Amarra

[ted_b]

How do you explain that, in that the type of "everything" makes a difference, the amplifier somewhere near the top of that list IMHO. I can't imagine that this product is so revolutionary (and physics bending) that impedances, power output, tube vs ss sonic character do not make a difference.

 

[tmfidelis]

Dear Eloise:

 

Regarding your question with playback electronics, a playback system's ability to resolve sonic differences using OSx and Amarra is not limited to solid state playback. One of the founders of the Tape Project uses exclusively all tube electronics in his playback system at home with Amarra.

 

If you check out http://www.bottlehead.com/, you may see another sector of the industry working with the Amarra software I/O.

 

Regards,

 

 

 

[Purite Audio]

The Itunes and Amarra demonstration was played through the same system , with the same amp, which happened to be valve, are you saying that the choice of valve amp somehow favoured the Amarra software?

 

[ted_b]

because I don't understand it...it's not related to my comments. I was simply commenting on your statement that (paraphrased) said "choice of amp type makes no difference". I am debating that statement. Seems amp type would make a difference in any signal path.

 

[lamarzocco]

Definitely turn off spotlight and dashboard via terminal commands. Spotlight will index everything, if you add a new drive or copy some files, it starts indexing eating CPU as well as creating more disk activity while listening. Same with Dashboard, all those widgets consume RAM even when you're not looking at Dashboard. Looking at Activity Monitor, you can watch them consume ram and occassionally as a widget updates the weather or stock info as its syncs, again consumes some CPU cycles.

 

IMO, to get the best performance out of a dedicated Mac Audio playback machine,

 

1. Reinstall Leopard, during the install you can customize what gets loaded, uncheck the printer drivers, the language packs etc.

2. Opt not to install any optional software packages, eg. the entire iLife package.

3. You can even save 600mb by deleting "alex" the voice.

4. Disable Spotlight and Dashboard via terminal

5. Disable screensavers and update energy preferences to not let your drive sleep.

6. If using a mac mini, remove the Airport Wifi / Bluetooth daughtercard to eliminate extra RF inside the mini. If you need bluetooth you can buy an external and put on on a usb extension cable.

7. Use an SSD disk, Intel X25M MLC or X25-E SLC, prices will get better once 32nm NAND Flash production ramps.

8. Max out on RAM, unfortunately the new mini will not take 2x4GB DDR3 modules, yet.. may be possible with an EFI update.

9. If you want to hog-wild, start deleting built-in apps. mail, ical, etc.

 

Gosh there were a few more things, but it's late and I'm on my work PC.

 

Glenn

 

 

[souptin]

Thanks for the reply - these are all good tips if you want to get the best out of your mac. In fact I think maxing out the performace is worthy of a topic all by itself. Of course it's also relevant to this topic, because Sonic Studio specifically recommend doing this sort of thing to get the best out of their products.

 

Some of the issues I (and I expect many others) would be interested in are peoples thoughts and experience on what gives the best direct effect on the sound quality. Since home computers are often multi purpose and shared between family members, there's always some compromise needed when you start switching bits off and uninstalling software.

 

More questions for anyone kind enough to answer:

1: Any preference on whether SSD disks are better dedicated to software & os, or to storage of the music files themselves?

2: Remove the Airport card completely or just disable Airport in software? (It's a tricky bit of surgery to open up a mini)

 

[PeterSt]

9. If you want to hog-wild, start deleting built-in apps. mail, ical, etc.

 

Gosh there were a few more things, but it's late and I'm on my work PC.

 

Remove iTunes perhaps ?

hehe

 

 

[lamarzocco]

Souptin

 

I agree, there's lots of potential to create a bare-bones OS X install that's optimized for audio playback and playback only. You could always have a second OS to boot from that has the whole OS and Applications that could boot into when needed and use bootcamp so you can try Peter's XXhighend player .

 

I'd use the more expensive SLC or non-Jmicron controller MLC drives as a boot / OS disk and storage elsewhere. I haven't tried this, but I'd suspect the cheaper MLC SSD's using the J-Micron controller may not be too bad for storage, as they are optimized for linear STR rather than randomized disk access. However with Amarra only working with WAV and AIFF, that's going to get pretty expensive to store any decent sized music collection on SSD.

