Schiit Bifrost 2

I was also hearing the aliasing both on the .ogg files in the browser and on the .wav files for the high-res sweeps using Foobar2K with WASAPI. This was over USB, my Bifrost 2 is still on the original firmware and had been left on since I received it two weeks ago.

I still have my Modi Multibit and Eitr, so I set those back up and, sure enough, the aliasing was completely gone coming through them. I then attached the coaxial from the Eitr to the Bifrost 2, and I ended up rebooting both the Eitr and the Bifrost 2. Then I retried all the tests and... behold, no aliasing. Seeing that you had heard the aliasing on both USB and SPDIF, I knew it couldn't be that simple and decided to go back to USB to see if I still heard it... and again, no aliasing.

A bit later, I again played back the .wav files in Foobar and noticed that the aliasing had returned. Another power cycle of the Bifrost and it again disappeared. So it appears that rebooting the Bifrost 2 refreshes whatever is causing the issue.

This suggests to me that the Bifrost 2 still needs the occasional power cycling that my Modi Multibit did every so often when it would start sounding very distorted. If this is the case, the Bifrost 2 version of this phenomenon is much more subtle.

 

Thanks for that. More salient for me, that led to a confirmation that my replacement unit has the bad cap problem which I'd read about as existing, but had no idea what symptoms were.

We have a bad batch of surface mount capacitors (ironically, the same 1206-sized part that we have used for years in other products). When they fail (shorted), they can cause one of the internal cards not to be detected. If you have the "three flashes" problem, contact us so we can swap out your Bifrost.

https://www.head-fi.org/threads/sch...bable-start-up.701900/page-3405#post-15200188

I've been testing mine with different combinations of power and USB source, but have gotten nowhere (it's most recently managed to go for 3 days between the three flashes problem). It's now occurred five times in the past ten days, but with the frequency of occurrence tapering off with time. It's been very odd. Edit: spoke too soon, it just went into failure again within an hour of the last reboot so the apparent pattern was just random.

I guess that means no more troubleshooting at home with different combinations of wall socket, cable, input source- the unit is going back next week.

 

It sounds more or less the same with both the ogg file and wav files. Louder on some than others, but easily heard without having to crank the volume.

A thousand likes!

After a power cycle it's clean as a whistle. With Modi Multibit, it was obvious when it needed a reboot. Playing music on BF2 there doesn't seem to be anything obviously audible, but I'm now going to have to get into the habit of regularly power cycling the BF2, perhaps leaving it on and doing it before each listening session. I'd be curious to see how this behaviour shows up in measurements, peaks at subharmonics of the test tone I suspect.

 

If you heard the aliasing I am, you wouldn't call it "nervosa". It really sounded "broken" and is easily audible while the tone sweep is well within the limits of my hearing range (I'll try and make a recording of it next time it happens). Worrying about whether or not it's doing it could turn into one. Thus the easy solution is a quick reboot. No need to wait 17 days.

 

Well hopefully this gets reported quick and fixed via firmware using the free cards sent out to everyone.

 

Try the new firmware before going further. What you’re hearing may be related to an input cycling glitch we fixed as part of the update. If you don’t want to write your own MicroSD card, just wait a few days, we are sending them this week.

 

@AudioNut - could the alias issue reported by @gixxerwimp and @TazerMonkey have been involved with your experience?

Could any or all of you help with what conditions reliably replicate the alias issue? If so, I'd appreciate knowing so I could repeat and measure.

Like @rtaylor76 I haven't heard this alias and my unit is powered 24/7 most of the time. BUT: I run Dante to spdif at 88K and only very infrequently switch sample rates or digital inputs.

 

Not sure about how to reliably replicate. I accidentally discovered this yesterday. Power cycled this morning and it went away, listened all day, just checked again now and it's back. RPi > USB all day today.

Edit: Power cycled just now and it's gone. Will check again in the morning (00:41 local now).

Edit2: Powered on all night, no music playing, no aliasing in the morning.

 

My replacement bifrost 2 developed the same problem as the original this morning (that's 10 days after the replacement). Three flashing lights, no sound. I turned the unit off, waited a few minutes and then back on. The bf locked on to a usb signal and has been playing without issue for about six hours.

