PC For Music listening

[alpha1]

Hydra, the difference, if audible, might be because of soundcard difference between the two PCs.

It's strange that there can be a sound difference between two software players stripped off any processing / DSP!
And lets see - if there is a difference, then why is there a difference?

Lets keep the sampling rate out, because it will not matter unless you play a source that is at 96kHz or 192 kHz! (this means the studio recording, mastering, and the final audio file all are at high sampling rate, and being played through the PC)

 

[hydra]

alpha1, there was no soundcard involved. Both machines were connected to the DAC (Beresford Caiman) via a generic USB cable.

Believe me, I'm stumped too! The only difference in setups here was in the computer hardware, operating system, and the media player software.

 

[jls001]

For the price Amarra costs it had perform better. Perhaps you can do a test comparing J River + JPLAY with Amarra and post the results.

Then there is an alternative in the Spatial Computer which promises better SQ with existing h/w.

+1 to the Amarra It's costly, but apparently worth it if one uses a Mac platform for music playback.

Spatial Computer also needs a Mac platform. It's much costlier than the Amarra and installing it involves remote tuning by the vendor. I heard Sridhar's Macbook running the Spatial Computer at ARNS, and to me it sounded better than the Ayon CD2s CDP. The Emerald Physics CS3 sounded much better when driven by the Spatial Computer.

@spiro: very interesting observation about the Lilith. I have not seriously tried it though I downloaded it after you posted about it. I will do a non-blind A/B with foobar tonight. My only grouse is the teeny-weeny UI.

Aside: My music PC which has been unused for some months now due to a stutter in playback has finally recovered. When playing any audio track on foobar, or a video on Win Media Player, there a periodic stutter which was very annoying. I didn't know what to do so I kept it like that till the other day. Plus incoming new LPs didn't exactly help matters. I thought it could be some hardware issue so I opened the PC and gave it a general cleaning (including RAM contacts). While putting it back, I blew the SMPS.

So in came a new SMPS with a bonus new OS (Win 7, finally). Now it runs WASAPI on foobar.

I did a quick comparo of the PC with my TD124. Tracks used were Steely Dan's Aja album. The digital files were ripped from my CD of the Euro pressing using EAC. LP is Mobile Fidelity. I didn't use the Euro pressing LP as it has irritating groove noise. I normally don't do these type of comparison but I wanted to see how the "new" PC sounded and also how the new cartridge (Shure V15 Type III with Ed Saunders stylus) stacks up against "competition". This was done when the new cartridge had just been put in and had played may be 2-3 sides.

Levelling for roughly equal SPL (by ear), the PC sounded very clean (obviously) and dug out more details while sounding clinically cold compared to the LP. The LP has its characteristic warmth and loses out a bit on the details department but had better bass foundation for the same volume level. I will do this again when my cartridge is more properly burnt in. Right now the cartridge is sounding absolutely horrible (may be 20 hours total now). I guess it is at the bottom of the burn-in trough.

Also, need to re-tune the foobar for low latency. Bhagwan advised me to get it below 50 msec. Will have to see if my sound card can do that. Plus need to disable some unused services in BIOS to improve system latency.

 

[alpha1]

^^^ Why would you want low latency from an audio player??

Low latency is required by live sound recording editing and production - due to the nature of job. Otherwise latency doesn't play any role in sound quality.

 

[jls001]

^^^ Why would you want low latency from an audio player??

Low latency is required by live sound recording editing and production - due to the nature of job. Otherwise latency doesn't play any role in sound quality.

Please read the section Background information: Why drop-outs occur : DPC Latency Checker

High latency can lead to audio and video dropouts. The current latency is adequate for music playback, but I am still trying to tweak it to get the best that it can give. Mine is a dedicated "music only" PC.

 

[jls001]

With respect to studio computers, here's a good article from Sound on Sound:

Optimising The Latency Of Your PC Audio Interface

These guys are talking about 5-6 msec!

 

[Thad E Ginathom]

Sound on Sound is an absolutely brilliant magazine. I used to buy the print copy in London, and I still check it for reviews of any equipment I'm interested in. However, music recording has different requirements to music listening.

Why would you want low latency from an audio player??

Low latency is required by live sound recording editing and production - due to the nature of job. Otherwise latency doesn't play any role in sound quality.

Absolutely correct. My analogy is pass the parcel (a party game). It doesn't matter how many people handle the parcel, or if it gets delayed, it is still the same contents that arrives at the end of the line.

Thus, latency has no effect at all on the sound quality. Whether there is a gap of milliseconds or minutes will make no odds whatsoever ...Except in purely practical terms: when we press Play or Stop, we like it to happen without a noticeable delay.

