[ttt123]

very interesting, so for example if I use a class A amp with it, will it be a hybrid output of class A and D? or will the class A take over?

I think the question is badly phrased.
- the output amplification is Class A, as that is what is connected as the final stage
- The Sony direct digital output fed to a Class A amp is a good signal, because the Sony output is very minimalist and clean, and avoids a double analog amping configuration.
- If you fed an analog A/B amp output into a Class A amp,
- the final output would be Class A
- but the cross over distortion, amp distortion of the Class A/B would just be amplified cleanly by the Class A output. The rule applies, GIGO. There is no magic that can alter the signal that has been created by the class A/B amp, to make the resultant signal better.

 

[Whitigir]

I think the question is badly phrased.
- the output amplification is Class A, as that is what is connected as the final stage
- The Sony direct digital output fed to a Class A amp is a good signal, because the Sony output is very minimalist and clean, and avoids a double analog amping configuration.
- If you fed an analog A/B amp output into a Class A amp,
- the final output would be Class A
- but the cross over distortion, amp distortion of the Class A/B would just be amplified cleanly by the Class A output. The rule applies, GIGO. There is no magic that can alter the signal that has been created by the class A/B amp, to make the resultant signal better.

Double amplifications still exists, but to a lesser degree. However, there is a more optimal way to utilize Walkman phones out as a dedicated line out

 

[justsomesonyfan]

I think the question is badly phrased.
- the output amplification is Class A, as that is what is connected as the final stage
- The Sony direct digital output fed to a Class A amp is a good signal, because the Sony output is very minimalist and clean, and avoids a double analog amping configuration.
- If you fed an analog A/B amp output into a Class A amp,
- the final output would be Class A
- but the cross over distortion, amp distortion of the Class A/B would just be amplified cleanly by the Class A output. The rule applies, GIGO. There is no magic that can alter the signal that has been created by the class A/B amp, to make the resultant signal better.

i'm very confused still, is there any way to explain it in very simple terms? how is it any different to normal double amping?

 

[Sonywalkmanuser]

To add to this, my work experiences confirm that data processing/error correction is not as straightforward as it sounds. Many years of my work were with digital Telephone CO (Central Office switches. The DMS-100) The one I worked on used ECC memory, Parity, and 2 machines running in synchronous mode, so they could match each other to detect errors. (NASA ran triple machines, and looked for a consensus of 2 out of 3 if there was a discrepancy). All these systems increase data reliability, but nothing could make it perfect. Even the NASA triple processor configuration failed sometimes.

My area was real time, and corruption. My first hand experience, over many years of chasing data corruption, is that "it depends". If a header structure is corrupted, there is a good chance it will be detected by integrity check software as it is accessed. (But only if somebody had built in checks due to previous failures.) The corruption can also be picked up by the built in hardware detection for Parity, and ECC. These mechanism detects many/most errors. The hardware detection was preferable, as it did not add an extra load on the processor. Though these would in turn trigger more extensive S/W diagnostics, so you never fully got away from needing to increase the CPU load.

The dual CPUs add another layer of checking, which only commercial machines could afford. But in spite of all these commercial grade protections, data corruption could still happen. Could be caused by EMI, could be component/hardware failure. Could be software errors. Could be overload delaying the diagnostic s/w from running, etc. etc.

Bottom line - data corruption can occur, no matter how you try to protect against them. The symptoms and effects depend on what bits are affected.

I used to get irritated, when I questioned a billing error, and the Service rep told me that "there cannot be an error, as it is all computerized". Really? But that was what I spent a great part of my work doing, trying to understand and correct errors in the system. But then I reminded myself that I don't want to erode the public's confidence in our systems reliability and integrity, by highlighting the "relatively" rare instances of data corruption we had worked on.

