pi3 sd card vs Pi4 wifi or ethernet sq

[Lacas]

Hi,

I found the sound quality of Pi3B+ playback from SD card vs from NAS through ethernet definitely better, more airy.
Is the sound quality of Pi4 + wifi or Pi4 + ethernet playback equal or better than Pi3B+ playback from SD? 
Is it worth to get a Pi4 1GB for better sound quality through wifi playback?

Thanks
Lacas

[Gekel]

If you have been able to measure a difference in the sound, based on the location and data transfer method used to copy the data from the server/storage location to the moode audio player, then it is easy to check your claims.

Save a file - doesn't matter what type - on a server. Save the same file on your local raspberry. Copy the file from the server to the raspberry and save it under a different filename. Repeat this and use WIFI instead of cable, exchange the cable, use cable with a different length. Exchange the PSU of the raspberry and/or save the file on a USB stick instead of a SD card if you want.

After you have done this, open a console and run

> shasum <filename>

for every file. This will calculate a checksum and show it on the screen. And only if this checksum is different, you have different data and this might result in different audio. Anything (what you think/believe you hear) is placebo.

However you can save yourself some time, because the files will be identical as long as you don't use damaged equipment (e.g. a corrupted hard drive). If my claim would be wrong, we would not only be able to experience differences in audio files but also in all other data transmitted between computer systems, and this would be the end of IT.

To add some more points: in the past days/weeks I tried a lot with my moode player (pi 4). Hooked it to different PSUs, used SSD drives, USB sticks, SD cards to store the system and data, used various network cables and wifi connections, changed the USB cable and even used the machine on three different DACs, all with the same result: The only possible way to change the audio quality is to exchange the chip used to transform the digital input to an analog audio signal. Any change on equipment on the digital side had and has zero influence on the sound quality, because only 0 and 1 are transmitted and these bits don't care if they are sent over copper, air or silver.

[KmPan]

If you have been able to measure a difference in the sound, based on the location and data transfer method used to copy the data from the server/storage location to the moode audio player, then it is easy to check your claims.

Save a file - doesn't matter what type - on a server. Save the same file on your local raspberry. Copy the file from the server to the raspberry and save it under a different filename. Repeat this and use WIFI instead of cable, exchange the cable, use cable with a different length. Exchange the PSU of the raspberry and/or save the file on a USB stick instead of a SD card if you want.

After you have done this, open a console and run

> shasum <filename>

for every file. This will calculate a checksum and show it on the screen. And only if this checksum is different, you have different data and this might result in different audio. Anything (what you think/believe you hear) is placebo.

However you can save yourself some time, because the files will be identical as long as you don't use damaged equipment (e.g. a corrupted hard drive). If my claim would be wrong, we would not only be able to experience differences in audio files but also in all other data transmitted between computer systems, and this would be the end of IT.

To add some more points: in the past days/weeks I tried a lot with my moode player (pi 4). Hooked it to different PSUs, used SSD drives, USB sticks, SD cards to store the system and data, used various network cables and wifi connections, changed the USB cable and even used the machine on three different DACs, all with the same result: The only possible way to change the audio quality is to exchange the chip used to transform the digital input to an analog audio signal. Any change on equipment on the digital side had and has zero influence on the sound quality, because only 0 and 1 are transmitted and these bits don't care if they are sent over copper, air or silver.

[WuBai]

Hi,

I found the sound quality of Pi3B+ playback from SD card vs from NAS through ethernet definitely better, more airy.
Is the sound quality of Pi4 + wifi or Pi4 + ethernet playback equal or better than Pi3B+ playback from SD? 
Is it worth to get a Pi4 1GB for better sound quality through wifi playback?

Thanks
Lacas

Lacas,

Some reports have stated that the Pi3B+ actually sounds better than the Pi4, but I understand that you wish to check if Wifi is better than ethernet.  Have you considered using a USB Wifi dongle?  

You can use a CF-912AC from Comfast and it works perfectly with Moode and it is very inexpensive to try.

Also, don't listen to those 'smart' people telling you that you are crazy and are hearing things that aren't there.  They are the same ones that think all bottled water tastes exactly the same because it is all H2O.  

Allen

[the_bertrum]

They are the same ones that think all bottled water tastes exactly the same because it is all H2O.  

Well since most bottled waters make a virtue of the the mineral content, anyone thinking they are "just H2O" needs to do a little reading.
As for how the bits get to the DAC altering the sound, well I'm in the "bits are bits" camp, but that is simply because I've never heard a difference and no one has explained to me the mechanism at play that could make such a difference.

[H-Award.32]

Hi,

I found the sound quality of Pi3B+ playback from SD card vs from NAS through ethernet definitely better, more airy.
Is the sound quality of Pi4 + wifi or Pi4 + ethernet playback equal or better than Pi3B+ playback from SD? 
Is it worth to get a Pi4 1GB for better sound quality through wifi playback?

