Marantz SA-10 Review (SACD Player & DAC)

[Blumlein 88]

There have been players that lit up an led to indicate interpolation rather than accurate data reading. They were never common because such stuff is very, very rare to encounter. The Reed-Solomon interleaved code works very well to minimize errors.

 

[srrxr71]

Stereophile magazine had in the 90s the most comprehensive explanation of CD technology. It’s complicated but I remember things like 8 to 14 encoding because having 2 lands in a row cannot be precisely determined. So the encoding is to make sure all 8 bit signals have some form of pit to land transition. So the 14 bit code is sent to a lookup table to get the 8 bit code. It’s pretty complicated and most of it was accomplished with analog electronics controlling the motor and laser sled. I believe they had to split the beam to create an interference that guided the sled to the maximum signal position.

I wish I had that issue with me.

 

[Blumlein 88]

It had nothing to do with two lands being hard to decode. Remember it is the pit to land transition that is being read. EFM, eight to fourteen modulation was to prevent unrecoverable errors. If you had a scratch that takes out 20 data samples and those samples are on the disc in order you cannot recover from it. EFM was part of what scrambled the order of the samples. That way a scratch that takes out 20 data samples will take out data samples in the original order than are not contiguous. So when re-assembled in order the missing data samples will be scattered about with good values on either side. The error correction can then re-create the value of the missing sample. It won't guess it will know what it was. If the data loss becomes bad enough it will interpolate, but the system was so good this rarely occurs.

 

[voodooless]

The Reed-Solomon interleaved code works very well to minimize errors.

It will fully correct a gap of up to 3.5 KByte . That almost 900 samples of audio!

 

[Blumlein 88]

It will fully correct a gap of up to 3.5 KByte . That almost 900 samples of audio!

That is why they used Reed-Solomon to transmit digital data from quite a few satellites.

 

[srrxr71]

It had nothing to do with two lands being hard to decode. Remember it is the pit to land transition that is being read. EFM, eight to fourteen modulation was to prevent unrecoverable errors. If you had a scratch that takes out 20 data samples and those samples are on the disc in order you cannot recover from it. EFM was part of what scrambled the order of the samples. That way a scratch that takes out 20 data samples will take out data samples in the original order than are not contiguous. So when re-assembled in order the missing data samples will be scattered about with good values on either side. The error correction can then re-create the value of the missing sample. It won't guess it will know what it was. If the data loss becomes bad enough it will interpolate, but the system was so good this rarely occurs.

I see. How would it know if read one land or 2 land? Was it keeping time?

 

[Sal1950]

Marantz are starting to look more and more like a brand trading on the legend and willing to sacrifice true performance in order to utilize circuits that have market influence power among audiophile buyers.

What grand ole US audio company hasn't been gobbled up by foreign entities making time on the brand name?

 

[Herbert]

Stereophile magazine had in the 90s the most comprehensive explanation of CD technology. It’s complicated but I remember things like 8 to 14 encoding because having 2 lands in a row cannot be precisely determined. So the encoding is to make sure all 8 bit signals have some form of pit to land transition. So the 14 bit code is sent to a lookup table to get the 8 bit code. It’s pretty complicated and most of it was accomplished with analog electronics controlling the motor and laser sled. I believe they had to split the beam to create an interference that guided the sled to the maximum signal position.

I wish I had that issue with me.

The analog electronics allow perfect alignment of the servo circuit. Splitting the beam into three was one of two methods.
The side beams helped in guiding the lens and sled.
But Philips, Denon, Toshiba, Technics used sigle beam lasers that track very well up to this day, even in portables:

CD player - Wikipedia

en.m.wikipedia.org

 

[Blumlein 88]

I see. How would it know if read one land or 2 land? Was it keeping time?

Part of EFM helps with that too by specifying how many ones and zeroes can be consecutively on the disc.

Try reading this:

Eight-to-fourteen modulation - Wikipedia

en.wikipedia.org

 

[Herbert]

This is a very good read: http://www.exp-math.uni-essen.de/~immink/pdf/cdstoryoriginal.pdf

 

[Xyrium]

Oh boy, where does the money go to?

Copper backplate and copper fasteners?

Edit: Just saw the interior photo. Add, copper xformer cover, copper standoffs, copper isolation walls, and of course, a smattering of ferrite beads for good measure.

 

[pablolie]

This is a very good read: http://www.exp-math.uni-essen.de/~immink/pdf/cdstoryoriginal.pdf

Seems to be unpermissioned for those of us without Essen TU accounts...