The worse one, that is still in my memory, is when an address bit got flipped in the Memory Allocator process, and it caused the system to overwrite/trample large amounts of active data, and crashed the system, which was a major long distance hub switch. This was not recoverable, and the only way to recover was to load in an older backup image. And of course this always occurred during peak traffic hours. (the first time, and 2 other times while attempting to fix the corruption)

This is an example of a single bit, in the wrong place, on the wrong system, able to totally crash a critical system. What are the odds for that happening? Very, very, very low. Buts odds are meaningless if you are the one that is responsible to find/fix it when it happens.

I think there are many people in this world, especially computer programmers will assume that any data errors in the system/electronics will always be detected and corrected or if it is unable to correct then an error code will be generated or the system will stop working all together, like how a C++/Python Compiler will do.

The underlying processing, transmission, memory and data storage is much more complicated than transmitting of "Ones and Zeros" through a wire and storing of files on a recording media. Flash memory after all, is a lossy storage medium, with a increasing probability of losing data (commonly refer to as bit rot) with wear and tear and also through the passage of time.

https://www.flashmemorysummit.com/English/Collaterals/Proceedings/2013/20130812_PreConfA_Camp.pdf

FLAC codec itself comes with limited form of error resiliency, though it doesn't mean that errors do not cause sound quality issues.

https://xiph.org/flac/features.html

• Error resistant: Because of FLAC's framing, stream errors limit the damage to the frame in which the error occurred, typically a small fraction of a second worth of data. Contrast this with some other lossless codecs, in which a single error destroys the remainder of the stream.

 

[Whitigir]

I think there are many people in this world, especially computer programmers will assume that any data errors in the system/electronics will always be detected and corrected or if it is unable to correct then an error code will be generated or the system will stop working all together, like how a C++/Python Compiler will do.

The underlying processing, transmission, memory and data storage is much more complicated than transmitting of "Ones and Zeros" through a wire and storing of files on a recording media. Flash memory after all, is a lossy storage medium, with a increasing probability of losing data (commonly refer to as bit rot) with wear and tear and also through the passage of time.



https://www.flashmemorysummit.com/English/Collaterals/Proceedings/2013/20130812_PreConfA_Camp.pdf

FLAC codec itself comes with limited form of error resiliency, though it doesn't mean that errors do not cause sound quality issues.

https://xiph.org/flac/features.html

• Error resistant: Because of FLAC's framing, stream errors limit the damage to the frame in which the error occurred, typically a small fraction of a second worth of data. Contrast this with some other lossless codecs, in which a single error destroys the remainder of the stream.

The moment you started out quantizing your analog signals, that is the moment you already have a lossy files. Everything there after will further that

 

[Sonywalkmanuser]

The moment you started out quantizing your analog signals, that is the moment you already have a lossy files. Everything there after will further that

At least with Sony, I feel that they are providing you with a whole suite of different signal processing options for you to alter this lossy file to suit your listening tastes. And also all these digital signal processing are very carefully designed, I find the DSPs to be very usable across all types of music. It’s not an over-glorified mp3 player in a luxury chassis that’s for sure. The Sony walkman is a proper hi fidelity portable music player.

 

[Whitigir]

At least with Sony, I feel that they are providing you with a whole suite of different signal processing options for you to alter this lossy file to suit your listening tastes. And also all these digital signal processing are very carefully designed, I find the DSPs to be very usable across all types of music. It’s not an over-glorified mp3 player in a luxury chassis that’s for sure. The Sony walkman is a proper hi fidelity portable music player.

And also, I agree with you that nothing is really bit perfect, technically speaking. It is only a loose term of which you can apply it loosely. Every noise shaping, feedbacks, error corrections are already digital processing or DSP. The matter is which one understand it the most, and processing it the better. In these matters, Sony is the King

 

[ttt123]

i'm very confused still, is there any way to explain it in very simple terms? how is it any different to normal double amping?