Thanks
Lacas

Hi, although some advice to avoid wifi data transfer for hi-end reproduction, the data transfer speed SHOULD theorically be more than sufficient for a perfect reproduction also for highest quality data files (still with an old 182.11 g/n protocol). What I personally experienced is that there might be some latencies in the wifi transfer due to various problems (radio interference, latencies in sending data from the NAS, etc.) and you can perceive that as little distortion: this was really terrible when I tried to use a Mi Box S with Android TV as digital player in my system (but you cannot really compare that stuff with a dedicated machine). Moreover, there are theorical possible interfences between the hardware componentes (Wifi, ethernet, USB) although Archimago in his blog posterd demonstrations that this does not actually affect the data trasnfer quality.
I personally chose to play my files directly from the SD card, because I found it a more convenient solution (it requires the minimum hardware possibile and you don't have to use multiple USB ports at the same time). But I still take the WiFi on for the connection to the LAN.

As first step, you could verify that you set the player to pre-load the file in RAM before playing it: this should solve any slow transfer of datas problems. If you still find a difference between playing from local SD drive versus ethernet/WiFi connection with the NAS, since good quality SD cards are nowdays relatively inexpensive, you should buy one and make this upgrade.

[WuBai]

They are the same ones that think all bottled water tastes exactly the same because it is all H2O.  

Well since most bottled waters make a virtue of the the mineral content, anyone thinking they are "just H2O" needs to do a little reading.
As for how the bits get to the DAC altering the sound, well I'm in the "bits are bits" camp, but that is simply because I've never heard a difference and no one has explained to me the mechanism at play that could make such a difference.

The_Bertrum,

Nobody has explained the mechanism at play because nobody actually knows.  There is so much when it comes to electricity and 'sound quality' which is already a subjective thing that has not been studied at an institutional level.  Unfortunately I don't think there are multi million dollar grants made available for this stuff, because if there was some poor college kid would have to listen to Keith Don't Go ten thousand times in endless rounds of double blind tests and would probably jump off the building to end his suffering.

I ran a test yesterday on my USBridge Signature.  I ordered Allo's EMMC which is an EMMC module with 64GB of memory attached to a SD Card reader.  EMMC is faster than the SD Card and it cost me $100 to have it shipped from India.  I loaded Moode onto it and plugged it in and played music through Wifi.  Guess what?  No difference in sound quality.  None.  I just wasted $100.  Now, if I was falling for placebo and wanting to lie to myself, wouldn't my mind just magically 'hear' a difference to justify the $100 I spend on the experiment?  

So then I transferred a very high res copy of Take Five to the SD Card and went back and forth playing it from my Synology NAS via Wifi and then playing it directly from the SD Card.  The SD Card version sounded calmer and more nuanced.  The Wifi sounded absolutely great, don't get me wrong but when I quickly switched back and forth between the two I found the Wifi sounded a bit louder like there was some extra energy in the music that was very faint.  I listened many times, and was the difference huge?  No, not at all.  I would say maybe 5% difference and the only way I could even hear it is by going back and forth directly and quickly.  So at the end, Wifi and SD Card are both great options and there is almost no difference between them, but I can't say there is 100% no difference between them.

Do I know why there was a slight difference between the sound coming from the spinning hard drive of my NAS to my router to another routher via wifi then to my dongle attacked to my Allo versus running straight from the EMMC SD card?  No.  I would assume that it is due to the transfer but have no idea why or how.

While I am 100% in the bits and bits camp, I must remain open to the fact that I don't know everything and there is still much left to learn.  That is all I am saying and I think that is the essence of science, the constant discovery and experimentation.  Sometimes it leads to crazy things which I don't believe in, like placing $500 magnets somewhere in the room and claiming "Oh, I hear so much more air in the recording" but sometimes it leads to actual discoveries where small incremental changes can be heard.

But at the end of the day, it looks like there are still some 'minerals' in the signal path that we still don't know how to fully account for that make some difference, however small.  Just keep an open mind and enjoy the experimenting process, it is the best part of this hobby in my opinion.  But if everything sounds the same to you, then in some ways I envy you!

Allen

[the_bertrum]

Nobody has explained the mechanism at play because nobody actually knows.  There is so much when it comes to electricity and 'sound quality' which is already a subjective thing that has not been studied at an institutional level.  Unfortunately I don't think there are multi million dollar grants made available for this stuff, because if there was some poor college kid would have to listen to Keith Don't Go ten thousand times in endless rounds of double blind tests and would probably jump off the building to end his suffering.

We don't actually need these tests to postulate on the mechanism.  To verify it perhaps, but not to develop a hypothesis which we could then test.  We could postulate that the spinning of a disc sets up resonances with the electrons in the read head that add bits to the signal that add warmth akin to vinyl, for example, then test to see if that's true.  I'm in the bits are bits camp precisely because there are no hypotheses that attempt to explain how anything in the digital signal transport can add or remove information from the signal that ultimately gets made into sound.
I have elements of my system that I believe make it sound better without proof, I'm not immune to it.  I upsample using SOX to a rate that causes my DAC to bypass its own internal upsampling because I believe that it sounds better when SOX does it.  Upsampling absolutely adds information to the digital signal, so I can see why that might have an effect.  I've no proof it does, but I have a valid hypothesis for why it might.

