New DAC has distorted/clipped audio w/new Mac and OS 10.10

aroide · 17989

0 Members and 1 Guest are viewing this topic.

Offline Bodyslam

  • Jr. Member
  • **
    • Posts: 24
Reply #45 on: May 11, 2015, 04:37:08 PM
Grainger,
Since you're using a DVD player, which is primarily designed as a video player, it very likely switches back to the 48K family of sample rates whenever it doesn't have a recognized CD in it. Can you look at the sample rate display on the BottleDAC when you open the tray and see what it's reporting?

The Oppo will probably do the same thing.

Paul Stubblebine
Paul Stubblebine Mastering, San Francisco
The Tape Project, LLC
serious student of the audio arts


Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #46 on: May 12, 2015, 11:22:05 AM
Two answers.  Paul, I just ran a test today.  I went up and opened and closed the player 6 times.  The displayed sample rate stayed at 44.1.

John, the above test was after the system played for 5 hours.  So it doesn't seem to have anything to do with the player being "hinky" after hours of listening. 



Offline John Swenson

  • Full Member
  • ***
    • Posts: 114
Reply #47 on: May 13, 2015, 02:08:55 PM
John,

I switched to using an older laptop (Ivybridge based) to work around the issue.  The DAC really shines through USB vs toslink and I didn't want to compromise.  When I get a chance, I'll hook my 2014 mac min back up and record the sound.

The distorted sound only happens for my using Audirvana.  I get that sound for a few seconds and the DAC no longer is recognized.  Audirvana uses integer mode of USB audio 2.0 and may be unique.  If I try and drive the DAC from the OS X system sound, I get NO SOUND at all and within 2 secs the BH DAC disappears as a device.  So the distortion may or may not give you a clue.

What I really want to do is to bring my DAC to work and get some traces on a USB gen 2 protocol analyzer.  If I can get some traces, are those useful to you?  I know for my team, its hard to use a trace without firmware source and driver source.

Tony

Hi Tony,
so it works fine with another computer? It's just this one Mac Mini that has the problem?

Yes a protocol trace on the machine that has the problem would be a big help in tracking this down.

The BH DAC has been run in integer mode with Audirvana many times without any problem, so I don't think that per se is the issue. It might be some interaction between that and other setting you have that don't exist on other's Mac Minis.

Thanks for sticking with this, I really want to get it figured out.

John S.


John Swenson
DAC designer
Well Tempered Record player -> Seduction
Moded Squeezebox->DIY DAC
BDT preamp->813 monster SE amp
DIY "Bazooka" Lowther speakers


Offline fullheadofnothing

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 1487
  • A noble spirit embiggens the smallest man
Reply #48 on: May 14, 2015, 07:19:53 AM
John-

This problem has been traced to ONLY be an issue when running a Mac with 10.10 AND USB 3.0. USB 3.0 works on OS 10.9. OS 10.10 works on USB 2.0. This is an issue with Apple's software. This is why I have been suggesting that those experiencing the issue contact Apple. The DAC that Tony has was tested here on multiple computers before it was ever sent to him. After he encountered issues, he sent it back to us, where we verified it further.

Grainger's issues are completely different from what Tony has been talking about. I believe it is the issue that I encountered and wrote to you about previously (with recordings, that you identified as "[...]is caused by the S/PDIF receiver (either optical or coax) having a signal, but the clock tracking circuit is prevented from tracking because USB uses a fixed frequency clock, when switched to S/PDIF (coax or optical) the buffer is now so far out of whack with the data that it is messed for a while until the tracking mechanism can get synched up.")
« Last Edit: May 14, 2015, 07:29:23 AM by fullheadofnothing »

Joshua Harris

I Write the Manuals That Make The Whole World Sing
Kit Packer Emeritus


Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #49 on: May 14, 2015, 10:53:52 AM
John and Josh,

I don't consider it to be much of a problem.  If it bothers you guys we can continue with it.  But I'm very happy.



Offline aroide

  • Full Member
  • ***
    • Posts: 61
Reply #50 on: May 20, 2015, 05:47:34 PM
I'll see when I can find the time to try and get a protocol analyzer hooked up.  I'm no expert at these details so I need help from my team (I'm the lowly manager type).

I wouldn't jump to a conclusion that this is exclusively an apple issue.  My other xmos dacs work just fine where the bh dac does not.  My team at work encounters these interactions all the time and normally it is some subtle disconnect in implementations or different interpretations of a protocol or spec.

I fear that apple changed USB hub chips and there has been a hardware change that is part of the issue.  It happens on all 2014 and 2015 macs starting with the Haswell processor versions.

For now I'm back to a mid-2012 MacBook Air to work around the issue and the bh dac works fine.  Btw this is USB 3 system running OS X 10.10.3 using the intel hub.  On my 2014 Mac mini the sub ports connect through a Broadcom hub chip.


Mac mini running Roon->
Mytek Brooklyn DAC+->
Darwin Truth RCA IC->
BeePre (BeeQuiet,  Mundorf Ag/Au/Oil, Sophia Royal Princess 300B)->
Audioquest Colorado->
Rogue Audio Atlas Magnum amp (Psvane small signal tubes, KT150s)->
Audioquest Gibraltar-speaker cables>
Magnepan 1.7 & REL T-5


Offline Doc B.

  • Administrator
  • Hero Member
  • *****
    • Posts: 9554
    • Bottlehead
Reply #51 on: May 20, 2015, 07:35:32 PM
Thanks Tony, it would be great if you could share any info you may find. Knowing how well the DAC works with every setup but this one combination will hopefully motivate us to work together toward a solution.

Dan "Doc B." Schmalle
President For Life
Bottlehead Corp.


Offline caffeinator

  • Sr. Member
  • ****
    • Posts: 270
Reply #52 on: October 11, 2015, 06:14:45 PM
I just saw this thread and started writing in response before I realized it is a few months old.

FWIW, what aroide was describing sounds very similar to what I heard when I upgraded my Mac Mini to OS 10.10.5 - a painfully harsh chopped kind of sound, like the music was literally going through some kind of chopper (intermittently cutting out time-slices).  Recognizable as music, yes, but beyond unlistenable.

In my case, after restarting, then being sure Amarra was running, the problem seems to have gone away, though the sample rate display seems to read 44.1 all the time, even when playing higher-res files - I am still trying to test that more thoroughly, though.