Amarra - Page 2

Further, If Amarra is perceived to sound differently from bit perfect configured iTunes through the same DAC, amp, and speakers, there are only two possible reasons. Either Amarra is not providing bit perfect playback in the configuration tested, or the perceived difference in sound is illusory.
Edited by Jaywalk3r - 11/30/12 at 10:47am

 

Have a read of this: http://www.amr-audio.co.uk/large_image/MAC%20OSX%20audio%20players%20&%20Integer%20Mode.pdf

 

It may go some way to explaining what might be happening. 

Link seems to be broken.

That author makes his comparisons with iTunes in a non-bit perfect configuration. The paper fails to explain how third-party software can do anything to improve on bit perfect playback already provided by iTunes. Even memory playback is a dubious "solution." It might work well for the first few songs of a playlist, but at some point, new songs have to be loaded into RAM. It also ignores that the music player is rarely the only piece of software using the hard drive, particularly if the computer is used as a music server, so even the benefit for the first few songs is questionable.
Edited by Jaywalk3r - 11/30/12 at 9:09pm

Link fixed (there was a space at the end).

 

The RAM-caching software, such as Amarra and Audirvana Plus pre-load the tracks into memory.  I inadvertently discovered one important thing inadvertently when I was playing around with measuring gear, and that is to ensure that the device selected for audio output is not the same as that selected for alert sounds.

 

I would consider exploring the technology used in pro audio for more information as software performance is critical for it.

Right. And what happens when playback of the songs preloaded into RAM is complete, but there are still (or new) songs in the playlist? What happens when another network user decides to access files in the library of the music server? What happens when the OS does work in the background?

New songs in the playlist are loaded into RAM before playback starts, that is: song is played, new song is loaded, new song is played. Amarra can be set to load more tracks ahead, which is useful for gapless albums. A network user accessing the files will have no effect on files loaded into RAM. As for the OS doing work in the background, that is why some people use computers dedicated for music playback and other software, such as Audirvana, has the option to shut off Spotlight and Time Machine when running.. You'd have to ask Sonic Studios or a programmer familiar with pro-audio software about that otherwise.

Right. But it does result in disk accessing. The whole point of preloading into RAM is to avoid disk use because of the associated electrical noise. But it doesn't prevent it. It just doesn't contribute, except for loading the file into RAM one time. Which is what iTunes does. It loads the file into RAM one time. Simply shifting the time at which the file is loaded into RAM doesn't accomplish much, since preloading still creates about the same amount of noise. After the first preload, the benefit is nearly zero, particularly with gapless playback. It sure is a nice marketing feature, though. :-)
Edited by Jaywalk3r - 12/1/12 at 12:06am

No, iTunes reads constantly from disk during playback. Amarra or others read from the disk once before playback and don't read from disk after that. I think the point is to cut down on the number of I/O processes and parts of the system involved during actual playback to minimise jitter upon output.

I opted for all SSD

No noise.

500Mb/sec read.

Gapless is no longer a problem.

Each reads the file into RAM exactly once, so the benefit of pre-loading is minimal, at best. There are certainly situations in which one would not want to read the file all at once if one wants to minimize noise caused by I/O processes. Plus, a single large read will create a larger noise (local) peak than a smaller read creates, and the larger peak is more likely to be audible. (To be fair, it's extremely unlikely either method is audible during playback, provided the machine has sufficient RAM and drive space available.)

Besides, the operating system doesn't just stop working in the background simply because Amarra is playing.

Regarding jitter, iTunes can be configured for bit perfect playback. As long as Amarra uses the same DAC, it will have approximately the same level of jitter (not that it's audible, anyway).


SSD gets rid of mechanical noise, which is already so low with modern HDD that it isn't a big deal (again, provided the machine has sufficient RAM and drive space available). An SSD still causes electrical noise. During gapless playback, additional files are read into RAM with Amarra (or any other player), so the benefit of pre-loading is gone.
Edited by Jaywalk3r - 12/1/12 at 11:04am

I experienced significantly better performance running Amarra on my SSD MBP retina than my M Mini with a HD.  IMO sonics were better but I think that has more to do with data access / performance than any perceived hardware noise.

