View Full Version : Do small dedicated computer players all sound the same?
If one selected a popular playback software like Moode and set it up in a number of the small dedicated computers like the Raspberry Pi, feeding identical DACs would they all sound the same?
Rothchild
19-10-2016, 19:28
Subjectively or objectively?
Subjectively or objectively?
If I knew what that meant I would reply. Leave me with Google for a few minutes :)
I think objectively fits the bill.
walpurgis
19-10-2016, 19:56
I think objectively fits the bill.
Is that a subjective opinion though? :)
Rothchild
19-10-2016, 22:05
Yes they sound the same (given the same dac/cables and software) - just as doing the same sum on different brands of calculator will give you the same answers.
Yes they sound the same (given the same dac/cables and software) - just as doing the same sum on different brands of calculator will give you the same answers.
Well I'm somewhat perplexed. From experience, almost every component of a Hi Fi system can have an influence on SQ, including cables, cabinet construction, phono sockets and even fuses.
Here, all of a sudden, we have a a wide choice of computer driven players and an equally wide choice of software to run them - yet everyone who talks on this topic claims they all sound the same. I find that very strange, although I have nothing to suggest otherwise.
To draw a comparison, CD transports are the nearest approximation and they certainly sound different.
dave2010
20-10-2016, 16:53
Yes they sound the same (given the same dac/cables and software) - just as doing the same sum on different brands of calculator will give you the same answers.
Up to a point, Lord Copper. I guess you've never come across calculators which give different results, or read about calculations done using Microsoft Excel, versions of which can sometimes give significantly different "answers".
Rothchild
20-10-2016, 17:11
Double post, sorry, see next page.
Rothchild
20-10-2016, 17:11
IMO a cd transport is not a good analogy, it's an electro mechanical device for retrieving digital data (and that data is subject to error correction so could already be arriving at the digital output of a transport in a slightly 'order' different time each time the disk is run).
In your initial q you specified 'feeding the same dac' so I assumed the same analogue signal path too (so all the bits you mention later such cable etc are discounted too)
With those variables out of the way my calculator analogy is actually the more accurate one. A digital music file is essentially a sum/problem and the chip on a computer is a calculator, the 'codec' (be that .wav .aiff .flac .alac .mp3 etc etc) is simply the instruction set that prescribes what order to do the sums in - it doesn't matter if the chip driving the calculator is made by Intel, AMD or ARM it should come up with the same answer to the same sums, otherwise it's defective.
Rothchild
20-10-2016, 17:27
Up to a point, Lord Copper. I guess you've never come across calculators which give different results, or read about calculations done using Microsoft Excel, versions of which can sometimes give significantly different "answers".
No, you're right, I've never heard of correctly functioning calculators where different models give different results to the same calculation, do you have an example of one?
Different versions of excel may well work differently, (causing inconsistencies due to what is essentially 'user error') but I'm not convinced that changes or challenges the fundamentals of how maths works - we'd be utterly buggered if mathematical constants (eg I=V/R) were not consistently calculable, nothing electronic would work or even be designable.
IMO a cd transport is not a good analogy, it's an electro mechanical device for retrieving digital data (and that data is subject to error correction so could already be arriving at the digital output of a transport in a slightly 'order' different time each time the disk is run).
In your initial q you specified 'feeding the same dac' so I assumed the same analogue signal path too (so all the bits you mention later such cable etc are discounted too)
With those variables out of the way my calculator analogy is actually the more accurate one. A digital music file is essentially a sum/problem and the chip on a computer is a calculator, the 'codec' (be that .wav .aiff .flac .alac .mp3 etc etc) is simply the instruction set that prescribes what order to do the sums in - it doesn't matter if the chip driving the calculator is made by Intel, AMD or ARM it should come up with the same answer to the same sums, otherwise it's defective.
I am really playing devils advocate here, but if all RPI type computers are 100% accurate in the signal they give out, how come their SQ can be improved with better power supplies.
Rothchild
20-10-2016, 18:32
Most of those claims relate to boxes where the dac is also taking from the Pi source (afikt) the signal improvements are (IMO) effected to the analogue part of the circuits in question, they have no effect on the decoding and (bitwise) movement of data around the device - the initial proposition / thought experiment sets the analogue part of the equation away from the question at hand (as I read it).
