Jump to content

Advice integrating DAC into Mac Pro / iTunes listening setup

Recommended Posts

I just ordered myself a new DAC -- the Lavry DA11. This is on the recommendation of my sound man, who built me a really amazing 3-way active TAD-loaded system. Until recently, my signal path was this:


iTunes > Mac Pro > Firewire > Apogee Ensemble > Analog system (via TRS-XLR balanced Mogami cables)


While the Ensemble sounded fine when my analog system was a pair of Adam P22 nearfield monitors, I found it to be disappointing on the big full-range system. Of note, I found the imaging to be poor, the upper octaves brittle and the overall tonality dissonant. So I tried something different:


iTunes > Mac Pro > Firewire > Apogee Ensemble > SPDIF/Optical > Krell HTS 7.1 > Analog system


The improvement was not subtle. The Krell's DAC blew away the Apogee Ensemble's DAC in terms of imaging, tonal balance and -- if you can believe it -- warmth. I say that because Krell are not known for being warm. They are known for being transparent.


But as an audiophile, I still wasn't satisfied I was getting the best out of my system. The Krell HTS is a home theater processor, and for 2-channel listening, I felt it would make more sense to go with a dedicated 2-channel DAC ... hence the Lavry DA11 is on order. I actually don't need the Apogee Ensemble anymore -- even for my hobbyist music product. I really only need 2 outputs at this point, so would consider getting a dedicated PCIe card to replace it. My question is this ... Of the following options, which is most recommended, and why?


1. iTunes > Mac Pro > Firewire > Apogee Ensemble > SPDIF/Optical > Lavry DA11 > Analog system

2. iTunes > Mac Pro > Lynx AES 16e > AES/EBU > Lavry DA11 > Analog system

3. iTunes > Mac Pro > RME AIO PCIe > AES/EBU > Lavry DA11 > Analog system


Any thoughts would be greatly appreciated. Thank you!


P.S. One more point ... I definitely intend to demo Sonic Studio's Amarra to get the best out of iTunes.





Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment

I'm curious about whether you've tried leaving the Ensemble out of the loop altogether, and if so what was the result? I suspect (not 100% sure) that your Mac Pro will have one of those dual purpose mini optical outputs, which could connect directly to the Krell or Lavry if they have optical inputs.


It will limit the resolution (osx won't go higher than 24/96 via this route), so may not be your ultimate best solution if you want to explore really high res music. But on the other hand it should be enough let you try out the Amarra demo, maybe help you decide what your best next purchase will be.


Sorry, can't comment on the relative merits of the Lynx or RME cards.


Link to comment

Thanks for your reply. I haven't tried taking the Ensemble out of the loop, as it is the only way I can make everything reach. Mac Pro and listening room are too far apart with no way to change it right now. This is why I'm thinking about doing the AES/EBU solution -- 110 Ohm cable can be run long distances (I need 30 feet total). Thanks!


Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment

I did as you suggested, and with very concrete and obvious results. I used the Toslink/Optical output on the Mac Pro and fed that straight into the Krell HTS 7.1. With iTunes the result was noticeably worse. As a side note, when playing with Amarra from Sonic Studio, the sound improved quite a bit. Here is what I think is going on after speaking to several people at various companies.


1. Optical outputs from the Mac Pro are poor in quality -- they output with a lot of jitter. The Krell HTS 7.1 is a home theater processor, not a 2-channel DAC. It does not re-clock the incoming data. As a result, you can't eliminate the jitter problems with the Krell.


2. I believe that when using a DAC with a good clock source, this problem can be reduced quite a bit. However, there is an important thing to keep in mind. When sending digital data in the professional environment, you generally tell each device which clock source to use -- (a) internal; (b) external; or © word clock. Internal uses the devices own clock. External sends the signal expecting the receiving piece of equipment to set the clock. The third option is used when you have many digital devices that need to by synchronized. By using a BNC connect to a rock-steady clock source, all devices will be in sycn. Very important for multi-channel digital audio tracking and playback.


3. This is why I have ordered a Lynx AES16e. It is a professional card, and you can output through AES/EBU (essential if you want to run long distances like I do). You can also set the clock source to external, which I don't think you can do on the Mac Pro's built in optical output. I have also spoken to some mastering engineers, and they like this card. They say the difference vs. an RME or other PCIe-based card outputting AES/EBU is really splitting hairs. The reason they like the Lynx card is that the drivers are solid and reliable.