 

glenn

 

[Opusover21]

Glenn what difference do any of those tip's make to the sound??

 

[lamarzocco]

Honestly, it's more of a what can we do to not degrade the sound. There's a lot of "potential" applications and services that could blip the audio processing, so those are some tips to help reduce some of the potential problems. People have mentioned that they can hear a difference between music stored on SSD vs. HDD but I have not done that test yet.

 

A couple of benefits of streamlining the install, saves disk space (if you're using SSD, capacity is at a premium) and faster boot times. Oh yeah, I forgot to add to

 

10. Turn off automatic updates.

 

I think we all know that computer based audio is the future, but the last thing I want when I sit down and listen is a bunch of computer problems or glitches when I'm running headless. It needs to be an appliance, power-on, play music, power off. I'll tweak and update settings or iTunes versions when *I* want or need to, but stability and reliability are important for me.

 

Glenn

 

[Opusover21]

Hi Glenn - I'm not trying to make hard work of this but I use my Mac Mini just for audio and everything is stored in RAID1 via a USB WD drive. I dont have any real issues "maintenance" issues with the mac so to speak...

 

Ok so do these tweeks impact the sound in your experience or is it performance related issues? I'm just looking for a simple yes or no based on your experience....thanks

 

[lamarzocco]

Opus,

 

So far in my limited testing it's been performance related. Once I get 192/24 capable DAC I'd be able to make more of statement to sound quality. In a few weeks i should have some nice DACs inhouse to play with.

 

Glenn

 

 

 

[The Computer Audiophile]

Hi Opus - In my experience it's to difficult to tell if certain items effect the sound. A response like "Turn off spotlight and get better highs" just isn't going to happen in my opinion from anyone I know. That said, I think the totality of performance adjustments likely prevents sonic degradation. I have been working with a beta test site for over one year on some of this stuff. The test site reports that on a very resolving system one can hear a difference between SSD and HDD. Take that for what you will, I'm just sharing some of my experiences :~)

 

[lamarzocco]

Chris,

 

In those tests are the disks are attached to the playback computer? Is there a difference even if the drives are networked on another machine? I guess this begs the question is there a difference streaming from a network drive versus localized storage?

 

Glenn

 

 

 

[Bezerkely]

Hi All,

 

A few pages back I suggested Quicktime/cross platform issues were the reason for iTunes poor performance. My listening test won't prove or disprove this hypothesis, but I'd like to post my comparison of four OS X applications for music playback: iTunes 8.1.1 (or whatever the latest version number), Songbird (ditto), Play by sbooth, and iTunes w/ Amarra in demo mode. Of course, all EQ, soundcheck, gain, etc. were off and volume of all players was at maximum.

 

First, a little about my system: white dual-core Intel iMac, 2Gb ram, 5Ghz wireless NAS, Audio/Midi set to 24/96->Belkin Gold USB->Empirical Audio Offramp3->Kimber Cable silver digital XLR->Tact 2150 (Maui Mods+)->Zu Ibis->Zu Druid MkIV (2008). iPhone as remote if using iTunes w/ or w/o Amarra. This is dedicated to music, and one day I will get around to stripping down the OS, add an SSD boot drive, etc. Maybe when Snow Leopard arrives.

 

Source material for comparison (all files AIFF): Fleetwood Mac, Rumors (DVD-A version, 24/96), Fleet Foxes, Fleet Foxes (16/44.1), Lyle Lovett, Joshua Judges Ruth (16/44.1).

 

First up: iTunes vs. Songbird. I started all tests with Rumors, specifically "The Chain", wanting to test out 24/96 material first. Well, couldn't really tell the difference between these two and gave up on Songbird. It's sluggish response time and all the extra garbage they add to make it the uber-iTunes is overkill for me. Who needs all that extra crap? I just want to play music, not look at a band's annoying MySpace page while listening.

