Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[pcstockton]

yeah, but its uncompressed on the fly...   
not sure how memory play affects this though?

Exactly....  I love how WAV and FLAC CANNOT under ANY circumstances sound different to someone.  Yet Memory Playback can, and is embraced/supported by MC...  Hilarious. 

Beauty is in the eye of the beholder.  All stop.  Period the end.

-FLACtacular Patrick

[JimH]

Exactly....  I love how WAV and FLAC CANNOT under ANY circumstances sound different to someone.  Yet Memory Playback can, and is embraced/supported by MC...  Hilarious. 

Supported, yes.  Not embraced.  It's there because some audiophiles wanted and it was easy to add.  It makes no difference in the sound.

It is a tricky business, building a high quality audiophile-grade player, without chasing every myth about audio.  We want all our customers to be happy.  But we also want to feel good about what we do.

[pcstockton]

Supported, yes.  Not embraced.  It's there because some audiophiles wanted and it was easy to add.  It makes no difference in the sound.

Is this the reason for the 64 bit path and how you handle volume etc???

Where do you draw the line? 

In the end is it not an exact case of 'Beauty is in the eye of the beholder'?

If someone thinks it sounds better, than it does [to them].  Isn't that all that matters?  Why must this discussion consistently devolve to accusation of improper testing, hallucination, lying, delusion, and broken DACs???

I cannot personally hear the difference and I tested it fairly significantly when I first started ripping CDs years ago.  But I am certainly open to the possibility that others can and do with their specific system setups.

Should all DACs, amps, preamps, cables etc, ALL sound identical.  I would happily have anyone to my house to show them clearly otherwise.

This extreme instant dismissing of others' experiences is as annoying to me as the opposing view that little blocks of wood placed throughout the room has an effect on the sound quality.

Something tells me the deltas are miniscule but perceivable on a resolving enough system.  RF, microphony, vibrations, etc can all affect analog signals which at some point we are all dealing with.

Cheers!
Patrick

[Matt]

If someone thinks it sounds better, than it does [to them].  Isn't that all that matters?

No, I don't think so.

I was talking with Chris Connaker (of ComputerAudiophile.com fame) last night about this very topic.  I said:

The thing I wrestle with is this: if I tell you something stupid, like that large page memory helps audio quality, and when you listen the placebo effect makes you really like the results more, does that mean it's alright to tell you the stupid thing?

Is it better if I believe the stupid thing myself and don't know how to or don't bother to do proper testing?

Personally I think both are unethical.

[InflatableMouse]

You once asked me the same thing and I said it mattered too. I was wrong. You bring up an interesting thought there, thanks.

[Matt]

I should clarify that I think it's fine for a user to reach any conclusion they like.  For some, part of the fun of the hobby is trying lots of little tweaks that may or may not make a difference.

But JRiver and other vendors should be held to a higher standard.  If we say something sounds better, we better be able to back it up with testing and logic.

[Vocalpoint]

Should all DACs, amps, preamps, cables etc, ALL sound identical.  I would happily have anyone to my house to show them clearly otherwise.

This extreme instant dismissing of others' experiences is as annoying to me as the opposing view that little blocks of wood placed throughout the room has an effect on the sound quality.

Something tells me the deltas are miniscule but perceivable on a resolving enough system.  RF, microphony, vibrations, etc can all affect analog signals which at some point we are all dealing with.

Whenever this topic comes up - there are generally three angles in which to interpret the end result - the provable, the environment and finally your brain. With respect to WAV vs FLAC:

1. The Provable is all mathematics (and code) and cannot be disputed under any circumstance. When analyzed by whatever method you choose (I like bit for bit compare in something like WaveLab) - the actual stream of 1's and 0's from either file are identical and as such - there cannot be any difference whatsoever in the source material.

2. The Environment - can have limitless variables and can contribute significantly to the "apparent" sound of one file over another. However - if taking #1 into account and using the identical "environment" (aka the same test system in the same room with the same equipment) - there should not be any difference in sound whatsoever in the source material.