I sent out an email to support and they responded back within an hour (on a Sunday no less) and suggested that I update the firmware. I assumed that the replacement unit would have already had the newest firmware, so I didn't install it when it was first posted here. Is there a way to verify the current firmware version and also that you've done a successful update?

So now with newest firmware everything seems solid. I'm adding this post as another data point for bf issues. I love the sound and functionality of this DAC, but...I've become leery about the BF. If this unit is rock solid for the next several weeks and I can use it without a hiccup, then I'm happy enjoying a wonderful sounding product. But, if it goes down again then I'm not sure if I'll have much confidence going forward. I don't want to be in a position where I'm waiting for the next failure. And if that is the case then I'd question whether the BF was honestly ready for prime time.

I'm not sure how many others are having the same issues but it doesn't seem that there are only a few isolated cases. Hopefully @schiit can get the issues resolved because I really do like the bf when it's working.

 

Unfortunately I don't really know how to replicate the issue either. Mine had been powered on 24/7 since the 14th until yesterday through USB. I haven't had any other glitches, just noticed the aliasing after reading @gixxerwimp 's post and attempting it myself. It originally came back just a few minutes after the first power cycle for me, but after the second one yesterday, I just checked and the aliasing is still gone almost a whole day later.

Either way, I'm going to go ahead and update the firmware per @schiit in a little while.

 

Posting this just to say mines been sweet since day 1, which I'm grateful for because I'm all the way down yonder, and you only ever hear about bad cases.

 

Because I hope I'm not the only stupid a$$ person in the world... a word of caution on the firmware update.

The microSD slot is recessed from the back of the case, allowing a gap around the slot and the case. This gap eats microSD cards. Don't be like me and have to take the case off the BF to get the card back.

Be careful.

 

Isn't taking the case of as simple as sliding the top piece forward? The Lyr 3 requires you take the volume knob off and then take the nut off the volume pot, then you can slide it forward. I would't think there is anything keeping the Bifrost 2 from just being able to slide forward since there is no volume knob to anchor to.

 

It is easy... unscrew 4 screws, slide case off, pull out microSD card, put case back together. It's not that it's hard, its just something that should be unnecessary (as long as you're not a dumbass like me)

 

@dasman66 Sounds like an easy enough mistake to make if the gap is big and you don't insert it at a perfectly horizontal orientation. Your forewarning will save the rest of us from making the same (dumbass) mistake. And maybe @schiit can put a cautionary note in their instructions.

 

Thank-you all for input on the alias issue.

For those asking about output impedance I measure on my SW51+:
Lo Z 27R
Hi Z 54R

 

This morning I left it on repeat playing the 4 hi-rez high frequency test tone tracks (wav) and came back a couple of hours later to find the aliasing had returned.* Luckily, I still have the Behringer UCA202 USB audio interface recommended by @atomicbob for determining Modi Multibit firmware version and was able to use it to record the aliasing (44.1 kHz, 16-bit wav converted to flac). Chain was RPi Zero W (piCorePlayer) > Bifrost 2 USB > SYS > SW51+ > UCA202 > Audacity. The tone sweeps are the ones I linked to earlier. I didn't bother with the 192 kHz sweep.
https://www.audiocheck.net/testtones_highdefinitionaudio.php

The zip file contains 6 tracks, 1 to 3 exhibiting aliasing (tone sweep played 2x) and 4 to 6 recorded after a power cycle with no aliasing (tone sweep played 1x):

• 01. 88.2 kHz sweep 1 kHz to 44.1 kHz
• 02. 96 kHz sweep 1 kHz to 48 kHz
• 03. 176.4 kHz sweep 1 kHz to 88.2 kHz
• 04. 88.2 kHz reboot sweep 1 kHz to 44.1 kHz
• 05. 96 kHz reboot sweep 1 kHz to 48 kHz
• 06. 176.4 kHz reboot sweep 1 kHz to 88.2 kHz

​

https://drive.google.com/open?id=1r5JrUnFg-4TiXRdAL6KV9AvQRSwuST3X

@TazerMonkey is this the same as what you're hearing?

* @atomicbob I don't know if this is a way to replicate the behaviour, but you can give it a try.

 

Great detective work, @gixxerwimp

 

Checking that out quickly, I can hear a little bit of that up and down aliasing tone on the after reboot files as well, but the first three are definitely within audible range for just about anyone and is a problem. Is this with the new firmware?