PS: Why is it so many CD player reviews turn into generic discussions on CD tech history and its merits? I'd hate to be one of the CD inventors always getting asked the same questions... :-D

PS2: It's quite telling that 40 years later nobody has come up with a more widely adopted standard for high end audio that is more widely agreed upon as being at the very limit of human hearing... and that innovation seems to steer rather towards monetization/DRM rather than promoting biotech approaches to improve human hearing... :-D I respect you, CD Format...

 

[SuicideSquid]

To bring it back around to the Marantz, I'd be very curious to see how it performs relative to the Reavon UBR-X200, which is a new universal disc player. Reavon looks like they're trying to position themselves as the 'New Oppo':

REAVON UBR-X200 - 4K UHD Dolby Vision HDR10 SACD Blu-ray player

The Reavon UBR-X200 Blu-ray player is compatible with the Dolby Vision format which provides HDR image technology, SDR/HDR conversion, SACD playback, dual HDMI, XLR output

reavon.com

Stereo XLR outs plus 7.1 RCA outs and universal disc compatibility including CD, SACD, DVD Video/Audo, bluray and 4k Ultra, and it's less than 1/4 the price of the MSRP of this Marantz unit.

 

[restorer-john]

Is this actually true that the error correction is mathematically perfect? Just because the CD will keep playing (filled in with interpolated data) doesn't mean it's accurate, or the original data?

The error correction is perfect. The data is not interpolated during error correction. Interpolation (linear) is a second stage fall back if the data cannot be mathematically corrected. The CD standard will allow for 2.4mm of linear burst data loss with 100% perfect correction. That's right, you can drill a ~2mm hole in a CD and it will play perfectly on a good machine.

My test discs have graduated data loss tracks used to test the abilities of the various players I restore/repair. They are certainly not created equal.

 

[restorer-john]

PS: Why is it so many CD player reviews turn into generic discussions on CD tech history and its merits?

Because people tend to forget (or don't know) how good and resilient the format really is/was.

 

[DWI]

Seems to be unpermissioned for those of us without Essen TU accounts...

PS: Why is it so many CD player reviews turn into generic discussions on CD tech history and its merits? I'd hate to be one of the CD inventors always getting asked the same questions... :-D

PS2: It's quite telling that 40 years later nobody has come up with a more widely adopted standard for high end audio that is more widely agreed upon as being at the very limit of human hearing... and that innovation seems to steer rather towards monetization/DRM rather than promoting biotech approaches to improve human hearing... :-D I respect you, CD Format...

PS3: I'm listening to J S Bach sonata for viola da gamba and obbligato harpsichord No. 3, a thoroughly delightful piece, on vinyl from an Archiv box set. It's such a quiet pressing and such glorious sound that I sometimes wish the whole digital thing had never happened at all.

PS4: 16/44 PCM has never failed me, but sometimes it's nice to shout at a blob plugged in under the sofa and get Amazon HD streaming in 24/192 PCM from 10 ceiling speakers.

 

[srrxr71]

Part of EFM helps with that too by specifying how many ones and zeroes can be consecutively on the disc.

Try reading this:

Eight-to-fourteen modulation - Wikipedia

en.wikipedia.org

Thank you. It’s always incredible to see how much work and thought went into making the compact disc.

 

[srrxr71]

The error correction is perfect. The data is not interpolated during error correction. Interpolation (linear) is a second stage fall back if the data cannot be mathematically corrected. The CD standard will allow for 2.4mm of linear burst data loss with 100% perfect correction. That's right, you can drill a ~2mm hole in a CD and it will play perfectly on a good machine.

My test discs have graduated data loss tracks used to test the abilities of the various players I restore/repair. They are certainly not created equal.

Could you please share which brands if any were better about error correction?

 

[Blumlein 88]

Thank you. It’s always incredible to see how much work and thought went into making the compact disc.

It was a huge under-taking. Philips was a huge company and decided they needed Sony in on it to introduce the 1st digital format for music playback. It could have been a disaster if it had never taken off and discs and players remained too expensive it could have disappeared as a failed format. There has been nothing like it in recorded music since.

 

[don'ttrustauthority]

So do you mean one should spend for looks, weight, size, brand, big front fascia etc etc instead of measured performance ? If measured performance is inaudible after a certain point then there is certainly all the more "no point" in spending $7k on an equipment. Isn't it ? Or you yourself will spend "$7k" on this even after knowing about this measured performance ?

I'd be more likely (which is not much) to buy something this expensive if it was hand made, or there was some kind of cool technology, but for this much money, it better be a speaker (pair). Seriously, this is too much for anything else.

My question is, how does measured performance add value after audibility is surpassed? This is the assumption behind your statement.