Another (slightly geeky) way to demonstrate this is to use an 'MD5 Hash' it's a sort of rudimentary security tool used (mostly) for ensuring that the file you're downloading hasn't been interfered with. The person who makes the file generates an MD5 hash number (a long string of letters and numbers) of the file and publishes it with the download, you download it and run the MD5 on it, if it's the same file it will generate the same number, one bit different and it won't.
I'm sure there are windows tools to do it so you can try it yourself, make a document and hash it, then edit one character in the document and hash it again, different hash number, again it doesn't matter whose silicon you run the program on the same file will yield the same hash. (indeed you can use it to demonstrate that flac is actually lossless, hash the .wav source, convert to .flac and then convert back to .wav and hash it again, if one single character in the whole file is different you can't get the same hash code back
If one selected a popular playback software like Moode and set it up in a number of the small dedicated computers like the Raspberry Pi, feeding identical DACs would they all sound the same?
They dont actually.
Have some experience in this area and computer audio is actually analogue, at least in the output stage.
So not all the same.
walpurgis
20-10-2016, 23:07
Would you care to read that through again? :)
(yes. I know the 's' is next to the 'a' and the 'i' is next to the 'o' ;))
My power supply improvement with the RPi is with an spdif connected dac so no power from the RPi is being used by the dac.
Would you care to read that through again? :)
(yes. I know the 's' is next to the 'a' and the 'i' is next to the 'o' ;))
yes..:)
Rothchild
21-10-2016, 06:53
Would be much easier if OP had specified subjectively, as we could all agree that the answer is at least 'probably' if not 'almost certainly'.
Just because a voltage pulse is used to fill or clear a register on a chip it doesn't make digital audio analogue.
In my experience the best gains are from taking SPDIF out of the computer, and instead using a dedicated USB to SPDIF converter with its own power supply. This way, the computer becomes much less important, proper software and configuration aside.
Sent from my SM-G920F using Tapatalk
Would be much easier if OP had specified subjectively, as we could all agree that the answer is at least 'probably' if not 'almost certainly'.
Just because a voltage pulse is used to fill or clear a register on a chip it doesn't make digital audio analogue.
The clock signals are very much analogue. They are timing events, not just a 1 or 0. It is about "when" the clock signal hits. This is why I believe jitter and clock drift are the main culprits in making transports sound different.
Sent from my SM-G920F using Tapatalk
I've used a Pi for a number of years as my streamer. Into, this past year, a Chord 2Qute. The sound I thought was always pretty good. I've tried USB straight from the Pi, Coax from a Digi+ and a number of other boards. I then bought on trial a Sonore microRendu..... computers do not all sound the same! This little thing has transformed my listening. It simply annihilated the Pi, full stop.
So much has it increased my pleaseure with music that I've gone on to buy a Mutec reclocker and Uptone LPS-1 PSU. Each time an increase was had in performance. I'm enjoying my music like I never thought possible again, I'm one happy bunny!
Well, that's my opinion, ymmv...
The clock signals are very much analogue. They are timing events, not just a 1 or 0. It is about "when" the clock signal hits. This is why I believe jitter and clock drift are the main culprits in making transports sound different.
Sent from my SM-G920F using Tapatalk
Indeed.
Which was what I was saying...maybe my timing was off?.. :)
Rothchild
13-11-2016, 17:09
Indeed.
Which was what I was saying...maybe my timing was off?.. :)
I think a lack of resolution meant not all the information was conveyed as accurately as you'd intended ;-)
I've used a Pi for a number of years as my streamer. Into, this past year, a Chord 2Qute. The sound I thought was always pretty good. I've tried USB straight from the Pi, Coax from a Digi+ and a number of other boards. I then bought on trial a Sonore microRendu..... computers do not all sound the same! This little thing has transformed my listening. It simply annihilated the Pi, full stop.
So much has it increased my pleaseure with music that I've gone on to buy a Mutec reclocker and Uptone LPS-1 PSU. Each time an increase was had in performance. I'm enjoying my music like I never thought possible again, I'm one happy bunny!
Well, that's my opinion, ymmv...
I love comments like 'annihilate' :) That safely removes any doubts that differences are subtle and tells what I always suspected, that a computer with dedicated build for audiophiles can outperform the RPi, which I already class as excellent. Out of interest did you compare it against an RPi with dedicated linear power supply?