3. Your Brain - here's where anything can and will happen. Even if #1 and #2 are being evaluated correctly - your brain can (and will) interpret things differently. If deep down inside - you want to hear a difference - you will - even if the math is provable and the room and system are the same. If you insist that the placebo effect cannot exist - you will hear something. If you spent 400 dollars on a audio cable - your brain reaches out to justify this purchase and will remind you relentlessly that if you spent that kinda of scratch - you NEED to hear something positive.

And so it goes in the world of the audiophile.

Me - I am all about practical so I start and stop with #1 in this instance. The other two angles have no bearing on this nor do I want to expend any energy worrying about them.

Cheers,

VP

[windowsx]

I don't see any problems in pleasing audiophiles with features that can easily be done. I love how MC18 having less loopback and wish one day they can offer option to completely bypass all DSP stuff leaving only playback behind optimized for lower secondary buffer size. Options in MC18 can be set down to 2000ms in which some apps like foobar can be set to just 50ms.

[retro]

Matt,
what was Chris Connaker's replies to your questions..?

I am one of many that actually hear differences between flac vs. waw, cables, vibration control etc.
On my main system, that is. On my girlfriends system, I can't. Different systems have different resolution capabilities.
Nothing strange with that, is it..?

As for blind-tests, they aren't as bulletproof as many here seem to think.

Chris Connaker himself prefer aiff, btw...

[pcstockton]

I should clarify that I think it's fine for a user to reach any conclusion they like.  For some, part of the fun of the hobby is trying lots of little tweaks that may or may not make a difference.

But JRiver and other vendors should be held to a higher standard.  If we say something sounds better, we better be able to back it up with testing and logic.

No one is asking you to declare something you dont believe.

Also the contrary must also be the case.  You cannot back up "with testing and logic" the fact that WAV does not sound different to that original poster.  Many people in the Naim world hear meaningful differences between the two in their systems.  Perhaps it is due to the method by which their servers decompress the FLAC files.  Maybe it taxes the system to the point where the analog sections are slightly compromised.

Should we assume a "noisy" electrical environment with badly engineered switch-mode power supplies can in NO WAY affect the analog output of the device?

In a purely digital environment FLAC and WAV might be completely identical and any extra processing to decode the FLAC may be far under any threshold.

But I find it odd that people dont think the analog signal can differ given what is going on.

Not all turntables sound the same, motors and power supply arrangements vis-a-vis vibrations and microphony are a HUGE factor.  Why would this be any different than a DAC?

I do agree that this is not noticeable to me with my Naim Kit (Naim DAC>102-Napsc>180>Arivas).  I cannot hear a difference between FLAC and WAV.  But if I connect my sound card with coax or BNC to my DAC and run everything on the same circuit and have tons on noise in the line with electrically noisy STBs, HDTVs and PCs.  I am open to the possibility that something can potentially affect my analog outputs on my DAC which certainly affects sound quality.

Cheers!
Patrick

[pcstockton]

Whenever this topic comes up - there are generally three angles in which to interpret the end result - the provable, the environment and finally your brain. With respect to WAV vs FLAC:

1. The Provable is all mathematics (and code) and cannot be disputed under any circumstance. When analyzed by whatever method you choose (I like bit for bit compare in something like WaveLab) - the actual stream of 1's and 0's from either file are identical and as such - there cannot be any difference whatsoever in the source material.

2. The Environment - can have limitless variables and can contribute significantly to the "apparent" sound of one file over another. However - if taking #1 into account and using the identical "environment" (aka the same test system in the same room with the same equipment) - there should not be any difference in sound whatsoever in the source material.

3. Your Brain - here's where anything can and will happen. Even if #1 and #2 are being evaluated correctly - your brain can (and will) interpret things differently. If deep down inside - you want to hear a difference - you will - even if the math is provable and the room and system are the same. If you insist that the placebo effect cannot exist - you will hear something. If you spent 400 dollars on a audio cable - your brain reaches out to justify this purchase and will remind you relentlessly that if you spent that kinda of scratch - you NEED to hear something positive.

And so it goes in the world of the audiophile.

Me - I am all about practical so I start and stop with #1 in this instance. The other two angles have no bearing on this nor do I want to expend any energy worrying about them.

Cheers,

VP

Agreed and most of this might be #3.  BUT and a big but, there are thousands of things within the purview of #1 that you are not considering.  Yes all of the 1s and 0s may be identical.  But there are tons of other quantifiable/testable components like RF, vibrations, distortion, unstable power supplies etc...  You must remember that at some point the digital becomes and analog signal.  

