[stjj89]

RE the Multibit Bifrost Sample Rate Change situation. I have the following interim news to report. The first comments relate to statistics – our sample size of MB Bifrosts upgrades and new units sold is well over 1000, not enough to round up to 2000 quite yet, but very, very, close. We replaced units all both units we could find where the sample rate change glitch continued after the units were a hundred or so hours of initial burn in time. You can do your own percentage calculations. Both users reported that the replacement units did not solve their sample rate glitch. Does this mean we think we have defective users?? No. There was too much smoke, even though we continue to believe that the overwhelming majority of these issues disappear after that burn in. That said, we have quanitatively figured out the problem, but not yet qualitatively.


 


[COLOR=000000]It took a while. First thing was to eliminate any failure exceptions. We found early on we could eliminate TOS and Coax S/PDIF, since there were generally no sample rate changes, but I burned a CD with different sample rates and we were clear. Dave and myself wasted hours trying to replicate the sample rate change glitch but could not do so – then it occurred to me – Dave and myself both use macs for USB streaming. So we each hooked up a PC. Voila! Failures! Not universally, but intermittent ones. Syncing my scope on the USB packets, the sr change behavior was quite different on either of of my two PCs running Windows(as well as Dave's), informing me of the lack of PC to PC reproducibility with the Redmond, WA OS. On the Macs, the output was continuous, just as it should be. In the below link, I discuss different OSes for audio use. As an experiment, I booted one on my PCs into Ubuntu without installation and tried the same sample rate changes and the result was that it behaved just fine with no glitches on USB streaming. The USB packets behaved normally.[/COLOR]


 


[COLOR=000000]http://www.head-fi.org/t/784471/what-a-long-strange-trip-its-been-robert-hunter/75#post_12132242[/COLOR]


 


[COLOR=000000]So it appears that the problems manifest themselves with USB2 audio over Windows victims. In all fairness both Dave and I were able to get a Mac to fail once, mine with the defective unit from Seattle and Dave's the one from Florida. It took both of us an entire evening to make that happen, continuously changing sample rates which is an odd use case. We can also eventually get a mac to fail on USB by rapid plugging and unplugging the USB cable.[/COLOR]


 


[COLOR=000000]Despite my contempt for Windows as an audio OS (until they add UAC2 to their OS), I will do all I can to figure out what is wrong with our confirmed (in the field) 0.16% continuous sr change glitch failures. I hearby assert is our intention to support Windows, no matter what it takes. I would also be eternally grateful if the Seattle user could burn me an audio CD of one of his glitching playlists c/o Schiit Audio. I promise to report back.[/COLOR]

Thanks for the update. It is good to know that you've managed to narrow down the possible causes of this issue. I appreciate the effort you are putting in to try to resolve this!

 

[Collot]

 
It took a while. First thing was to eliminate any failure exceptions. We found early on we could eliminate TOS and Coax S/PDIF, since there were generally no sample rate changes, but I burned a CD with different sample rates and we were clear. Dave and myself wasted hours trying to replicate the sample rate change glitch but could not do so – then it occurred to me – Dave and myself both use macs for USB streaming. So we each hooked up a PC. Voila! Failures! Not universally, but intermittent ones. Syncing my scope on the USB packets, the sr change behavior was quite different on either of of my two PCs running Windows(as well as Dave's), informing me of the lack of PC to PC reproducibility with the Redmond, WA OS. On the Macs, the output was continuous, just as it should be. In the below link, I discuss different OSes for audio use. As an experiment, I booted one on my PCs into Ubuntu without installation and tried the same sample rate changes and the result was that it behaved just fine with no glitches on USB streaming. The USB packets behaved normally.
 
http://www.head-fi.org/t/784471/what-a-long-strange-trip-its-been-robert-hunter/75#post_12132242
 

Feels great to see that things are moving in the right direction. Note that I have myself encountered the problem on mac and Linux as well. BTW, I've read your linked post:

Originally Posted by Baldr /img/forum/go_quote.gif
 
Linux, you say? Yup. The modern distros are getting easier and easier to use – even rivaling/beating newer PC stuff. It shares to Mac advantage of Unix roots. Stability and consistency. Unlike Windows, USB2 audio is built in. The newer Linux distros for the most part figure out your hardware and build it as kernel modules as it installs. At higher sampling rates it is still not as good as the Mac (requires some tweaking) but for Redbook it is just fine. Oh, and did I say? It is free, just provide an old PC to install it on.