Now, if I was falling for placebo and wanting to lie to myself, wouldn't my mind just magically 'hear' a difference to justify the $100 I spend on the experiment?  

Not necessarily, the $100 spend might have primed you to hear a clear and marked improvement, and even a subtle one would have disappointed your expectation enough for you to think there was no effect.

Do I know why there was a slight difference between the sound coming from the spinning hard drive of my NAS to my router to another routher via wifi then to my dongle attacked to my Allo versus running straight from the EMMC SD card?  No.  I would assume that it is due to the transfer but have no idea why or how.

And I would add the other component in the system - your brain - into the mix and assume that the well known and documented bias in the imperfect way we hear was the cause; on account of there being plenty of hypotheses that explain how that alters perception of sound versus none for how data signal transport alters the reproduction of sound.

While I am 100% in the bits and bits camp, I must remain open to the fact that I don't know everything and there is still much left to learn.  That is all I am saying and I think that is the essence of science, the constant discovery and experimentation.

Openness, yes.  Discovery and experimentation, yes.  But also and importantly, scepticism and Occam's razor.  While I don't doubt that you hear something when you use Wi-Fi versus SD, I will put the cause somewhere between your ears before I put it between the files and the DAC.  And there is nothing at all wrong with that.  The enjoyment of music is, by definition, subjective and so what goes on between the ears is without doubt the most important part of any Hi-Fi equipment.  Obviously listening for differences in components is part of your enjoyment, and joining in this debate with you is part of my enjoyment.  Maybe one day I'll replace a router and hear things in my music I never heard before, but even then I expect I'll attribute the better music to what I perceive rather than what is actually reproduced unless the router has a label on it that says it automatically adjusts data that flow through it.

[WuBai]

All solid and respectful and reasonable responses and I cannot disagree.  A lot of what I discover in audio 'for me' does appear to be insane since I have no reasonable explanation for it, but I can assure you to the extent that you will believe me, that I am not looking to be proven right or wrong once I test any changes in the system.  As much as possible, I try to remove confirmation bias and I've run into many instances where some expensive component or tweak ended up sounding worse to my ears and the cheaper option turned out to sound better in my system.  But ultimately nothing can be proven because there is no way for one person to get inside the head or perceptions of another.  Even the shape of people's ears and ear canals can lead to different people hearing different speakers in dissimilar ways.  But like you said, as long as we are happy then that's all that matters.
I had once instance where my system just sounded magical, whatever that means.  It had depth and tonality and was amazing to listen to and I really enjoyed it.  Then I said "I'm going to make it even better!"  So I replaced the wire from the IEC connector inside of my my linear power supplies with a VH Audio Airlok silver wire of the same gauge.  These wires had done wonders for my DAC and I was replacing the wire from the IEC to the PCB inside a power supply, probably 2 inches of wire.  Did all the work, put it back together and was so excited to listen to the 'upgrade' after removing that stock 'crappy generic' wire.  Talk about egg on my face, the whole thing went straight to hell.  All the 'magic' was gone and the depth was no more and the tonality shifted and no longer sounded realistic 'to my ears.'  I've made many such discoveries and mistakes along the way.  Just happy to meet someone on the forums that doesn't resort to personal attacks and call people crazy for perceiving things differently.

[Gate45]

Vorrei contribuire a questa interessante discussione.
Ho una casa con quattro stanze cablate con cavo ethernet CAT6 e l'ingresso della casa ha la raccolta delle quattro linee dove ho il modem che esce con uno switch a 4 porte. In due stanze ho due switch "Tp Link" per collegare vari dispositivi compreso il Raspberry PI 3+.
Per archiviare la mia musica utilizzo un QNAP NAS che ho tenuto sullo Switch nella stanza in cui non si trovava il Raspberry PI, perché fa rumore quando il NAS legge i dischi e quindi ho scelto di spostarlo, ma così il percorso dei dati dà il Il NAS su RPi era lungo e passava avanti e indietro tra tre Switch, ma ho pensato che, essendo digitale, non ci sarebbero stati problemi.
Mi sbagliavo, non avevo considerato il rumore che viene aggiunto ai dati passando tra 3 serie di interruttori.
Quindi ho deciso di mettere il NAS sullo switch dove c'è anche l'RPI. La musica è molto migliorata, molto più pulita, più dinamica, ecc. Così ho pensato che ad ogni passaggio da un dispositivo elettronico, la situazione peggiorava.
Il rumore da dove proviene, in gran parte dagli alimentatori e visto che gli alimentatori volanti degli switch non sono di grande qualità, ho sostituito l'alimentatore da 9V con un ottimo alimentatore LDO a basso rumore e ottimo PSRR, che ha migliorato ulteriormente l'ascolto.
Tutti i commenti sono i benvenuti.
Grazie a tutti