In the digital domain you basically have possible flipping of bits and jitter/timing issues. Even if these are real issues, I doubt WAV vs FLAC has much to do with this.

But you dont listen to the 1s and 0s.  If you are telling me there is NO WAY you can affect analog signals I think you are very wrong.

If that was the case every single amp, preamp, in the world would sound identical.  I dont think anyone is claiming that.  Although there are many people who think speakers are the biggest factor.  They are not.

-Patrick

[Vincent Kars]

I should clarify that I think it's fine for a user to reach any conclusion they like.  For some, part of the fun of the hobby is trying lots of little tweaks that may or may not make a difference.

But JRiver and other vendors should be held to a higher standard.  If we say something sounds better, we better be able to back it up with testing and logic.

I can only second both point of views.

Unfortunately most vendors have all kind of bold claims not backed up by any testing (no, I don’t think user testimonials can be considered evidence).
The alternative is the Foobar’s Peter or JimH way, claim that there is no audible difference.
Saves a lot of testing 
The only one I know trying to backup his claims with measurements is Cisc’s Cplay.

The problem is that our systems are different.
I can imagine that e.g. a discrete soundcard, fully exposed to what is going one inside a PC will yield different results compared with a galvanic isolated outboard DAC feed by an async protocol.
This makes proper testing very complex.

Logic dictates that there is a special case where we might expect identical results: no DSP because different algorithms yield a different sound.
However, even if this is the case we can’t rule out differences. Processing in bursts or doing it smooth might have an impact.
Logic is not easy, what looks logical at face value might be far more complex the deeper you dive into how the system works.

Personally, I do regret that JRiver doesn’t have a true memory playback.
If it was possible to have all I/O and all decoding done before playback starts, we can eliminate a hell of a lot of variables regarding lossless file formats.
IMHO, true memory playback would make JRiver independent sound quality wise of (lossless) file formats.

[Vocalpoint]

Agreed and most of this might be #3.  BUT and a big but, there are thousands of things within the purview of #1 that you are not considering.  Yes all of the 1s and 0s may be identical.  But there are tons of other quantifiable/testable components like RF, vibrations, distortion, unstable power supplies etc... You must remember that at some point the digital becomes and analog signal.

I look at #1 from a pure math and code standpoint. Math is absolute and code cannot change when being executed. And the 1's and 0's are just that.
  

In the digital domain you basically have possible flipping of bits and jitter/timing issues. Even if these are real issues, I doubt WAV vs FLAC has much to do with this.

Possible does not mean "is".

But you dont listen to the 1s and 0s.  If you are telling me there is NO WAY you can affect analog signals I think you are very wrong.

If one is playing a WAV thru Media Center to a specific system in a specific room at a specific time and then plays the FLAC directly after it without touching anything - there is no possibility of anything affecting any analog signal output. It should and will be identical....unless of course - something - anything is changed.

But if you are inferring that a typical D/A interface suddenly decides to "change" it's character because it thinks it "sees" a WAV coming thru...or decides it's going to process that set stream of 1's and 0's differently one day over the next - I am going to call BS immediately.

If that was the case every single amp, preamp, in the world would sound identical.  I dont think anyone is claiming that.  Although there are many people who think speakers are the biggest factor.  They are not.

Of course - each system imparts it's own sonic signature - but as described above...if #1 is constant (and it always will be) and the #2 (the Environment - that is imparting the "signature") remains constant - it is impossible for a WAV and FLAC created from the same source data to 'sound" different.

If they do - or your brain makes you think that they do - it's definitely #3 at work...which is the case 99.9% of the time when this topic comes up.

Cheers,

VP

[Vincent Kars]

If one is playing a WAV thru Media Center to a specific system in a specific room at a specific time and then plays the FLAC directly after it without touching anything - there is no possibility of anything affecting any analog signal output. It should and will be identical....unless of course - something - anything is changed.

There are 2 changes, the amount of I/O and the amount of processing needed for the decoding.
You can’t rule out that these differences can affect the timing of the DAC e.g. more input jitter, stray signals over the ground plane, RFI, EMI, etc.