Here is one way to look at it:
It may be simpler to understand if I use the definition of Double Amp to mean: Analog amp1 output to Analog amp2 input. This is a double amp (2 analog amps) configuration.
- A typical double amp config is: DAP Digital signal -> D/A -> analog signal -> internal analog amp1 -> DAP Phone out -> External analog amp2 -> IEM . This is a double amp configuration. In this configuration, there is distortion from the D/A conversion, the first Amp stage, and also the 2nd Amp stage. This is considered undesirable, and people try to avoid it. Though if it sounds OK, because each stage is done very carefully, then it is still usable. Just theoretically not desirable

The Sony does not have any analog amp.
- Sony direct digital pulse stream -> low pass filter -> IEM is a signal that is very close to the original Digital signal. So if you amplify this signal(double amp it), it does not have the same negative effects as a typical double amp configuration. There are theoretical negatives still, but much, much less than a typical 2 analog amp config. Effectively, you can look at it as the equivalent of a single amp configuration, as there is only the single external analog amp in play.
- Sony direct digital does not have a D/A conversion stage, nor an analog amp stage (as there is no analog signal involved). The digital signal stays digital all the way, from input to output, and is never converted to Analog. Except at the output, where the LPF does a simple convert by filtering out the non audible frequencies.
SONY DAP digital signal -> reclocked/shaped for more accurate timing/dither -> High current Digital Pulse Stream - > LPF(Low Pass Filter) ->IEM.

There is no Analog amp involved. It uses a high current digital pulse stream to create the high current output to drive the IEM (after Low Pass filtering). The digital stream is passed from input to output. Thus the Sony Phone out is not like the output from a typical DAP.
In fact, it is simpler than a normal DAP single analog amp configuration, as there is no analog amp involved. If you double amp this, on paper, it is a double Amp configuration, but internally, it does not have the same drawbacks as a regular DAP that is double amped. This is the advantage of Sony's Direct Digital implementation.

Another way to look at it is that it is very true to the digital signal and is very R2R like in how direct it is in processing the digital signal, with the added advantage that it does not have an analog amp stage like a typical R2R DAP(or any DAP) would have. Of course a direct digital amp is not a simple/perfect solution either, and it takes a lot of work and engineering knowhow to make a good one. Otherwise, if it were easy to do, there would be a lot more DAPs using a direct digital configuration, and simplifying the hardware parts count.

EDIT: The Sony direct digital configuration is simple, but understanding it and explaining it simply is a challenge!

 

[Sonywalkmanuser]

@Whitigir

I guess the truth may be more shocking to purist analog audiophiles. Analog vinyls can also be derived from DSD digital masters. They have been listening to lossy digital source without knowing about it.
https://www.washingtonpost.com/music/2022/08/05/mofi-records-analog-digital-scandal/

 

[Whitigir]

@Whitigir

I guess the truth may be more shocking to purist analog audiophiles. Analog vinyls can also be derived from DSD digital masters. They have been listening to lossy digital source without knowing about it.
https://www.washingtonpost.com/music/2022/08/05/mofi-records-analog-digital-scandal/

That what I have been saying, the only form of Vinyl is A-A-A records, and they are so rare to find anymore, in good conditions

 

[maceto]

I think the 1AM2 will serve you well enough. It’s not always necessary to go for the top tier models unless you are really going for its specific sound characteristics. After all, there’s plenty of DSP adjustments you can make in the 1AM2 to make the sound to synergize to your iem/headphones.

I’llq

Though Sony still sponsor review products, but very rarely, only reputable personnel like @twister6 that I know by far. The thing about the M2 is the stock shortage

still waiting for him to review it… he’s had it for months

 

[Whitigir]

most of the people can not tell the differences between a digital mastered vinyl VS an analog recording one. It is most obvious to me that the Analog recordings has each instruments uniqueness (tone, reverberant and house resonances) to stand out much further apart and more vividly where as the digitally mastered one almost have a very generic tone and house sound. I have seen and read people saying that by sticking to the Redbook standard, 16/44.1 Nyquist theory, they can replicate the 20-20Khz bandwidth 100%. That is correct, but that is only 100% compared to the time when it was Quantized only . Then allow me to repeat, right when you started quantizing, your analog is already Lossy…errors corrections and so on will add salt to the wounds (it is futile to think you have bit-perfect)

Mathematically if you used the same algorithms to quantize and reverse it to decode, then it can 100% replicate what the original algorithms could have done. But what lost was lost, and no way to retrieve it, except for guessing it based on principle factors (Sony DSEE-HX)

 

[psikey]

600 pages for a DAP that not even 60 people have.