So just to get it straight the micro rendu take a signal from a NAS via a network lead and outputs a USB signal.
Stratmangler
14-11-2016, 18:23
So just to get it straight the micro rendu take a signal from a NAS via a network lead and outputs a USB signal.
No it doesn't.
There isn't a signal on the NAS.
There isn't any signal until the Microrendu generates it.
Rothchild
14-11-2016, 18:23
I agree that clocks are important in digital audio and will clearly make a difference. They are also another potential source of 'apples and oranges' issues in answering the OP. Where the clock source is in the chain differs in different setups and routings and the clock is not necessarily part of the the computer driving the playback.
For instance an i2S buss (such as is used for most 'onboard' PiDACs) doesn't actually communicate clock (which is why all those add on cards have a crystal clock generator on them), SPDIF is an unclocked transfer protocol too (so relies on there being a clock in the DA element), USB is something of a basket case in that there is synchronous and asynchronus protocols which work differently etc.
I will still happily argue that there is not an audible difference caused by using (for example) an AMD implementation of an X86 chip over an Intel one, they are just big fast calculators and sums is sums. Differences occur because systems and setups are diverse and different, in a lot of ways, but if clocks and convertors are dealt with 'off-site' then they should all sound the same.
I will undertake to test this myself as I've ordered a Nano Pi Neo (actually for a non-audio project) it has an i2s buss so I should be able to use one of my PiDACs on it and will be able to check for myself and report back.
No it doesn't.
There isn't a signal on the NAS.
There isn't any signal until the Microrendu generates it.
Sorry, just my lack of knowledge. Would I be right in saying the that the Microrendu uploads the music from say a NAS using the ethernet connection etc. If not, how should I phrase it.
Never got on with Pi's for computer audio to be honest... :scratch:
Sounded "good", don't get me wrong, but (to me) can be bettered quite easily.
Great for what it is, and if I hadn't heard any other kit I'd have been happy.
Have a good one myself, built in linear PSU, IQAudio Dac+, on paper pretty up there I guess.
Was looking to look into developing a Pi based Server at one point (The mythical Squeezebox replacement!), but direct comparison with my own kit; Logitech Transporter fed by Ethernet from my Windows based server/NAS into my Dac...really showed it the door.
Was hoping it would be as good/better but past a certain level I felt it struggled a bit.
Ah well!
Its the diminishing returns thing I think?
They DO sound good, but for that "extra" magic you need to aim a little higher.
Things like quality of power supply, regulation, isolation of power rails, electrical noise elimination and the like wont be state of the art from Pi.
Just my take on it, I've heard and compared the Pi and going back to the OPs question, in my experience they DONT all sound the same.
Stratmangler
14-11-2016, 19:20
Sorry, just my lack of knowledge. Would I be right in saying the that the Microrendu uploads the music from say a NAS using the ethernet connection etc. If not, how should I phrase it.
It's more accurate to say that the Microrendu accesses its data from the NAS and generates a signal to communicate with an offboard USB DAC.
Your NAS doesn't have any music on it - it's all just data.
There shouldn't be any requirement for the Microrendu to be physically connected to the NAS either.
The NAS needs to be physically connected to your network, as does the Microrendu, but the two can exist remotely from each other.
I love comments like 'annihilate' :) That safely removes any doubts that differences are subtle and tells what I always suspected, that a computer with dedicated build for audiophiles can outperform the RPi, which I already class as excellent. Out of interest did you compare it against an RPi with dedicated linear power supply?
So just to get it straight the micro rendu take a signal from a NAS via a network lead and outputs a USB signal.
Well, annihilate seemed like the proper word at the time... I certainly would not go back to a Pi! Sorry for the delay in replying, forgot I'd posted in this thread. I use a Pi with Max2Play running on it with a 2TB WD Red drive attached in a cupboard. That runs LMS. This is attached via Ethernet into a gigabit switch. Then an Ethernet cable into the mR and USB into my Mutec. Then coax out into the 2Qute DAC. I use Squeezepad on my iPad (which I think is an absolutely brilliant app). Hope that clears that bit up.
I have tried an LPS on the Pi, but it was a while ago, before I got the mR so I never did a direct comparison. I wasn't bowled over with it tbh.... it might be different for you of course. Once the mR was in the system though, the difference was instant and obvious, no need to try and hear a change, it was readily apparent.