[Vocalpoint]

There are 2 changes, the amount of I/O and the amount of processing needed for the decoding. You can’t rule out that these differences can affect the timing of the DAC e.g. more input jitter, stray signals over the ground plane, RFI, EMI, etc.

So you are implying that MC is changing the variables "on the fly" based on load and file type? That I/O and "processing" is resulting in a different streams of 1's and 0's per file type? If that's the case why would anyone would use or trust MC as an audiophile player at all.

And yes - I would immediate rule out any "overhead" causing timing issues until such time that it could mathematically proven otherwise. The 1's and 0's either make it to the DAC and or they don't. I do not believe you could have one set of 1's and 0's for a WAV and a completely different set for a FLAC.

If that could be proven - I would of course accept it.

VP

[Vincent Kars]

PCM audio is sample + time step.
Alter a sample and you get a different sound.
Alter the time step and you get a different sound.
The “problem” with digital audio is that only the bits (the samples) are digital.
The clock is fully analog.
Any disturbance of the clock driving the DA conversion will have its impact.
The bits are the bits as much as the timing is the timing.
At bit level we don’t have any problem, any lossless format produces identical bits.
But do the different electrical patterns always produce the same sample rate jitter?
If that could be proven - I would of course accept it.

[pcstockton]

So you are implying that MC is changing the variables "on the fly" based on load and file type? That I/O and "processing" is resulting in a different streams of 1's and 0's per file type? If that's the case why would anyone would use or trust MC as an audiophile player at all.

And yes - I would immediate rule out any "overhead" causing timing issues until such time that it could mathematically proven otherwise. The 1's and 0's either make it to the DAC and or they don't. I do not believe you could have one set of 1's and 0's for a WAV and a completely different set for a FLAC.

If that could be proven - I would of course accept it.

VP

I am talking about things that affect electrical signals AFTER the DAC.  Although digital output can vary in it quality.  But lets not get into that.

But yes let us assume the 0s and 1s are ALWAYS perfect.  No flipped bits etc.  Everything checks out.  Identical copy.

I dont see how that is the end of it.

If the computer/streamer/soundcard, or any combination thereof, or whatever is outputting the digital signal is contributing noise, RF, vibrations etc that are passed along to the analog outputs of the DAC, it can affect sound quality.  This could be through a coax or BNC cable, it could simply be through the mains, or other source of transmitting vibrations, e.g. computer sitting on the same rack as the hifi.  For an example of mains pollution you can come listen to what happens to my 5.1 HT system when my lady uses her hairdrier.  Noticeable distortion and sibilance that ANYONE can hear.  This is an extreme case of course.

A electrically noisy PC can certainly pollute mains in the same way.  If you happen to be DAC'ing inside said computer with a PCI soundcard or something, or you connect via USB, these vibrations and whatnot can be passed along to the DAC where in the analog realm they can affect SQ.

Like i mentioned, I do not personally hear a difference between FLAC and WAV, but the fact that FLAC requires more processing power and therefore potentially more noise does make sense to me.

Once again, I TOTALLY agree on the 1s and 0s being exactly that.  I am not talking about flipped bits or jitter.  I am talking about why putting a Mac Mini directly on top of my DAC and plugging it in the same mains block, might make SQ differences.  This can happen without any consideration of WAV vs FLAC.

I think we may be talking past each other here.  I agree with everything you say up to the point where you jump to it being "impossible" for FLAC and WAV to sound different. 

In the digital domain and within MC everything is identical and kosher.  With memory play, I dont think we are doing anything within MC or digitally.  I use Memory play because I am under the impression that the less the PC is doing, the quieter it is and the less influence it could have over my analog signal (outputs) from the DAC or even the pre.

A turn table is a very good example of how analog signals can be compromised by very minute RF, vibrations etc...  The DAC is not different in this regard.  It is certainly more robust and less susceptible, and the effects may be close to the threshold of perception, but impossible???  I dont think so.

Cheers!
Patrick

[struct]

If someone thinks it sounds better, than it does [to them].  Isn't that all that matters?  Why must this discussion consistently devolve to accusation of improper testing, hallucination, lying, delusion, and broken DACs???