Please read the section Background information: Why drop-outs occur : DPC Latency Checker

High latency can lead to audio and video dropouts. The current latency is adequate for music playback...

Any latency is adequate for playback.

High latency cannot lead to dropouts, unless your machine is working so badly that you are continually overflowing or failing to fill buffers. That would not, primarily, be a latency problem, it would be a lousy-PC problem! Most PCs, and most OS setups are not at all lousy: they are perfectly able to play a song, or several, with no hiccoughs.

The word latency means different things in different contexts. DPC Latency is something else entirely. It is one of the things that can turn your PC into a truly lousy PC for playing sound. It can even result from an unfixable combination of hardware. It is a very nasty thing indeed: I've had it.

but I am still trying to tweak it to get the best that it can give. Mine is a dedicated "music only" PC.

You may gain some technical satisfaction, and learn some stuff along the way, but you will not gain in music quality by shaving a few ms off your latency. Just, the same music will arrive at your speakers a few ms earlier. This is a difference that cannot possibly be perceived, except for the feel of the controls as mentioned earlier, because the relative timing of the music is still correct.

Consider one of the current "audiophile" PC fads: read the entire song into memory before playing. Ultimate latency!

If you do have DPC-Latency problems, then I really sympathise

 

[alpha1]

Please read the section Background information: Why drop-outs occur : DPC Latency Checker

High latency can lead to audio and video dropouts. The current latency is adequate for music playback, but I am still trying to tweak it to get the best that it can give. Mine is a dedicated "music only" PC.

With respect to studio computers, here's a good article from Sound on Sound:

Optimising The Latency Of Your PC Audio Interface

These guys are talking about 5-6 msec!

As I said earlier - latency is of importance only to a person making music.
It doesn't change anything for a person listening to music.


The second site is outright about music production.
The first site also talks about real time streaming - again this is of importance only when you set-up pipeline in your workflow:

Instrument -> mic -> (pre-amp) -> ADC -> DAW -> DAC -> feedback monitors.

In this case you wish to listen to the instrument in real time.


But for us, mere mortals, latency does not matter, because we are not playing instruments that we expect it to sound instantaneously. The CD player - if it plays after 5 seconds, this 5 second delay will be present throughout and it doesn't affect us in any way.

In fact, latency is an inherent property of your hardware. If you try to "set" latency less than what your hardware can take - then we can definitely expect problems.

 

[jls001]

Any latency is adequate for playback.

You may gain some technical satisfaction, and learn some stuff along the way, but you will not gain in music quality by shaving a few ms off your latency. Just, the same music will arrive at your speakers a few ms earlier. This is a difference that cannot possibly be perceived, except for the feel of the controls as mentioned earlier, because the relative timing of the music is still correct.

Consider one of the current "audiophile" PC fads: read the entire song into memory before playing. Ultimate latency!

If you do have DPC-Latency problems, then I really sympathise

There are two types of latencies - one where packets arrive variably, and second where they arrive with fixed delay. Buffering entire song and playing it comes in second variety.

What I am trying to attain is to keep the variability small so that it does not affect playback. I did some reading and have already learned basic tweaks to achieve better PC latency. I will try them out. Of course whether 150 msec versus 50 msec produces better music quality remains to be seen. I really don't expect to gain in audio quality but if it does come, it is welcome. I am only trying to avoid the glitches I had faced.

 

[alpha1]

^ I just Hope you are not buying anything extra to ensure that.
Please keep in mind that a huge chunk of "audiophile" industry runs on placebo, peer pressure, ego massage and FUD (and thus objectivity is looked down upon)

Surprisingly, none of these things have anything remotely to do with sound/music. (audio, audire = hear; philia = love/affection)

 

[jls001]

@spiro: very interesting observation about the Lilith. I have not seriously tried it though I downloaded it after you posted about it. I will do a non-blind A/B with foobar tonight. My only grouse is the teeny-weeny UI.

I listened to same tracks on foobar and Lilith using an IEM on my laptop. To my ears Lilith does have better sounding mids. I could also hear better defined bass. I will try tonight on my music PC.

 

[jls001]

^ I just Hope you are not buying anything extra to ensure that.
Please keep in mind that a huge chunk of "audiophile" industry runs on placebo, peer pressure, ego massage and FUD (and thus objectivity is looked down upon)

Surprisingly, none of these things have anything remotely to do with sound/music. (audio, audire = hear; philia = love/affection)

No pressure from any quarters. I had issues of audio dropouts in my PC which deprived me of a huge chunk of my music collection for some considerable time, and that is why I don't want a repeat. No bruised ego to heal or an oversized one needing a massage, either

But I might buy something

 

[jls001]

As I said earlier - latency is of importance only to a person making music.
It doesn't change anything for a person listening to music.