I have two systems  room and headphone. The mini(room) is connected to a Cambridge Audio DacMagic Plus via USB2. I don't suspect any hardware related noise, if there was any, could be passed to the DAC over USB.  It's bits not sonics.  The MBP is connected to a Schiit Bifrost using TosLink. I think it unfathomable that any noise, if there is any, could be passed to the DAC over FO. 

I perceived an improvement on the Mac Mini when I upgraded the HD to SSD.  Again IMO, this had more to do with performance.  As an example,  gapless issues went away. It's  not a stretch to think it helped other data related issues such as interface / clocking where bits might be dropped.  

I did a side by side of the 2 DACS using the MBP running Amarra with output to a Schiit Lyr with 65' Amprex tubes to Senn HD650.

The DM+ used USB2.  It was clocking the MBP at 192kHZ.  The Bifrost was clocked by the MBP over TosLink at 96kHz.

I preferred the Bifrost.  IMO this had more to do with the DAC than the interconnect. It did debunk the whole 192kHz must be better that 96kHz.  To me.

My logic suggests there could be issues when other processing tasks running might have an impact on the processors ability to provide a consistent clock.  If the clock slips it  would effect sonics.  SSD provides a performance benefit. I have no way of telling if there ever were any clocking issues.

All this is very subjective.  In the end, I like my kit and how it sounds.

 

Edited by Iamnothim - 12/1/12 at 12:29pm

Remember, the MBP Retina is Apple's (newest) high end portable. The Mac mini is their entry level desktop (running laptop components). Oftentimes a company's higher end offerings perform better than their entry level stuff.

That's exactly why iTunes works perfectly well when bit perfect playback is the goal.

Not surprising at all. Of course there's no reason to assume the perceived improvement should be attributed to anything other than expectation bias.

That's a huge stretch, actually, with no justification for such a conclusion. Like you said yourself, "I have no way of telling if there ever were any clocking issues."

I'm open minded and like trying different things.

 

Help me with this... sincerely.

 

I googled "bit perfect" and all I got was http://bitperfectsound.com which is another 3rd part player, like Amarra, PureMusic, etc.

Is there a link / description where the phrase " bit perfect" is used with iTunes?

 

I believe you mentioned setting up iTunes for bit perfect.  Where do I find that and what are the parameters?

I'm running iTunes 11

 

I enjoy HD Tracks high clock 24 bit recordings.

I also enjoy remastered CD's   Sgt. Pepper, Rubber Soul, etc.

I don't think I can tell the difference between the two.

 

 To me it's about the source.

 

The stuff on "mastered for iTunes" sounds very good.

I don't think I can tell the difference between those selections and HD Tracks.

Still if a recording is available on HD Tracks I will buy it there rather than iTunes.

If there wasn't a difference iTunes wouldn't have "mastered for iTunes". 

I think the sound is noticeable over the vanilla versions.  Ditto HD Tracks.

 

Regardless of my preferences,  some pointers on iTunes optimization would be appreciated.

What are your preferred setting for ripping a CD into iTunes.  I'm using AIFF and error checking.  (Ya, big files)

 

I haven't looked, and I can find out for myself,.... But I'm asking.  Can iTunes be configured for an outboard DAC?

Is is just a matter of going into Midi and setting the outboard port?

 

Thanks

It's not really about doing anything special for Itunes to playback bit perfect.

 

Turn off enhancers and EQ. Max volume. Done. 

 

 

Bit-perfect audio/video does not perform any digital signal processing (DSP) such as channel matrixing, filters and equalizing and does not do any resampling or sample rate conversion (such as upsampling or downsampling). In audio this means that the digital output from the computer sound card is the same as the digital output from the stored audio file. Unaltered passthrough.

The data stream (audio/video) will remain pure and untouched and be fed directly without altering it.

Bit-perfect audio is often desired by audiophiles.

In case of error correction it might apply error concealment through interpolation which results in the result not being bit-perfect.

Poor device drivers often alter the data, resulting in it making not bit-perfect. This is especially true for device drivers used in consumer-grade sound cards.

Bit-perfect is good when sending the data output digitally (such as via S/PDIF) to a external digital-to-analog converter (DAC).

 

 

Link          

 

 

As for your HD Tracks VS "Itunes master, it doesn't really matter if it's crap from the beginning.

 

Loudness war database 

 

You should grab XLD for ripping your CD's

 

 

Yes, use the Midi settings.

Edited by paradoxper - 12/1/12 at 2:05pm