Sonywalkmanuser likes his mSD's

 

[psikey]

Soooo... after 3 pages of science class, people still loving music out the M2 or just using it as an audio test instrument!

 

[Sonywalkmanuser]

most of the people can not tell the differences between a digital mastered vinyl VS an analog recording one. It is most obvious to me that the Analog recordings has each instruments uniqueness (tone, reverberant and house resonances) to stand out much further apart and more vividly where as the digitally mastered one almost have a very generic tone and house sound. I have seen and read people saying that by sticking to the Redbook standard, 16/44.1 Nyquist theory, they can replicate the 20-20Khz bandwidth 100%. That is correct, but that is only 100% compared to the time when it was Quantized only . Then allow me to repeat, right when you started quantizing, your analog is already Lossy…errors corrections and so on will add salt to the wounds (it is futile to think you have bit-perfect)

Mathematically if you used the same algorithms to quantize and reverse it to decode, then it can 100% replicate what the original algorithms could have done. But what lost was lost, and no way to retrieve it, except for guessing it based on principle factors (Sony DSEE-HX)

I feel that this 2019 interview with the Late Ken Ishiwata of Marantz is really worth a read. His views makes a lot of common sense to me. DSD isn’t as bad as a music format as others would make it seem to be, there is a good reason for this format to be used.

https://www.whathifi.com/features/ken-ishiwata-forty-years-of-preserving-the-marantz-sound

Do you have a preference for DSD or PCM music files?
I’m sure you already know the answer – I prefer DSD! I’ve never been crazy about using PCM for music. PCM uses a constant number of bits and a fixed sampling rate – for example, 24-bit/192kHz. Every sample uses the full 24-bits, regardless of the requirement of the signal. I’ve long recommended the use of variable bit rates to avoid waste. As it is now, as much as 90 per cent of a 24-bit/192kHz data stream is wasted.

On top of that, the sound of DSD is close to the analogue original. That’s is why we decided to convert all PCM signals to (1-bit) DSD inside the DAC of our KI Ruby player.

Can a DAC’s sonic performance be judged from its specifications?
Unfortunately, specifications don’t tell you about sound quality. That’s not just for DACs, it’s for everything. Those specifications are all based on static measurements, but music is dynamic and there are many other parameters that influence performance.

The biggest problem of DACs of today is not their specifications. Those are all wonderful. If you only look at specifications, then you will feel confident it will sound good.

Back in the 80s, we had DACs using BiMOS technology, but today all the ICs are based on CMOS. With CMOS you can get high speed, low consumption and high specifications. Today’s handheld devices and computers require those things, but for serious audio? I have a big doubt.

You know musical dynamics require power, more power means high consumption... well, it’s just my view.

Do you think Class D circuitry works for high-end products?
There are so many different amplifier designs. My personal view is that we can make high quality amplification with different technologies. There are so many designs of conventional and Class D amplification today, they all sound different and measure differently. Both can work excellently, but you have to take care if you want the sound of Class D to compare to traditional analogue amplification technology.

Measurement or listening... which takes priority?
All measurements we do are static. Yes, they are essential to find out how amplifiers are working, but they don’t tell you about the sound quality. You have been in my listening room in Eindhoven. We’ve spent so many hours in that room listening and fine tuning all Marantz products. Sometimes that listening forces us to change our basic design, despite the fact that the measurements are okay.