@Mat Nice job on your assessment. I think for me it was more pronounced because I am using the Frost for my headphone rig, and I think things are more critical there.
To build on this a little bit I did the test this weekend where I re-cabled everything so that the Frost was now driving the DA-8S / speakers and the SU-8 was driving the TA-20 for the headphones. So let’s start on the speaker side.
There was an immediate overall improvement in sound quality when listening to the speaker setup through the Frost. Sound stage was wider and a little deeper, and I thought there was better overall refinement and a more natural quality in the sound. However, I liked the way the SU-8 delivered the lower end just a little bit better, while pretty much holding its own in other areas. So while noticeable for sure, not dramatic enough where you say “This has to stay like this”.
The dramatic improvement came on he headphone side. Listening to my phones over the SU-8 was good, but lacked that enveloping response and fullness that the Frost delivers. Just multitudes better in every respect.
So the net result is that the setup will stay put; The SU-8 will continue to drive the speaker setup and the Frost will run the headphone rig. When the SU-8 has depreciated enough I will go ahead and replace it with something better. I think part of the deal for me is that I think I enjoy the more natural and complete sound of a multi-bit/R2R implementation over a chip implementation. This means I am gonna have to save up for the next purchase! But I am in no rush, as the SU-8 is doing just fine. It pairs with the DA-8S very well, and honestly in my current environment I can’t do much better without really pissing the neighbors off, so I am set for now. I am glad I did the test, though. I learned quite a bit from it.
So I have been playing around with my Bifrost 2 for quite a while now and have found there are actually a couple ways to get a subtle but audible increase in quality. One of the most obvious being to follow what the heads at Schiit said themselves about using short usb cables on preferably a usb 3 or 3.1 jack. Which to my ears I have found to actually be true when it comes to perceivable time distortion in the upper end of the spectrum at least. The second less obvious one, is seemingly to set the Bifrost to one of its maximum sample rates, be it 32bit - 192000hz or 176400hz. For my purposes I chose the ladder. These are the only two sample rates than can be used to set the Bifrost into NOS mode (Disables its internal oversampling method) And instead utilize your own oversampling method of choice for lower sample rates. In my case, I chose to use a version of the “Sound eXchange resampler” custom made into a plugin for foobar2000. I set the target sample rate to what my windows settings are currently set to. To my knowledge this specific resampler utilizes your specified sample rate with 32-bit float calculation to get the job done.
While not a huge change, I did notice an overall increase in sound quality. Notably, above all, I perceive an actual increase in soundstage when I avoid the regular oversampling methods. If you run the Bifrost 2 utilizing its own oversampling. It sounds great on its own, albeit best with wasapi exclusive modes only because it forces the bifrost 2 to set itself to those specific sample rates (notably when you hear the click) and oversample entirely on its own. This seems to be entirely unique to this DAC and possibly other NOS capable devices as wasapi modes and changes in oversampling don’t usually seem to produce these kind of noticeable results with regular sigma delta dacs that I’ve had. Wasapi shared mode utilizes its own standard oversampling much the same as when utilizing Direct Sound, and unsurprisingly both sound exactly the same. They do not sound as good as letting the Bifrost 2 handle it all by itself to my ears. The key difference here seems to be the specific type of oversampling I am using in place of anything else so far (at least concerning foobar2000) To my ears this is yet another step up above the Bifrost 2’s own oversampling methods. I’ve been toying around with this at these settings since before I even knew the Bifrost was capable of NOS at these frequencies, and I am still in absolute disbelief with the difference that I am hearing, because the difference is still fairly subtle, yet I’ve been able to replicate this difference to my ears time and time again. If anyone has the time to sink, I hope someone else can help me verify whether this is really the case to my ears and that I may or may not be going absolutely insane.
@Doegiz That is some nice work. I’m feeding my Frost via digital coax, I decided to get away from USB. And it’s incredible. But tonight I tried the Unison USB module for the first time. I got an Audioquest Cinnamon USB-C to USB-B cable and I’m using my cell as a streamer. This cable is super nice, btw. I’m gonna get another one and feed the U2 Pro via the laptop’s UCB-C interface. But yeah, the cell displayed a headphone symbol and went into some kind of HiFi mode as soon as it saw the Frost. Can’t get any easier than that!
I am very impressed with how my trusty old LG V40 performs in this capacity. I have tested both the home Wi-Fi and the LTE network to stream Amazon Music HD and it sounds fantastic. No EQ or anything, and it’s fantastic. This now gives me an alternate source should something happen to the PC. The mobile Amazon music app is really cool, too. It adds interactive features that the PC version doesn’t offer. So it’s fun to use. Honestly, though, I’m really impressed by how good it sounds. I love this DAC, man.