 

Next: iTunes vs. Play. Not sure what I expected here. I had tried earlier versions, but wasn't impressed with the basic interface and lack of iTunes-like organization abilities, built-in CD extraction, no remote control, etc. It's support of AudioUnit plugins was interesting. If you need high quality EQ compared to iTunes, then this was probably your best bet. However, since everything else was lacking, I never gave it a serious listen until now. Again, I started with "The Chain" and while I could clearly hear differences-mostly in tone and sound staging-I wasn't quite convinced which was better. So, browsing the library I hit the Fleet Foxes and thought how I had always been disappointed at how this album had sounded. The first track starts out a cappella, then you get some guitar chords coming in...well, I was shocked. I wasn't hearing a guitar anymore. I was hearing strings, frets, a resonant box. Went back to iTunes-flat, dull dynamics, everything was less distinct/smeared together. There are hints of what I hear with Play, the difference is I don't have to listen for it with Play like I do with iTunes. Flip to Lyle Lovett, well recorded, but somehow never quite as satisfying as I expect. Once again, Play to the rescue. Dynamics, better. Tone, no contest. Sibilants, gone. Piano is a hammered instrument. I'm suddenly thirsty for ice water. These things are obvious in live music, we've just forgotten since we listen to music in cars, over crappy iPod headphones or over cheap computer speakers (I have Klipsch Heresys in my office w/ NuForce Icon). I have to say I haven't gone back to listening to vanilla iTunes since. Recordings I've listen to a hundred times seem fresh again and I can't stop listening. Okay, I did go back briefly, just to make sure I wasn't fooling myself, but it was immediately obvious something way wrong. It's almost like after starting your car up for five years, you finally take it in for a tune up and when you drive it that first time you suddenly realize the sound/performance is changed in a subtly but very positive way. Your car pool buddy may miss the difference, but to you it's like night and day. Interface hasn't changed, and no remote. But the playback is so nice, I'm starting to thing more in a vinyl mode-set up the playlist and sit your butt down until it's over-rather than my hyper-I can access any track at any moment mode-with iTunes.

 

Last: iTunes w/Amarra demo vs. Play. The main difference here is that Amarra switches to the native sample rate of the source material whereas Play is always outputting 24/96. In my system, the Tact accepts up to 24/192. However, internally, everything is processed as 24/96. Some time ago I compared 16/44.1 vs. 24/96 upsampling on the Mac and determined that I preferred that my iMac do the upsampling (core audio) over the Tact internal upsampling. Fleetwood Mac, not much difference. I don't think I could A/B Amarra vs. Play for this file. I must confess that I find the frequent Amarra dropouts very distracting for the purpose of making a comparison. Amarra vs. vanilla iTunes and the difference is clear. Amarra vs. Play, I can't quite call it. Moving to 16/44.1 material, Amarra loses on my system. The internal upsampling of the Tact defeats Amarra's native sample rate switching. I e-mailed Jon at Sonic about this, and he suggested Amarra wasn't the product for me if I wanted for force 24/96 output of all sample rates.

 

This leaves me a great conundrum. I really like iTunes and being able to remotely control iTunes from the couch. Adding Amarra seems to be a no-brainer for me if only I could fix the sample rate at 24/96. One option would be to upsample my entire library, but this means more $$ on top of Amarra and an extra step in the ripping process with every new CD purchase. For right now, I'm stupendously happy using Play for serious listening and switching over to iTunes when I need remote capabilities for parties, etc.

 

I'd love to hear feedback from others concerning iTunes/Play/and Amarra. Many here scoff at Amarra's price, and I find that Play is very close if not as good as Amarra. Again, the demo mode makes it hard for me to declare Amarra the absolute winner, but I am very willing to give the Sonic guys the benefit of doubt. My modest equipment probably doesn't hold a candle to the gear they listen to day-in and day-out, but I can say that the difference between Amarra and iTunes alone is very clear on my system. I don't really care why it sounds better and I appreciate everyones hard work at making computer audio better.

 

I also want to add I quick endorsement for the Empirical Audio Offramp3. This gizmo took a long time to break in, but once it did, my system became a living, breathing, natural, dynamic monster and I fully give credit to the Offramp as being the heart of my system and why I am able to clearly hear differences on the software end.

 

Thanks for reading this long post and happy listening.

 

 

 

[PeterSt]

If you read back your own review of what you did, you will find that you started to imply all playback is to be "bit perfect", but near the end stated that "of course" everything is upsampled in your system from 44.1 to 96. I hope you can see that this invalidates all, and that the most profound sound you will receive from any setup, is the upsampling. Better/worse it not all that important. Important is it won't allow comparisons.

 

Now, go back to the beginning again, and see that, out of all, your 24/96 Rumors didn't bring a diference. Isn't that a coincidence ...