For what it is worth (i.e. nothing), my comment about the dac was not meant to devolve the conversation.  If there is difference heard, so be it, no problem with that. 

The system must be wonderful to hear a difference and there must be a good pair of ears listening.  This makes me jealous .  It does however seem reasonable to me that it is unlikely that the wav/flac difference would be the only thing to fix if a difference could be heard.  Conversion is easy and the OP has done this, one improvement made, good decision, good first step.  But I would be left with a significant nagging thought that maybe I could try something else that may also remove this difference and make the whole system better.  Maybe an asynchronous DAC, maybe more isolation, maybe a different motherboard, different CPU, maybe any number of things.  If such a subtle difference in what the computer is doing to serve the DAC makes a difference, something in the system needs to change (or it is exceptionally close to perfect, which of course it may be).

Craig

[gazjam]

Hi Craig, OP here.
good point re: making first step, but why WOULD you want to question everything else..what can I do to "fix" the rest of my system because there must be a failing somewhere...

why not just enjoy the (real or perceived) improvement for what it is?

audiophilila nervosa is a terrible thing
(I know I used to have it! The only measuring equipment I need is my ear/brain interface)

I live happier coming from the heart than the adrenalin gland.
not everything that matters can be measured and not everything that can be measured matters...

[Matt]

Personally, I do regret that JRiver doesn’t have a true memory playback.
If it was possible to have all I/O and all decoding done before playback starts, we can eliminate a hell of a lot of variables regarding lossless file formats.
IMHO, true memory playback would make JRiver independent sound quality wise of (lossless) file formats.

First, JRiver has true memory playback.  Everything needed for playback is loaded to memory.  I'm not saying this matters, only that our description is accurate and honest.

But you're really talking about where decoding happens.  If you could promise me that by moving decoding to the other side of memory playback that I'd never have to discuss whether lossless compression sounds the same again, I'd do it tomorrow.  Should we make a bet

[struct]

why not just enjoy the (real or perceived) improvement for what it is?

Totally, agree.  It is any easy thing to go to wav, you should do it.  Costs nothing, its a win.

My comment is pure conjecture, really don't know if anything could be done, or more importantly that anything should be done.   I just thought that in a well operating system, that wav/flac "shouldn't" make a difference.  I would see this as a cue for potential future experiments, not meant to be a negative.  (Hope I am feeding your nervosa )

It seems to me that these audiphile questions are very complex because there are a great many variables and non-linear interactions (and this is without including the ears), hence the difference in opinions.

Craig

Re Matt's qn...
I would take a bet that it will make a difference for someone but defintely not for all.  I will not take a bet that it would stop discussion.  You should do it though.  It is legitimate science to prove something wrong, indeed many a good thing is discovered this way.

[Vincent Kars]

First, JRiver has true memory playback.  Everything needed for playback is loaded to memory.  I'm not saying this matters, only that our description is accurate and honest.

But you're really talking about where decoding happens.  If you could promise me that by moving decoding to the other side of memory playback that I'd never have to discuss whether lossless compression sounds the same again, I'd do it tomorrow.  Should we make a bet

No dishonesty at all, but there is room for improvement.

As facts cannot beat perception, it will not help.
Nevertheless, I do think you have a very strong case in the lossless file argument if you eliminate any possible difference like I/O and decoding.

Do DAC ready output today and I gladly lose any bet.

[Skogkatt]

If we start to think that on the fly FLAC decompression makes any difference vs. WAV, then we should consider any type of processing, within and outside
JRiver, as a possible source of sound degradation.

Any feedback JRiver team will receive on possible SW modification to improve reproduction will be subjective and not based on objective observations:
some of us will be happy and others will continue to say there is no difference at all.

If we could agree on the fact that sound degradation is related to jitter induced by SW processing, then the only (partially) objective test would be hooking-up a jitter
analyzer to a few common configuration (PC / digital interface / DAC) and tune SW during play for the lowest jitter.

By doing that you aim to improve not only lossless file reproduction vs. uncompressed but JRiver as a whole. 

