Amarra sound quality improvements, a practical view

[AudioExplorations]

I read many accounts of the software's SQ improvement over things like iTunes/foobar with kernel streaming (both supposedly bit perfect).

 

IF the software does infact give an improvement it can only be due to one of the following scenarios:

 

1. The other solutions are not bit perfect and Amarra IS

2. Amarra is applying some kind of DSP/equalisation that people perceive as a SQ improvement

 

if neither of these are true, then:

 

3. There is no SQ improvement and people wouldnt be able to ABX the difference

 

Any thoughts on this? Is sonicstudio getting away with charging people $695 for FLAC support in iTunes and easy switching between 16bit/24bit file formats?

 

[DanRubin]

Your reasoning seems practical and sound, and you may be correct. But my experience is that all of the players (Amarra, AyreWave, Pure Music, Audirvana, iTunes) sound different, yet I believe they are all bit perfect. If true, it may mean that bit perfect is only part of the sonic story. Not sure what the other parts are.

 

[ondesx]

Well, it seems several other factors affect the sound some of them are :

- the protocole type (USB vs. FireWire)

- the locking and slaving of the audio device (i. e. the DAC)

- the processor, the OS kernel and drivers, 64-bit will give improved results over the 32-bit versions.

 

FireWire is a better candidate since it is a Peer to Peer protocol with less CPU accesses. Locking and slaving the audio device will also reduce jitter obviously...

 

The very interesting result with several of these soft players (free or not, like Amarra, Pure Music, AyreWave and Audivarna) is that they all are better in 44.1 kHz/16 bits than any oversampled version of the tracks with say iTunes or another "old" player ! Even when they play lossless (FLAC or Apple) tracks...

 

[elcorso]

The other parts are, matching their software with the huge choices of interfaces (USB, Fire Wire, etc.) and DAC's in the market.

 

That's why they ask you (when you register the software) the Mac, RAM, OS, interfaces and/or DAC you have, to see wich is the most popular setup and try to match his software to this one, but without forgetting the other ones: A very hard task!

 

I'm sure that the mentioned software improve the SQ, and then SNR, by kernel streaming, then the "kernel panic" you get sometimes mostly in Amarra. And, when you switch from a music server app to another one, what I do for comparison: I have to shutdown and restart again my Mac in order to flush the cache, to avoid an even worst "kernel panic".

 

I also feel that the increased "better to me" sound in Amarra is for the increased SNR (perceived as loud output) they implemented by a better SQ improvement.

 

BUT, I'm also very happy with Pure Music last version (1.65a). I found a best match to iTunes, playing together, because I do love iTunes interphase, and the only way I can get a "better sound to me" in Amarra is playing it trough the finder, not iTunes.

 

[AudioExplorations]

Thanks all for the comments.

 

Trying to make sense of the various accounts indicating the many hardware related factors that seem to affect the sound quality, it seems that these have an effect on the jitter present in the signal.

 

In theory, if the 1's and 0's being outputted to the DAC are bit-perfect, the only other factor that could affect the SQ is the amount of jitter present in the signal (i.e. the timing errors of the bits that are being sent 44,100 times per second).

 

I can understand how different hardware configs may affect the jitter levels, but am still puzzled as to how different software solutions could have significant (audible) differences in jitter levels given they are running on the same hardware platform.

 

[DanRubin]

>In theory, if the 1's and 0's being outputted to the DAC are bit-perfect, the only other factor that could affect the SQ is the amount of jitter present in the signal

 

Again, I think our collective experience suggests that there is more to it than that. Or we are all deluding ourselves.

 

[Moses]

There is an old CA podcast where Chris asked this very question of Jon Reichbach of Sonic Studios, and an interesting discussion of Reichbach's answer, at:

http://www.computeraudiophile.com/content/CA-Podcast-Episode-2-Jonathan-Reichbach-Interview-Amarra

John

 

[sq225917]

All Johns comments in that p-cast refer to mastering, they have no relevance to straight music file playback where the source file is not being modified/processed.

 

Amarra and others either improve jitter, the eye pattern on spdif, reduce voltage noise on a usb cable, or they make no difference, there aren't any other options if they are as they claim bit-perfect.

 

I'm happy that they appear to sound different to say basic iTunes, but I'd love to see jitter spectra for all of them.