 

I don't say that the differences you perceived can't exist, but I do say that the differences incurred by the upsampling WILL exist. So now it is just hard to know to what degree your observations are caused by the software. I think (not sure) that the only thing you are allowed to compare here is 24/96 files. And a commercial album like Rumors may not be the best for it ?

 

Peter

(and thanks for all the effort ... the writing alone takes an age)

 

[The Computer Audiophile]

Hi bekerkley - Thank you very much for the detailed comments.

 

I'd like to continue this discussion and get more information about your upsampling choices. I couldn't quite figure out what you were doing or why you were doing it, but I am really curious about your decision to force 24/96 output. I'm not criticising it at all, as that usually ruins all good discussions before they start.

 

Anyway I'd love to hear more details if your interested in continuing the discussion.

 

[Bezerkely]

Hi Peter & Chris,

 

Let me explain a bit more about the Tact 2150 amp (http://www.tactlab.com) and this should make my upsampling choice clear. The Tact is a class "D" digital amplifier. Tact calls itself the only true digital amplifier in that it uses a TI chip that takes a digital signal and turns it into a signal strong enough to directly drive speakers without changing it to analog. Panasonic makes home theater receivers based on this same TI part. In simplest terms, it is a "power DAC". However, this is not quite true since a DAC is creating an analog waveform (that then needs amplification to drive a speaker). It is my understanding that whatever bit depth/sampling rate that is feed to the Tact is transformed to 24/96 internally and that signal is fed to the TI chip. The 24/96 signal is then upsampled to 384Khz and output as PWM (pulse width modulation) signal-essential a burst of voltage that varies in length-that drive your speakers. Somehow this sounds like music coming out your speakers, and is wonderful sounding to my ears.

 

To summarize, if I feed the Tact anything other than 24/96 it will be converted to 24/96 automatically by the Tact. I have no choice there. Where I do have a choice is where the upsampling occurs: on the Mac or in the Tact. Listening to the Tact with a 16/44.1 signal vs. 24/96 upsampled on the Mac (by core audio), I prefer the sound of the Mac doing the upsampling. The two are close, but the Tact fed 16/44.1 sounds a tad more cold/digital/clinical to me.

 

Therefore, in my listening test, upsampling was always in occurring and unavoidable. Only when listening to the Amarra demo was the upsampling taking place on the Tact. In all other instances the upsampling was happening on the Mac through core audio. From my past listening experience, the difference in upsampling between the Tact and Mac are small. The differences in hear in software, iTunes vs. Play, is due to the playback engine since both are being upsampled by core audio. This playback engine difference is far greater and much easier to distinguish than the Tact vs. Mac upsampling. In my Play vs. Amarra listening, I suspect that the Tact upsampling hurt the Amarra playback slightly, leading to my wish to allow Amarra to upsample or not change sample rate and allow core audio to continue to upsample. This would allow me to do an apples to apples comparison of Amarra vs. Play, rather than the apples to oranges that I am doing now where two variables are changing at once (playback engine and who is upsampling).

 

I hope that I have made my choice about upsampling and the rational behind that choice clear.

 

To address some of Peter's questions directly: bit-perfect is kind of a red herring in my current system, it will never be possible unless I buy a separate DAC and amplifier to replace the Tact. I agree that Rumors wasn't the best choice of 24/96 material, but my selection of high rez is extremely limited and I wanted to stick to music I was intimately familiar with. Maybe I'll listen more to the HD Tracks sampler I downloaded and give the 24/96 iTunes/Play/Amarra bake off another spin. I would like to reiterate that the difference in 16/44.1 playback was very clear between iTunes and Play with core audio doing the upsampling for both. Poor choice of 24/96 source material made the difference smaller, but Play was still preferred over iTunes.

 

Chris, I'm in Berkeley and if you have any spare time during the symposium weekend, I'd be happy to have you drop in for a listen.

 

I would encourage all the OS X users out there to do an iTunes vs. Play comparison. If you find that Play is the superior playback engine, then I think you will like Amarra as well and may hear additional improvements. The main benefit of Amarra is that it maintains the functionality of iTunes, but adds superior playback. Even though Amarra isn't ideal on my system, it's very close and keeping the iTunes functionality may, in the end, persuade me to buy.