£70 for a Pi setup and £650 for the mRendu, I'd hope to hell it would beat it.
I looked at the mRendu but it can't do the streaming services I use, and I'm not going to pay the ridiculous price for Roon only to be tied to Tidal ( which is great ) but unsuitable for the 5 of us that use my Google Music account.
The SoTM M-200 is supposed to be even better than the mRendu and it's £200 cheaper. But again it's only a Roon/Tidal endpoint so no good for me either.
I'll stick with my Pi/Piano 2.1/Kali setup as it sounds way better than my previous Pi/IQAUDIO setup, at least that is until either Sonore or SoTM add more streaming services ;)
£70 for a Pi setup and £650 for the mRendu, I'd hope to hell it would beat it.
I looked at the mRendu but it can't do the streaming services I use, and I'm not going to pay the ridiculous price for Roon only to be tied to Tidal ( which is great ) but unsuitable for the 5 of us that use my Google Music account.
The SoTM M-200 is supposed to be even better than the mRendu and it's £200 cheaper. But again it's only a Roon/Tidal endpoint so no good for me either.
I'll stick with my Pi/Piano 2.1/Kali setup as it sounds way better than my previous Pi/IQAUDIO setup, at least that is until either Sonore or SoTM add more streaming services ;)
mike, you using google music now? is that through pi or the chromecast?
mike, you using google music now? is that through pi or the chromecast?
Pi using Bubble uPnP ( on android ), I only use Google so my 2 daughters, 'potential' son in law and 2 grandkids can all use the same account at £14.99 a month, ok it's only 320kbps but decent enough for the price and I mostly use it for 'browsing' artists with a view to buying LP versions or CD if no LP available ;)
£70 for a Pi setup and £650 for the mRendu, I'd hope to hell it would beat it.
....and it does! You pays your money...
....and it does! You pays your money...
Don't get me wrong, I'm not dissing the mRendu or the SoTM, I'd love either but they're too limited in what they do to be any good in my system.
That and the post Brexit price rise of over £150 on the mRendu is a bit off-putting.
£70 for a Pi setup and £650 for the mRendu, I'd hope to hell it would beat it.
I looked at the mRendu but it can't do the streaming services I use, and I'm not going to pay the ridiculous price for Roon only to be tied to Tidal ( which is great ) but unsuitable for the 5 of us that use my Google Music account.
The SoTM M-200 is supposed to be even better than the mRendu and it's £200 cheaper. But again it's only a Roon/Tidal endpoint so no good for me either.
I'll stick with my Pi/Piano 2.1/Kali setup as it sounds way better than my previous Pi/IQAUDIO setup, at least that is until either Sonore or SoTM add more streaming services ;)
Mike this change in tone regarding the SQ of small computers is a bit of a revelation on here. Up till now most posters have said essentially that all computers sound the same apart from a change of power supply.
Long may it continue. I am happy to hear of the SoTM M-200 as well as the microrendu, and any other albeit expensive dedicated computer players that can improve SQ.
£70 for a Pi setup and £650 for the mRendu, I'd hope to hell it would beat it.
If you look back at the start of this thread the common response was that they all sound the same.
If you look back at the start of this thread the common response was that they all sound the same.
Yeah, must admit it's not a view I agree with as my Pi setup just now is superior sounding to how it was with the IQAudio DAC I had on it to begin with.
Bass is much better defined with more slam, everything sounds much more defined and soundstage is much wider, all this without the loss of musicality, it really gets your foot tapping:)
If you look back at the start of this thread the common response was that they all sound the same.
They don't...seriously.
I designed, built and sold computer audio kit that had "Sound quality that few CD Players can dream about, A degree of refinement that only the best record players can deliver". (Apparently)
I wouldn't disagree with him...though it wasn't me that said it, but him.
A bog standard laptop would do the same job ON PAPER, but wouldn't sound the same.
Point is...all digital sources don't sound the same.
They really don't.
Held back from getting involved but I see a lot of oversimplification about digital audio based on misinformation read online.
(and no...I'm not selling anything :))
They don't...seriously.
I designed, built and sold computer audio kit that had "Sound quality that few CD Players can dream about, A degree of refinement that only the best record players can deliver".