[MrHaugen]

I was so sure that this would have no impact before, but that was until I saw a video that explained the problem with music from electrical components. Jitter and inaccuracy in music reproduction in electronic components is a fact, and it's measurable. The analogue clock sinuses and jitter and what not can make a difference. I've worked as a electronics engineer before, and it looked very logical and very real. I'm not telling you that what I saw and heard was the truth. I will not be held accountable for the "thruthness" of this video. But a decade long skeptic was convinced.

The question is if this is measurable by the human ear. If I remember correctly, the examples used in this video showed pretty significant levels of jitter. So I would not rule out that it's possible to hear.
I certainly does not have such ears, so I do not really care... But I will never again say that such talk is nonsense. Personally, I just think it's to take it a bit to far, worrying about such things.

[Vincent Kars]

The link below I do think a nice example: an engineer developing a USB DAC.
As bits are bits, this is a straightforward task.
Have a good look at part 2 where he sees the bus rate in the analog out of the DAC!

http://eetimes.com/design/audio-design/4009467/The-D-A-diaries-A-personal-memoir-of-engineering-heartache-and-triumph

[InflatableMouse]

TL;DR, I think if developers need to make a change based on user input, like the way MC is playing from memory as discussed here, they shouldn't make these changes based on unsubstantiated claims.

Unless you're a sound engineer the only discussion I see you can really have in this respect is whether or not things can be heard and how that is established, anything else is just words. You can't control what someone claims to hear but you can control how that is established. The problem with that is that many people with "the golden ears" that claim to hear things won't submit themselves to a scientific approach in a controlled environment to prove whether they hear a difference or not. Their approach to establish whether something is heard or not often aids in their claims. That's not to say they do this on purpose it's just that they are unaware of how they should be testing and more importantly why they should take a certain approach. I didn't understand until I was confronted with the tricks my own mind played on me while taking listening tests.

A "golden ear" is not some magical ear that can hear things that other people don't. The only difference is that it is trained to spot things. And even that is incorrect because it's actually the brain that is trained and not the ear itself. It's like how some people have an eye for detail; they spot things in a picture that other people don't notice, even if they look for it. They don't have better eyes their brain is trained to spot these things. Once you realize that, it becomes easier to understand that the mind can play tricks on what you hear as well. How often do you see things in the corner of your eye that are not there when you look at it? Optical illusions make static images move (search the internet there are plenty to be found). The brain does the same thing with sound and unless you take a scientific approach there is no certain way to tell whether there's a difference or not.

[gazjam]

Agree with this.

As Jim said earlier, JRiver has to be held to a higher standard.
As always though, there's so many phenomena we DONT know about yet..

[crisnee]

I just want to add one comment to this discussion that almost everyone overlooks.

Everything is almost never, if ever, the same, re audio tests or anything else really, no matter how meticulous the set-up. Why? Time. You want to compare A with B, you have to listen to them sequentially. During that sequence, no matter how short, everything in the world has changed. And besides that, you have changed in a very definite way toward your test material. You have heard A before the other (B) and therefore reacted to it in a specific way, which is in a different way then if you had heard B first. There's no escape from the time thing that I can see.

Just wanted to be clear that it is probably impossible to set up a truly objective test and when we're talking about minuscule differences that may well matter.

[Skogkatt]

I just want to add one comment to this discussion that almost everyone overlooks.

Everything is almost never, if ever, the same, re audio tests or anything else really, no matter how meticulous the set-up. Why? Time.

True, time! Listening tests remain a subjective exercise, quite far from an objective evidence.

However, do not underestimate the capabilities of ear + brain in front of the reproduction system you use to listen every day
while reproducing music you know very well: if something is different or just not up to the standard you'll detect it, sooner or later.
Certainly not all types of music are suitable to make audible tiny differences. Also the quality of the reproduction system is a determining
factor together with the fact that some listeners aren't used, interested, or exercised enough to identify some specific details.

As you can see there are several threads started from people claiming to hear differences between FLAC and WAV: are all
those people visionary? Probably they are just genuinely willing to exploit the maximum potential from their systems as some of
them are also ready to go through the hassle of converting their entire music library to WAV with all the problems it creates (more than double
file size just to name one). Most of them did not started a thread asking if they were right or wrong about their sensations, they just asked for
directions or suggestions on how to perform correctly the library conversion.

We, the users, can only be capable of subjective opinions. It would be nice to say: "my JRSoundMark is 7658"…