 

Plus, my wife owes me after all the in-law visits these last two months.

 

 

Thanks again for reading.

 

 

 

[Audio_ELF]

I may be wrong here, but it is my understanding that itunes up/downsamples internally everything to the frequency that is set in CoreAudio at the time iTunes is started. I.e. If you set in Audio Midi Settings to 24/96 then start iTunes, it is iTunes that is upsampling to 24/96 NOT coreaudio.

 

This may explain the difference in the sound "quality" you heard.

 

Not sure if sbooth's play works similarly or if it just uses coreaudio to upsample.

 

If you're confused, then don't worry you're not the only one...

Eloise

 

P.S. And not that it matters. But Tact is (IIRC) now known as Lyngdorf. Or is this just Europe - looking I'm confused now.

 

[Bezerkely]

You're right. I forgot iTunes and audio midi had such a strange relationship. I should redo the comparison with audio midi set to 16/44.1. However, I still think Play will win. iTunes has always sounded flat to me.

 

As far as I know, Play is just a front end to allow easy playback through core audio and any manipulations of the digital signal (other than the volume slider) would have to be through Audio Unit plugins. I suppose I could buy an AU upsampler instead of using the core audio upsampler via Audio Midi.

 

Tact was Boz & Lyngdorph, who parted way several years ago. Boz is still going under the Tact moniker, while Lyngdorph has spun off his own designs using his work at Tact as a starting point and now has a distinct line of his own.

 

[serengetiplains]

Eloise, Boz was the brains behind the original Tact. He's a DSP guy. Lyngdorf is a sales guy. I own Lyngdorf myself, but I suspect Tact has the sonic edge, particularly with room correction and crossover functions.

 

[Audio_ELF]

Thanks to you both for clarifying the Lyngdorf/Tact relationship.

 

 

[One and a half]

is lit when the HDCD is detected?

 

[audioengr]

Owning a Mac Mini as well as a number of PC's and laptops, I can make some recommendations:

 

When you change the sample-rate or other device properties in Audio MIDI, you must close iTunes, make these changes and then re-open iTunes. This means eliminating it from the toolbar. It is not really closed until the iTunes Icon is not anywhere on the desktop.

 

iTunes should be avoided for PC. Mac is better for iTunes, but I have found, at least with my Mac Mini that the jitter from iTunes is much higher than the jitter using a PC and Foobar 0.8.3 with device unmapped.

 

I did a number of comparisons to determine this:

 

PC -> USB cable -> Overdrive USB DAC

Mac -> USB cable -> Overdrive USB DAC

Mac -> USB cable -> Tascam US-144 -> Pace-Car reclocker -> S/PDIF -> Overdrive DAC

PC -> USB cable -> Tascam US-144 -> Pace-Car reclocker -> S/PDIF -> Overdrive DAC

 

The straight USB scenerios compare jitter. The Tascam/reclocker scenerios compare data integrity.

 

In all cases the PC won for best audio quality. The reclocker with the Mac was close, but not as 3-diminsional or live sounding as the PC using either USB direct or with reclocker. A bit dry from the Mac. This indicates that Core Audio is somehow changing the data as well as generating a lot of jitter. If I were to put numbers on the effects, I would say that jitter was about 80% of the problem with iTunes and 20% Core Audio data integrity problems.

 

The fact that the Mac Mini is grounded and the PC laptop that I used was not does not matter because the S/PDIF input on the Overdrive DAC is galvanically isolated.

 

I have ordered the Amarra test software. Next week I will see what effect this has on both jitter and data integrity.

 

Steve N.

 

 

[cfmsp]

 

 

Steve says:

"This indicates that Core Audio is somehow changing the data as well as generating a lot of jitter. If I were to put numbers on the effects, I would say that jitter was about 80% of the problem with iTunes and 20% Core Audio data integrity problems."

 

Steve,

 

I stand in awe of your ability to ascribe percentages to the contribution of problems based on your subjective observations that the music is a bit dry and less "3-diminsional" (sic).

[sarcasm mode: off]

 

iTunes on the Mac has been proven (rather easily) capable of bit perfect playback, except when not set up correctly.

 

Perhaps the data integrity issues are 'user error'?

 

Not having iTunes setup for bit perfect playback would seem to invalidate your entire test, yes?

 

just curious