Happen to agree with the magazine reviewer...though it wasn't me that said it, but him.
A bog standard laptop would do the same job on paper, but wouldn't sound the same.
Point is...all digital sources don't sound the same.
They really don't.
Held back from getting involved but I see a lot of oversimplification about digital audio based on misinformation read online.
(and no...I'm not selling anything :))
Thanks Gazjam for that insight.
There are a few areas when working with computers that seem to be obvious;
1. Best linear power supply you can afford
2. No devices (other than the computer) to draw on the computers current.
3. Computer to be dedicated to audio duties, nothing else.
4. All cabling including USB to be the best quality you can afford
But what other areas need to be considered?
Thanks Gazjam for that insight.
There are a few areas when working with computers that seem to be obvious;
1. Best linear power supply you can afford
2. No devices (other than the computer) to draw on the computers current.
3. Computer to be dedicated to audio duties, nothing else.
4. All cabling including USB to be the best quality you can afford
But what other areas need to be considered?
Computer and DAC 'isolation' :eyebrows:
Computer and DAC 'isolation' :eyebrows:
Yes, for me the best gains are from using an independently powered, galvanically isolated (using a pulse transformer) USB to SPDIF converter, as well as proper software and configuration. This will make more difference than investing on the PC side.
The whole exercise is about "minimising loss" in my experience.
Hardware?
Switchmode PSU's aren't fit to be near a hifi system. Linear or at least quality regulated is what you want.
PCs are noisy, especially the cheaper Pi's, so counteract it as much as you can with shielded internal cables.
(Not that you can do much tweak wise with a PI...but it can sound good.)
Maplins CB Radio linear power supplies tweaked to run 12V (as per Stan Beresfords original reccy for his Dacs) feeding individual power rails to the Motherboard are to me the best "dont sell a kidney for your hifi" options.
Linear 5V psu's for the SSD.
Sure there would be improvement running £000's worth of linear PSU's, but these are good enough.
Use SSDs for music storage if you can, if not then dont use "green" or cheaper hard drives.
Windows based?
Have a dedicated PC stripped down to the absolute minimum and tweak windows to have as little running as possible. Windows Server 2012 R2 if possible.
Switch off all power saving states in BIOS and in Windows settings.
Always thought USB output was one of the worse sounding ways to get audio from a computer, asynchronous or not.
If you have to, use a USB/SPDIF converter as it isolates any noise from the USB (of which there is a lot), the choice of USB cable matters a lot less with this.
Unless you get into a separate USB card with it's own linear psu's...which is different.
All about isolation and minimising loss...
My favourite was the Musical Fidelity V-Link 192, not the cheaper one..this sounds significantly better.
If your really feeling clever, modify the Vlink to have an balanced XLR output, sounds better again.
Way better IMO than direct USB output to your Dac.
Best sounding way to get your file based audio files to your Dac IMO is via Ethernet.
Checksums make sure what goes out is what went in.
That's why Squeezebox works so well and other manufacturers are finally catching up.
From there, if you can use balanced XLR then do.
To me, XLR is the best sounding digital hookup to your Dac.
Spend a couple of years (and some ££'s) looking for the best sounding digital hookup to my Dac from my Transporter.
AES/XLR shows everything else the door, which is weird when you consider digital is digital? :)
Digital cable CAN sound different..went through a lot till I found the one that sounded the best.
To me, there must be mechanical/electrical/analogue factors which affect sound quality, even with digital hookup.
XLR balanced has protections built in which (on paper) shouldn't kick in unless its a very long connection...my experience says different.
My ears tell me this.
Software?
Buy Roon and forget about it.
Seriously...
Tried most of them (Jriver, 2 PC setup running JPlay, LMS, Mac mini running Audirvana Plus, Ammara, and XXHIghend) and Roon just sounds better that anything else I've tried.
It just works.
Computer audio (once you get sound quality to a certain level) is all about the interface...and Roon is both the best sounding and most pleasurable way to experience Computer Audio.
If you dont want to self build, look into Roon and buy a Roon "Endpoint" that suits your budget.
All the way from Rasberry Pi's to DcS gear by way of Squeezebox.
Take yer pick. :)
The whole exercise is about "minimising loss" in my experience.
Hardware?
Switchmode PSU's aren't fit to be near a hifi system. Linear or at least quality regulated is what you want.
