[ac67]

Any change to how third party apps play native 44.1/16 files?

Yes so far apps we tested are 192KHz/32bit up-converted

 

[ac67]

So will UAPP or Tidal extract MQA detail with the first unfold and that then being fed out at 32/192 ? Are you hearing the unfolded MQA like intended? The upscalling will not degrade the MQA?

We suspected the first unfolding still occured, as this is handled via software. The unfolded 88.2KHz/96KHz get upsampled.

 

[psikey]

We suspected the first unfolding still occured, as this is handled via software. The unfolded 88.2KHz/96KHz get upsampled.

So if the 88.2 or 96 then gets up-converted to 32/192 your certain there is no quality loss/change ? Sounds as good as it would if bit-perfect ?

Same with Spotify? A 16/44 up-scaled to 32/192 no loss/change of quality ? I was always advised that say converting a 320kbps to higher quality is actually bad, are they not same thing? I'm no expert on the theory.

 

[Sonywalkmanuser]

I have been using Neutron player to playback 16bit 44.1KHz flac outputing using OpenSL ES driver 32bit floating Point 192000.

So far it sounded fine to my ears. No wobbly or wonky sounding ASRC sound from upsampling 44.1kz to 192khz.

Infact I find that the 32bit floating point output has a very smooth DSD like sound.

Recommend for you guys to download neutron player evaluation copy to try out for yourself.

 

[ac67]

So if the 88.2 or 96 then gets up-converted to 32/192 your certain there is no quality loss/change ? Sounds as good as it would if bit-perfect ?

Same with Spotify? A 16/44 up-scaled to 32/192 no loss/change of quality ? I was always advised that say converting a 320kbps to higher quality is actually bad, are they not same thing? I'm no expert on the theory.

Upsampling changes the sample-and-hold. With correct algorithm, the converted analogue signal will be smoother.
Conversion to 320kbps is different.

 

[ac67]

Am wondering, with Crinacle and Toranku changing IER-M9 ranking created much interest, likewise will the new firmware for ZX507 bring more people to be 'poisoned' by this DAP?

 

[Sonywalkmanuser]

Anyone notice that with the new firmware even with direct source on, there's two soft popping sound when you switch tracks or pause and play?

Previous firmware does not have popping sound when you use direct source mode.

 

[T.R.A.N.C.E.]

Anyone notice that with the new firmware even with direct source on, there's two soft popping sound when you switch tracks?

Previous firmware does not have popping sound when you use direct source mode.

I have had the soft pops with direct source on with all the firmware's. I like it, the sound is reminiscent of vintage gear. I only use direct source on, I prefer it's sound.

So if the 88.2 or 96 then gets up-converted to 32/192 your certain there is no quality loss/change ? Sounds as good as it would if bit-perfect ?

Same with Spotify? A 16/44 up-scaled to 32/192 no loss/change of quality ? I was always advised that say converting a 320kbps to higher quality is actually bad, are they not same thing? I'm no expert on the theory.

MP3 is different, converting MP3 to "higher quality" is bad because it starts of as a lossy file, it cannot be converted to lossless, by up-sampling the MP3 you are only exacerbating the artifacts in the lossy file.

Like others already mentioned, MQA is first unfolded in app such as tidal or uapp, the unfolded file is then up-sampled using float, there is no degradation, same with any PCM like 16/44 and up.

 

[Fran Jau]

Anyone notice that with the new firmware even with direct source on, there's two soft popping sound when you switch tracks or pause and play?

Previous firmware does not have popping sound when you use direct source mode.

I used to have this problem, but now after the update its gone, the other way around, I was noticing this when using 3.5mm to 4.4mm adaptor

 

[Sonywalkmanuser]

Previously it wasn't a good idea to upsample mp3 but technology has improved as we are now in the Artificial Intelligence age of things. Machine Learning and algorithms. Sony has been feeding it's DSEE AI upsampler with its own Sony music record label music. You are talking about Sony's own DXD masters and etc. Now the upsampler knows how to upsample better than before(see article below).

Of course DSEE AI won't make mp3 sound like 24bit 192KHz flac or something but in my experience with DSEE HX on itunes 256kbps AAC files, I must say DSEE did make AAC files sound much more closer to it's 16bit 44.1khz CD originals.

Google translate required:
https://www.sony.jp/feature/products/dseehx/

Next, please tell us how you created AI for "DSEE HX".
Chinen: In order to build a good DNN, a process of “learning” using a large amount of data and appropriately “evaluating” the results is required. And on both sides, the power of the Sony Group, which has a music label, was very helpful. Sony Music Entertainment's many high-resolution sound sources are used to create optimal algorithms. As a technical issue, it was very simple. We upscaled the CD sound source and the compressed sound source, and pursued as a numerical target how close to the high resolution sound source, which is the “correct answer”.

 

[GRUMPYOLDGUY]

I see your second point, more psychoacoustics territory.

However, I'm not sure why your bringing filter arguments into upsampling debate? My understanding was that upsampling was done in early digital audio days in high end equipment to avoid harsh filters that cause sonic issues.

Calculated Interpolation Upsampling;
The interpolator is going to be incredibly accurate - especially whilst also increasing the bit depth at the same time as the sample rate. Just up-ing the sample rate would cause issues however raising the bit depth at the same time is a pretty good solution;

Easy maths examples;
Lets take two samples from a 16 bit audio file at 48kHz Nyquist. Lets double the sample rate (so for simplicity in the example; adding another sample point in the middle)

Sample Point A = 2345
Sample Point B = 2355

Interpolated middle sample (Lets call it 'i1') would be i1 = 2350

However same example, but it results in a non-whole number
Point A = 2345
Point B = 2346
i1=2345.5

Interpolated sample would be 2345.5, which would be rounded down/up as 16bit and 24bit audio uses Integers (which can only be whole numbers - all depending on implementation, but almost 100% universally) and introduce quantisation noise (not an artefact as such), but add to what you referred to as correlated noise (which, as you say would be well below what we can hear, just as 16bit quantisation noise is in a normal recording).