4. As a side note, from actual tests, a good sound engine like Amarra can help immeasurably with stabilizing the image/data coming out of any output. And to be clear, Amarra is not even approved for use with the built-in optical outputs on the Mac. However, I was able to get the built-in optical output to send 24bit/96kHz data to the Krell DAC, and it sounded a lot better than iTunes, which seems to truncate the data -- at least on the built-in optical output.






Let's see what happens overall ...


Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment


"However, I was able to get the built-in optical output to send 24bit/96kHz data to the Krell DAC, and it sounded a lot better than iTunes, which seems to truncate the data -- at least on the built-in optical output."



it's possible that you're not aware that you need to properly set the Sample Rate in Audio Midi Setup prior to launching iTunes. That may be at least partially responsible for the poor sound you are experiencing.


That's the only reason I can think of in which iTunes would 'truncate the data' on ANY output.






Link to comment

Hi Clay,


Yes you are right about that. I figured that out afterwards. After taking out the Ensemble, AMS looked so unfamiliar, I didn't realize how to do that until afterwards. Even at the higher sample/bit rate settings, however, the sound isn't so good with the Krell HTS. A fair test will be with the Lavry DA11. Their engineers claim the sound should be quite good this way.




Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment




Did you make sure to set your Mac Pro such that iTunes was doing the upsampling and not OS X? It has been shown that Itunes upsamples way better than OS X does. Here's how to make sure it happens;


1. Quit iTunes.

2. Launch Audi midi setup inside Applications/Utilities (assuming Mac OS X 10.4 Tiger, 10.5 may be slightly different)

3. Set Audio midi for your output device to 24/96.

4. Launch Itunes.



From now on, itunes will take whatever audio it has and deliver it to the OS at 24/96 before the OS gets a chance to upsample it. The OS will then take that cleanly upsampled audio and deliver it to the optical outputs untouched.


Try that and see how things sound on your system.




Link to comment

Thanks for doing the test and reporting back, sunjitwhy. I am finding it very interesting and look foward to your thoughts on the Lavry as well.


I also think that clocks might be a very important factor when it comes to translating the pure data of the computer world into the analog world of sound (I hate to use the j-word!), but to be fair I'm just speculating at this point.


The idea that Amarra, or other software, can somehow present the data in a more effective manner is an intriguing one, isn't it? The price kind of puts it out of my league for short term purchases, but I keep going back to the demo, switching it on and thinking yes, it's definitely got something.


Link to comment

Hi again. Per your advice - thank you - I did go through AMS and change sample rates appropriately to playback properly from iTunes. This works insofar as iTunes will stream 24bit/96kHz data when AMS is set properly. Unfortunately, this still does not seem to improve the sound staging.


As an aside (sorry ... I am prone to these!) ... It is very silly that iTunes doesn't do this automatically. Logic Pro - the studio program I use for my own "music" productions changes the sample rate automatically depending on the project you open. iTunes doesn't do this, which is silly ... they have the technology. They bought the German company eMagic, which invented Logic. That happened two years ago, so it would seem so easy for them to offer the functionality in iTunes, or even in "iTunes Pro" for a modest fee. Given their user base, they could easily spend $10-20m building the best sound engine for an "iTunes Pro" and only need to charge $100-200 for it.


I really do believe that sound engines matter within software. Logic is beautifully designed (by the Germans) and sounds brilliant. It sounds very close to Amarra, I think, though mastering engineers will still tell you that SoundBlade is among the best, and this is the engine that Amarra is built on. I have always noticed that when I produce some music, it sounds wonderful (to me at least, since I wrote it -- ha ha) coming out of Logic, but once mixed down and played out of iTunes, it doesn't sound as good. Too bad it is so expensive ... I think they should sell it for 75% less and spend more time expanding the user base to generate the payback on R&D.


Anyway, that was a long aside ... Thanks for the tips!



Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment

I just got an email from the folks at Lynx, and here are their thoughts ... They sell both firewire and PCI cards, so I wouldn't totally discount what they are saying. Mastering engineers seem to feel this way too.


1. FireWire is not as stable a format as AES, especially at higher channel counts and/or at higher sample rates.

2. FireWire has more latency than AES. Our AES16e has even less latency than our AES16.

3. SPDIF is an unbalanced format, AES is balanced. AES also works at a higher voltage, which adds to the robustness of the signal.

4. Optical has higher jitter. Using AES would also avoid this.


FYI : Latency shouldn't matter to audiophiles. It matters in the studio. Latency is the delay between the time you hit a note on the keyboard and then hear it back through the speakers. If there is too much latency in a system (particularly a problem when you have several musicians playing at the same time while tracking), it is impossible to play with any sense of timing.




Sanjay Patel | Ciamara Corporation | New York, NY | www.ciamara.com

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Create New...