You mention tweaking, would you mind sharing the settings you have been using while testing rate changes on Linux? Did you stream audio to the device directly (hw:xx,xx) or to the 'virtual device'? The latter implies software resampling and probably wont trigger the problem.

 

[defbear]

RE the Multibit Bifrost Sample Rate Change situation. I have the following interim news to report.

 


[COLOR=000000]Despite my contempt for Windows as an audio OS (until they add UAC2 to their OS), I will do all I can to figure out what is wrong with our confirmed (in the field) 0.16% continuous sr change glitch failures. I hearby assert is our intention to support Windows, no matter what it takes. I would also be eternally grateful if the Seattle user could burn me an audio CD of one of his glitching playlists c/o Schiit Audio. I promise to report back.[/COLOR]

I didn't care to quote the entire post. Thank you for this plain spoken report. It is very beneficial to me to hear how serious you have taken the problem. Thank you for detailing the steps you have taken to resolve the issue.

 

[billerb1]

I didn't care to quote the entire post. Thank you for this plain spoken report. It is very beneficial to me to hear how serious you have taken the problem. Thank you for detailing the steps you have taken to resolve the issue.

 +1

 

[aamer23]

I still think we are missing the main point say there is a hardware / software bug causing issues on rate changes just use a play list with a single rate until Schiits understand this Schiit. what i have found that music is so much better i am listening to more of it for longer due to the MB upgrade. the sound is more natural like a live performance and i dont hear distortion as much. even youtube streamed concerts sound better 

 

[Ableza]

Random issues that are user-system-dependent are often nearly impossible to reproduce in the lab making them impossible to "fix."  Normally if the product works to spec and this occurs with my company's equipment, and we cannot reproduce the failure, we refund the purchase and end the issue.  Fortunately we sell only to industrial users which means the users' systems tend to be fairly regular.  Consumers will present a nearly infinite variety, and no company can design for the endless combinations of PC/software/networking gear that exists in the world  And you can't fix every users' system for them unless you eliminate their system, and this would drive your price points up by hundreds of dollars if it could be done at all.  Not all products will work for all users, sadly.

 

[earnmyturns]

  Random issues that are user-system-dependent are often nearly impossible to reproduce in the lab making them impossible to "fix."  Normally if the product works to spec and this occurs with my company's equipment, and we cannot reproduce the failure, we refund the purchase and end the issue.  Fortunately we sell only to industrial users which means the users' systems tend to be fairly regular.  Consumers will present a nearly infinite variety, and no company can design for the endless combinations of PC/software/networking gear that exists in the world  And you can't fix every users' system for them unless you eliminate their system, and this would drive your price points up by hundreds of dollars if it could be done at all.  Not all products will work for all users, sadly.

This is exactly why Apple controls all of their hardware+software stack and charges a premium. In the audio domain, custom-built digital sources typically sell for multi-$K (see Auralic, Linn, Bel Canto, ...) to feed their DACs. One might (maybe optimistically) hope that in those cases source and DAC are better matched than what Schiit can achieve between the infinite variety of PCs and a $600 DAC. OTOH, the quality I get out of that $600 DAC is awesome for the price. I've observed the same kind of tradeoff in specialized skiing and mountaineering gear: you can get something that works OK for a broad range of uses but as a result it is too heavy or too expensive; or you can get something that fails in some uses, but it is great for its range of applicability, and is lighter or cheaper. Engineering is all about compromises, and sometimes the precise boundaries of the compromise do not become apparent until the product is fielded (in the ski realm, the notable and long-lived case are Dynafit touring bindings, which are super-light but do not satisfy the release expectations of some aggressive skiers).