PCs are noisy, especially the cheaper Pi's, so counteract it as much as you can with shielded internal cables.
(Not that you can do much tweak wise with a PI...but it can sound good.)
Maplins CB Radio linear power supplies tweaked to run 12V (as per Stan Beresfords original reccy for his Dacs) feeding individual power rails to the Motherboard are to me the best "dont sell a kidney for your hifi" options.
Linear 5V psu's for the SSD.
Sure there would be improvement running £000's worth of linear PSU's, but these are good enough.
Use SSDs for music storage if you can, if not then dont use "green" or cheaper hard drives.
Windows based?
Have a dedicated PC stripped down to the absolute minimum and tweak windows to have as little running as possible.
Switch off all power saving states in BIOS and in Windows settings.
Always thought USB output was one of the worse sounding ways to get audio from a computer, asynchronous or not.
If you have to, use a USB/SPDIF converter as it isolates any noise from the USB (of which there is a lot), the choice of USB cable matters a lot less with this.
Unless you get into a separate USB card with it's own linear psu's...which is different.
All about isolation and minimising loss...
My favourite was the Musical Fidelity V-Link 192, not the cheaper one..this sounds significantly better.
If your really feeling clever, modify the Vlink to have an balanced XLR output, sounds better again.
Way better IMO than direct USB output to your Dac.
Best sounding way to get your file based audio files to your Dac IMO is via Ethernet.
Checksums make sure what goes out is what went in.
That's why Squeezebox works so well and other manufacturers are finally catching up.
From there, if you can use balanced XLR then do.
To me, XLR is the best sounding digital hookup to your Dac.
Spend a couple of years (and some ££'s) looking for the best sounding digital hookup to my Dac from my Transporter.
AES/XLR shows everything else the door, which is weird when you consider digital is digital? :)
To me, there must be mechanical/electrical/analogue factors which affect sound quality, even with digital hookup.
XLR balanced has protections built in which (on paper) shouldn't kick in unless its a very long connection...my experience says different.
My ears tell me this.
Software?
Buy Roon and forget about it.
Seriously...
Tried most of them (Jriver, 2 PC setup running JPlay, LMS, Mac mini running Audirvana Plus, Ammara, and XXHIghend) and Roon just sounds better that anything else I've tried.
It just works.
Computer audio (once you get sound quality to a certain level) is all about the interface...and Roon is both the best sounding and most pleasurable way to experience Computer Audio.
If you dont want to self build, look into Roon and buy a Roon "Endpoint" that suits your budget.
All the way from Rasberry Pi's to DcS gear by way of Squeezebox.
Take yer pick. :)
Sorry its a term I havent seen before, but what is a Room Endpoint?
Roon Endpoint is the 'receiving' hardware/software end in the Roon-iverse. It usually outputs either a digital audiostream or an analog signal.
A bit like a Squeezebox Classic if that makes any sense?
(which in itself could act as a Roon Endpoint, for that matter!)
Regards /Mike
Thanks guys. I've read a bit about Roon which certainly seems impressive in a multi media, multi room system.
But if the only player is a hard drive storage linked to a computer, then feeding a USB signal, does it offer benefits over say Moode.
Multi room is only one of the things Roon does well.
To me, computer based audio is all about 2 things...
Sound quality and User experience?
To me, Roon is unsurpassed in both of these.
https://roonlabs.com/
Squeezebox user here for the last few 5 years or so, dabbled with Jriver and Mac previous to that and before THAT...Squeezebox user since 2007.
Having discovered how good the Transporter was as a digital transport, LMS was all I would ever need (and it was free!!)
Then Roon came along...
I DO miss some of the plugins with LMS that I used, as well as using both the cool Transporter displays for album / artist info etc, but Roon really is an "endgame" solution for me.
Sound quality, interface, access to your music library...browsable links between songs,artists,albums...all unsurpassed.
I use Tidal, but no other music service though, but those that do might find Roons lack of interaction with Spotify/Qobuz etc a problem.
But its constantly developing and other music services are on the roadmap.
Big hitters like Ayre, PS Audio, dCS, Auralic and Bryston have teamed up with Roon, and there's a helluva recommendation right there.
Not free, but worth every penny IMO.
Nice posts Gaz interesting experiences.
