm2704.pdf?ts=1647208300254

Nearly 20 years old now and out of production.

 

[PeteL]

USB signals have no clock related to the playback of digital audio. SPDIF does. DAC chips don't take SPDIF directly, they have to synch clocks to it before using the signal.

I understand that well, but well, as you said USB is Asynchronous, DACS aren't, so yes USB will need some sort of "processing" Sure most DAC will use a SPDIF receiver for clock recovery, at least lately, but it's not obligatory, yes many DAC chips have SPDIF inputs.

 

[threni]

A USB interface is asynchronous. It has the data for each sample in order, but no clock information. The clock is in the DAC. So you feed it with a buffer to keep the data flowing without running out or filling up and the DAC clocks it out. Generally internally an I2S interface does this. If a DAC only has SPDIF, then the DAC has to synch the clock to the incoming data rate. There are various ways to do this, and most increase jitter even if only slightly. The result of synching to that SPDIF clocked data stream then goes over an I2S interface to the DAC chip itself. It is the SPDIF connection which is more complex. You have it backwards in regards to conversions and complexity.

Data from a CD is fed usually into a buffer with the speed of the spinning disc modulated to maintain a more or less constant data rate. From memory the CD will spin between 200 and 500 rpm speeding up as the disc plays.

I don't know how people can argue CD data isn't read into a buffer. A buffer is some storage - usually ram but it could be anything - used temporarily. A CD player reads light into bits, stores that in ram, does a bunch of error detection/correction (which is part of how it can play scratched disks without audible artefacts), then sends that data on for further processing (D->A, or encoding it into packets for USB output). Some have just enough ram to handle decoding; others, such as a little £30 "diskman" I had once, have more and read a few seconds ahead so if you knock it about when walking/running the music can play on uninterrupted (up to a point). Anything which operates on bits IS a computer, essentially. There's a data, which is operated on, and there's the computer, doing the operating. Apart from quantum machines, all computers are turing machines which means they are all 100% identical in terms of what they can do. Some can do a given piece of work in a few microseconds which would take another machine days because of architecture, scale etc but they are in theory all capable of doing the same thing. The archetypal turing machine is simply a long piece of tape which can be played backwards and forwards and there's a machine which operates on data on the tape, reading and writing a bit at a time. Everything else - ram, caching, USB, lasers, buffers, parallel computing, etc etc are just meaningless abstractions.

 

[PeteL]

https://www.ti.com/lit/ds/symlink/pcm2704.pdf?ts=1647208300254

Nearly 20 years old now and out of production.

Sure yes, but the discussion was about Asynchronous USB, It was assumed by me UAC2 what you are referring to is obsolete.

 

[SuicideSquid]

How do you know is the same DAC? For this price/size there could be 10 DACs inside it. I mean, do we assume something about how the device is build and works internally or not? The manufacturer sells this device mainly to play CD/SACD, and I believe users than can be seriously interested in buying this kind of device will be interested in that too. Besides, so there is no need for any assumption if the device is tested with a CD. Testing only the USB input and not the other functionalities, as digital output or headphone jack is the result of a biased approach, ASR biased I mean. Not the end of the world of course, and it can be justified by saying that there is not enough time or that such details are of no importance here, but it makes no sense, it's simply wrong, to say is not necessary because we know the result in advance. No big deal, but let's not fool ourselves.

I don't think this is entirely fair for a few reasons. It's reasonable to assume the DAC is the same because there's simply no good reason for it to be different. If it were, one would think Marantz would advertise this fact with some puffery, but they don't. Second, Amir's using higher-than-CD-quality files to ensure that the results he's reporting aren't simply format-limited, so he's showing off this player in its best possible light by using the USB connecting and 48-192kHz files, rather than using a test CD at 44.1kHz. Finally, lets be respectful of Amir's time - do we really need to double his workload just so he can say "yeah SINAD when playing a CD is close to theoretical limits and jitter is below audible levels, as is true of virtually all CD players manufactured for the past 30 years"?

 

[voodooless]

How do you know is the same DAC? For this price/size there could be 10 DACs inside it. I mean, do we assume something about how the device is build and works internally or not?

Because there is no indication otherwise: The marketing does not mention it, the specs don’t show it, and the inside pictures seem to point at just one DAC implementation.

Not the end of the world of course, and it can be justified by saying that there is not enough time or that such details are of no importance here, but it makes no sense, it's simply wrong, to say is not necessary because we know the result in advance.

Largely yes. It will show the same crappy filter response, especially bad for Redbook. The rest is then immaterial. Playing SACD does probably not use the magical upsampling, so there’s that.

No big deal, but let's not fool ourselves.

Yes sure, I’d like to see some test on the CD drive part, just to make sure it performs equally. But we can also deduce a lot from the data we already have.

 

[Blumlein 88]

Further the SA-10 takes the data from all sources, whether CD, SACD or USB and upsamples them to 256 DSD prior to converting them. There certainly is no direct SPDIF to DSD capability. All sources are converted. It is the MMM or Marantz Musical Mastering.

SA-10 SACD Player - CD/SACD & Digital Hi-Res Formats | Marantz™

The Marantz™ SA-10 CD player features HDAM technology and optimized circuit designs providing quality audio from a disc or streamed from a computer.

www.marantz.com

And yes that includes replay from Discs.

 

[voodooless]

Further the SA-10 takes the data from all sources, whether CD, SACD or USB and upsamples them to 256 DSD prior to converting them.

Got any sources for the SACD upconversion? To do that, one would first need to convert to PCM. That’s basically sacrilege

The spectrum plot Amir made of his SACD also seems to suggest otherwise.

 

[Blumlein 88]

Got any sources for the SACD upconversion? To do that, one would first need to convert to PCM. That’s basically sacrilege

Just the spiel on the Marantz product page linked above.

 

[Chrispy]

Yes but it’s still an optical signal not electrical.

It's digital in both cases.....not sure what your point is.

 

[BDWoody]

It's digital in both cases.....not sure what your point is.

To troll.

Let's stop feeding it.

 

[Chrispy]

To troll.

Let's stop feeding it.

Seems. like many "hifinuts", not to know much about how it actually works Just keep wondering what he's understood in the past of the process and why it matters here. Then again we're talking about a pretty silly cd player.