This is not true.

For example, if one were listening to music from a PC that is unnecessarily burdened with so many non-music processes that it can't play the music coherently, it is a problem. And it is perfectly possible to arrive to such a pass. In my specific case, I need to be more careful since my PC is a very old one salvaged from a colleague's home. He was about to gift it to the friendly neighbourhood raddiwallah. I brought it home, gave it new RAM, new keyboard and new mouse. And of course a studio-grade RME sound card. That's it. And as of two days ago, it has new OS.

Perhaps I am blowing things slightly out of proportion. I wish to put on record that my PC sounds fine playing foobar/WASAPI. I am only trying to tweak it for better performance so that I don't see recurrence of audio stutter. I am not trying to be anal retentive and be a pain to myself.

 

[Thad E Ginathom]

There are two types of latencies - one where packets arrive variably, and second where they arrive with fixed delay. Buffering entire song and playing it comes in second variety.

Latency, as in the time delay caused by processing the sound from signal-in or hdd out to speakers or DAC should be fixed. If it is not, then we are outside the field of normal, and into the field of faulty, which, obviously, needs to be fixed.

If you have glitches, then yes of course, you have a problem that needs to be fixed. I think that somebody mentioned already: reducing latency might make the problem worse rather than better. You are giving your PC a tougher job to do. PCs do not have a real-time architecture, but the more you incline latency towards zero, the more you are trying to get real-time processing from the machine.

Would be interesting to know more about your problems, how you are determining latency, and what steps you are taking. Perhaps might be offtopic to this thread. Elsewhere?

 

[alpha1]

JLS: what you say happens only on my 10 years old P3 machine with 128 MB ram.
When I load too many applications, the PC hangs, and the sound also hangs.
I get jarring sounds of "bit by bit" sound reproduction.

But then ... its because of CPU and RAM (and thus decoded signal from Harddisc data not reaching the soundcard), not the soundcard latency. The sound card never gets loaded as such.

 

[Thad E Ginathom]

Perhaps I am blowing things slightly out of proportion. I wish to put on record that my PC sounds fine playing foobar/WASAPI. I am only trying to tweak it for better performance so that I don't see recurrence of audio stutter. I am not trying to be anal retentive and be a pain to myself.

Absolutely never thought you were

Whereas PCs, even old ones, should be able to play audio just fine, as it isn't a very demanding application, there are still things that can go wrong.

Added to which, tweaking PCs is a pastime in itself, and why not. What we can do is share and better understand the principles

 

[Decibel]

The Asus Xonar STX with ASIO support can go down to a latency of 5ms, well below the acceptable limit of 11-12ms.

Cheers

 

[jls001]

Added to which, tweaking PCs is a pastime in itself, and why not. What we can do is share and better understand the principles

Or at least mess it up (as is my wont:lol::lol::lol

 

[Thad E Ginathom]

:lol:

The Asus Xonar STX with ASIO support can go down to a latency of 5ms, well below the acceptable limit of 11-12ms.

When latency is mentioned for soundcards/interfaces, it is usually the time it takes to feed the input to the output. This is what is critical if you are, for instance, listening to a track whilst recording another track to mix with it, or monitoring yourself whilst playing or singing.

again: for playing, how can there be an "acceptable limit?" In fact, given the above, what are people even measuring? And how?

 

[jls001]

I did the tweaks which I'd threatened to do. Average latency before tweaks was around 145 msecs. More or less same figure post tweaking:lol: So I will not waste more time on this.

But I anyway slid the foobar latency from about 240 msec to 50 msec just to see if it can handle it, and it does without any problem. May be I'll get more adventurous and bring this down to 30 msec for no other reason than to see if the PC can handle it.


SoundPlayer Lilith v/s foobar2000:

Doing an A/B on my main rig closely reflected what I already heard on my laptop-headphone combo. Lilith has discernibly better bass weight. The mids are on the warmer side of neutral. However it loses out on resolution when pitted against foobar, and therefore sounded a bit fuzzier.

Later I will also do a Lilith - foobar - TT comparo to understand if it Lilith more closely resembles analog sound.

Learning from this exercise: I realised yesterday that the headphone output of my beat up laptop (Dell XPS M1330) sounds quite good. I think I heard more details than from my music rig. IEM used is my 3-year old Sennheiser MM50.