Sent from my SM-G920F using Tapatalk
Rothchild
23-11-2016, 18:12
If you look back at the start of this thread the common response was that they all sound the same.
The original question was; given all things being the same (DAC, Software etc) would changing the computer part of the equation make a difference. I still believe the answer to this is no.
But the thread has moved on in a natural way and the question has adapted itself to (something like) can one make computer based systems sound different? (generally by swapping and changing elements of the system) well then yes, clocks, power supplies etc all make a difference, but these are not 'the computer' per-se, this is why studios spend good money on these elements, but I've never seen recordists debating whether AMD sounds different to Intel (because it doesn't).
There was also a lot of debate a few years ago about whether or not recording and mixing software has a 'house sound' associated with it (as many thought they could hear differences) this led to quite a lot of fairly deep testing that resolved that there wasn't (most differences were down to different implementations of 'Pan-Laws' which adjust levels when signals are panned around the stereo field).
speed of processor too, so in a way it does make differences...but I generally agree, the processors aside I guess the computing part just... computes, and its answer should be same every time or not at all.
Fair point Grant...
But.. :)
I've noticed differences on things you wouldn't expect: Speed and type of Ram used, type of PCi-E internal cable used and even shielding internal ribbon cables and different options for internal grounding can make a difference.
Always surprised me what made a difference, even in things like processor socket TYPE, never mind processor speed!
Fastest processor wasn't always the best option for audio, but sometimes (just to confuse things) it was depending on the type of files you were playing.
I'd say have an open mind and don't buy into some of the dogma you read online.
Experiment and try things for yourself...in my experience most things make a difference.
Aye, my win 7 laptop doesnt stream amazon music well for some reason although it really should as its just a highish mp3 code and pc is able to play video and audio off utube(not streamed it back over the ethernet) yet the better pc does it easily with little proc power used. No idea why... so your right too.. some things that are not maybe obvious count I guess at times. sound wise, sure the cables etc will have an effect.. external ones do so internal ones should.. I was just speculating on the actual processing should be same, albiet that faster ones will do it faster. Might go back to using lappy actually and move on my rpi if I move my pioneer into main system. Connection wise something will have to go, and i can double up on the chromecst with the laptop streaming my flac music via fubar and bubble. Be a shame but there is limited room.
Fair point Grant...
But.. :)
I've noticed differences on things you wouldn't expect: Speed and type of Ram used, type of PCi-E internal cable used and even shielding internal ribbon cables and different options for internal grounding can make a difference.
Always surprised me what made a difference, even in things like processor socket TYPE, never mind processor speed!
Fastest processor wasn't always the best option for audio, but sometimes (just to confuse things) it was depending on the type of files you were playing.
I'd say have an open mind and don't buy into some of the dogma you read online.
Experiment and try things for yourself...in my experience most things make a difference.
realysm42
24-11-2016, 05:14
**I've not read the entire thread**
I think the OP's original question is flawed (flaw highlighted):
"If one selected a popular playback software like Moode and set it up in a number of the small dedicated computers like the Raspberry Pi, feeding identical DACs would they all sound the same?"
It could be interpreted in a number of ways;
1.What does small mean?
2.What does dedicated mean (this is a spectrum, not binary)?
3. What are you actually asking, I read this as "if you were to set up a couple of identical computers and run this test, would you get any variation?"
You've not clearly defined what the variation you'd like to test is.
Regardless, I've made a number of changes to my vanilla pc:
Upgraded from standard usb output to SOTM pcie usb out
Upgraded the clock for above card
upgraded from generic psu (to all sorts but ended up with) a paul hynes sr7ehd with 4 rails powering all components separately
All of these have provided a more than subtle sound upgrades, espcially the paul hynes psu. I was surpirsed at how good the clock upgrade for the pcie card was, too. This has lead me to buy a clock upgrade for the motherboard as well (not done yet).
Even if they all sound the same, they can all be made to sound a great deal better.
Hope this is of some use.
I use a Super Flower Leadex Gold 750W PSU for my self-build PC, the fan doesn't need to run, it's SMPS, however: http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=370 (stunning performance and measurements, no switching spikes that you normally get). Regardless, I have a filtered power strip for 'noisy' stuff like TV, router, computer, away from the power strip for the HiFi.
realysm42
24-11-2016, 21:52
"It's all in the implementation, maaan"
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.