 

[aamer23]

  This is exactly why Apple controls all of their hardware+software stack and charges a premium. In the audio domain, custom-built digital sources typically sell for multi-$K (see Auralic, Linn, Bel Canto, ...) to feed their DACs. One might (maybe optimistically) hope that in those cases source and DAC are better matched than what Schiit can achieve between the infinite variety of PCs and a $600 DAC. OTOH, the quality I get out of that $600 DAC is awesome for the price. I've observed the same kind of tradeoff in specialized skiing and mountaineering gear: you can get something that works OK for a broad range of uses but as a result it is too heavy or too expensive; or you can get something that fails in some uses, but it is great for its range of applicability, and is lighter or cheaper. Engineering is all about compromises, and sometimes the precise boundaries of the compromise do not become apparent until the product is fielded (in the ski realm, the notable and long-lived case are Dynafit touring bindings, which are super-light but do not satisfy the release expectations of some aggressive skiers).

yep horses for courses for 600 usd for top class SQ is very good better than getting 10% extra SQ for 60000 usd

 

[Soundsgoodtome]

A bump to jump the fanbois.
 

 

Spoiler: fanboi alert

Why do all the Schiity fanbois post a bunch of useless garbage on this thread as if to bury all information good or bad? What may be being posted here could help Schiit in fixing this issue but when there's an additional 5 pages per day and half of the pages are garbage, then you can see why this is getting tiresome for some.

 

Even though what's being broadcasted by Schiit directly as less than 1% being effected, I think that may be a bit optimistic. Reason for saying is MY EXPERIENCE alone would leave me to believe that this is not a rarity thing to occuring. I've been in contact with 3 different Schiit Bifrost Multibits. 2 of them being mine and another being another member here in Seattle. I can't say for the other member here but I believe he has a new unit and not an upgraded one. As for my 2 Bifrost Multibts, one was an upgrade from a Bifrost Uber and my current one being a new build. All three of these units have issues with the sampling rate changes, with one having the possibility of fixing itself (because it's still around 150hrs or so).

So with 2 of 3 units I've come in contact with along with the third unit still experiencing this issue...you can see as to why I believe the 1% Schiit says to have an issue may seem optimistic (in my experience). If Schiit can't in fact get this to occur at their bench so they can find a fix, as some would say, why don't they ask the owners of who are experiencing issues for a little bit of information with their setup. I've had 0 questions asked about my setup to maybe find a culprit.

Someone said this may be your system-by-system basis, meaning it's the users setup screwing up. Well I've had this on 2 different computers on 2 different Windows system with different USB cables. Both experienced the issue including the new unit sent to me as I suppose I'm to 'sink test'. I believe the issue here is somewhere in the USB chain of the unit or within the drivers itself, I'd take the probability of the former to be higher.   The reason I believe the USB is the issue? Using a Gustard U12 transport (USB to optical or coax), I've had 0 issues with sampling rate change buzzing and the BIMBY. The relay still clicks with the new sampling rate like it would via USB but absolutely no buzzing what-so-ever.

 

[aamer23]

  A bump to jump the fanbois.
 

I take it that USB2 is the being used. When i upgraded Bifrost to Uber I could not get USB2 to work so I used USB3 on the i/o at the back of PC this fixed the issue noise and Bifrost  did not like so called audiophile grade USB cables had to use standard PC grade USB cables this fixed my problem. Windows 8 and 8.1 pro 64 bit

 

[jimmers]

I'm still waiting for Baldr to give his opinion on which old Mac makes the best music server.
 
 
P.S. Baldr, if you are waiting until you have a back-up unit before you tell us, I will try to find one for you .... if you tell me what it is.

 
 
Edit: He just said on his thread - should have posted/checked there - all these streams of Schiit, so confusing.
http://www.head-fi.org/t/784471/what-a-long-strange-trip-its-been-robert-hunter/120#post_12198527

 

[afrochops]

I have been seeing the sample rate issue about 1-2 times per week. I'm on a MacBook Pro laptop and use a USB 3.0 hub, the hub could easily be the culprit since I have lots of usb problems. Just replaced the cable to the Bimby with a Straightwire USB cable.

 

[Soundsgoodtome]

Did I hear that right? I thought Apple had no issues like this lol

 

[aamer23]

Did I hear that right? I thought Apple had no issues like this lol

you missed usb 3 hub 

 mac usb is perfect 

 

[Soundsgoodtome]

Fact or possibility? Mac-in-trash is what we used to call em.