However, 32bit audio is done with the float format (it can be done with integers, however it is normally always done with Float), meaning you can store a decimal number (unsure if Sony uses single or double float, but assuming single)
Therefore
Point A - 2345
Point B = 2346
i1 = 2345.5

The decimals in a float will be accurate up to 7 decimal points.


So, lets increase the sample rate from 48kHz to 192kHz (4 times greater, therefore 3 more samples between point A and B)
Point A = 2345
Point B = 2246

Interpolated sample point i1 = 2345.25
Interpolated sample point i2 = 2345.50
Interpolated sample point i3 = 2345.75

The 32-bit float DAC will be able to read these additional points (Walkman spec sheet says it can read 32bit float PCM .WAV files).

I think that Sonys reason for upscaling all the individual apps audio output, is that it's a work around solution. The apps are not able to directly control the DACs sample rate, therefore if the DAC is set to the highest sample rate it can handle, and if the apps audio output is upsampled - the listener will still receive the higher audio. This is of course conjecture as I have no idea about the specifics of how Sony has implemented this.



I'm happy to stand corrected on this, but the above is why I do not understand the aversion to upsampling in this scenario. If its your opinion you don't like the sound of upsampling, more than happy with that because at the end of the day, everybody's ears are different and there is no point listening to something if you don't like the sound of it.

The spec for the 507 states is decodes floating point wav files. I do not see where it states that the DAC directly synthesizes floating point words. Those files may very well be (and likely are) quantized at the output stage.

More simple math, 24 bit input data (fixed point) and max 32 bit output data means 8 bit coefficients for just a single tap. That doesn't include bit growth through the add tree which is log2(number of taps). Obviously 8 bits is not going to accurately quantize the coefficients, so there is necessarily some rounding and or truncation in the process.

Again, I highly doubt the artifacts are audible... I've already said as much. But they are unnecessarily present when the alternative (bit perfect) does not have this problem. Just play back my files the way they are!! I don't want some engineers at Sony telling me they know I'll like something better and making the decision for me.

 

[GRUMPYOLDGUY]

Anyone notice that with the new firmware even with direct source on, there's two soft popping sound when you switch tracks or pause and play?

Previous firmware does not have popping sound when you use direct source mode.

I had it show up on the previous firmware after the first time I disabled and re-enabled direct source. Even after re-enabling direct source I had the popping between tracks. I assumed it was a bug in the firmware.

 

[Sonywalkmanuser]

I think bit perfect is still something desirable as it helps to save on battery and reduce heat output.

 

[rlw6534]

Does native Walkman player support DLNA on that bad boy?

Not that I have found. I wish it did for NAS streaming around the house...

 

[AlternativeNose1]

I was a little confused by some of the points in your reply. So I've tried to respond as best I can.

The spec for the 507 states is decodes floating point wav files. I do not see where it states that the DAC directly synthesizes floating point words. Those files may very well be (and likely are) quantized at the output stage.

WAV is just file name for a raw PCM representation. I'm making assumptions as to how Sony have implemented their upsampling, I don't actually know. The DAC is listed as being able to handle 32bit float wav, but that doesn't mean that the DAC is doing the conversion (or synthesis as you put it). It might be software based or even being sent to a different chip in the Walkman (I suspect the CPU as its an ARM chip and Float processing is bread and butter for modern chips, lighting fast and very accurate https://developer.arm.com/architectures/instruction-sets/floating-point ).

Also I'm confused by the use of Quantized at the output stage? Quantiztation is performed when the signal is converted from Analog to Digital. https://en.wikipedia.org/wiki/Quantization_(signal_processing)

More simple math, 24 bit input data (fixed point) and max 32 bit output data means 8 bit coefficients for just a single tap. That doesn't include bit growth through the add tree which is log2(number of taps). Obviously 8 bits is not going to accurately quantize the coefficients, so there is necessarily some rounding and or truncation in the process.

32 bit float is not just 8bits more;

In my post I do mention that most audio uses 32bit float vs int and therefore I assume Sony is doing that here.

Additionally I very much doubt that Sony are just multiplying/adding binary words against each other and therefore dealing with bit-growth. They would probably be using the C language and converting the integers to floats, then calling up samples from the audio buffer, using the Nyquist formula combined with an interpolation formula, filtering the finished result and then exporting the final result to the buffer. But when we are talking about quadrupling the sample rate to 192kHz - we are dealing with halves, quarters, eights which is were floating point numbers excel as they naturally work in Base 2.

Again, I highly doubt the artifacts are audible... I've already said as much. But they are unnecessarily present when the alternative (bit perfect) does not have this problem. Just play back my files the way they are!! I don't want some engineers at Sony telling me they know I'll like something better and making the decision for me.

If thats how you like listening to music I would recommended avoiding any streaming services then and just use the DAP to play local files with Direct Source turned on. However you have to take into account that every DAC has a different design and therefore sound. Sony have their own DAC whereas Fiio, Drangonfly, Apple all use off the shelf DACs. Buying a Sony audio product very much gives you a different sound (one that I like a lot).

Like I said, I was a little confused by some of the things you put in your response, so if I've gotten the wrong end of the stick and I'm missing the point of your argument please let me know.