So I decided to look into this a bit more. Not so much about the Frost’s oversampling technique itself, just the resampling plug-in bit. I had no idea this DAC did that either, btw. As it turns out, there are two resamplers that come with Fubar, so I tried one from RetroArch.
And no, you are not insane! There was a noticeable overall improvement in the sound the minute I initiated it. Very cool tweak. Thanks for sharing that. I like tinkering with these things…
@Raptor168 Very interesting re the V40 and the Amazon HD app. I bought a Kindle fire
10 a few months ago on sale, partly because the web said it can play Am Hd music. I have tried it w. Spotify so far and have discovered that the usb output is 24/96,according
to my dac screen. That’s without an additional player and sounds quite good. Plan to try the Am. app too.
I was really surprised at how good this phone does in an audio delivery role. I have this cool little holder that props the phone up just right so it looks like the system’s control head, which is essentially what it is at that point. The way the app presents the songs and content is really cool, too. And though there is no EQ, the sound quality is incredible. I have been playing around with amp/headphone combos and I am still trying to figure out which one is my favorite. I guess those combos will be my EQ! With 2 amps and 4 headphones to choose from (5 if you throw in the Lagoons), that is a lot of different sound signatures. Don’t need EQ at that point! Want a certain sound? Combine this headphone with that amp, and BOOM! yer good to go. I’m lovin it!
I have recently revived an old iPad that my Wife retired after getting her new Surface Pro and have been wondering what that machine might do in this capacity. I’m not an Apple guy, but I have heard that these things deliver excellent audio. So this might a rabbit hole I go down in the future.
@Raptor168 Yeah I have an old LG g5 which outputs at 24/192 by default! Almost bought a V20 back when. Glad you’re having such fun, and that you’re eclectic tastes
include classical music. SO many miss out on that tremendous array of tone colors…
sometimes just because of the term. Important to expose kids early to classical as just
another (gorgeous) genre of music. PS I am pretty eclectic too and try new things often.
PPS Would love to have a bifrost for my Asgard…someday, lol.
32bit - 192000hz or 176400hz. For my purposes I chose the ladder. These are the only two sample rates than can be used to set the Bifrost into NOS mode (Disables its internal oversampling method)
AFAIK this is inaccurate, there is no way to disable oversampling on the Bifrost 2 and the 192/176k tip applies to the older model
You’re claim is accurate as not long ago I had already realized that this may have only applied to the original Bifrost, however my point still stands that these settings I played with do may a difference, albeit very minor. Whether or not you choose to believe that is your call. Only your ears can say on that one.
I know it does seem that way, that is basically why I called my self crazy. Especially with the fact that in Windows, so long as you aren’t using wasapi exclusive modes, you are still dealing with re sampling on top of what the bifrost may be doing on its own, and I was able to hear a difference between these different settings even when using wasapi and leaving it to the bifrost on its own. As far as this goes, I’m not knowledgeable on how these things work to decipher exactly why this is making a difference to my ears, and whether or not others can tell the difference as well and I’m not just crazy. Its of course, still up for debate. Even so, I’ve managed to dig up countless other forum posts on separate websites as well of users who evidently tried toying with the same things, even mistaking the Bifrost 2 for having NOS mode as well as the Denafrips Ares which I believe does not have NOS capability either, but they still claimed to hear a difference as well as I did. All and all I’m not willing to throw away the possibility that this could be a mass placebo effect if some people want to argue that point. Especially since the slight differences amount to as little as one might hear when swapping cables around, which is also highly debated in the audio realm as well. So far I am one of those stereotyped believers in the idea that cables actually do make a difference much the same as this idea with resampling, however nothing is factual and its all based on personal experience with what I heard, or at least thought I heard.
Well, I posted what I did above. I heard a difference; one that I actually preferred and still have running. I am not going to pretend to know the science behind the change! I changed a parameter that possibly allowed this DAC to process the signal in a slightly different way. Or it just changed something internal to the app. Either way, It’s actually pretty cool that you can do that.
The even cooler thing is that I can set it back to the way it was with just a few clicks. I do like Fubar. It’s been a great player for me. And of course, I love this Bifrost.
I was switching out headphones and noticed I wasn’t hearing as much as I should in my left ear… did a bunch of switching and swapping and ultimately realized the left balanced input on the Bifrost requires some “wiggling” or tension on it to get a 100% connection. I swapped the cables out to no available, it’s definitely the connection. Pretty bummed out about it as I love this thing, but ultimately decided to put in a work order for it and contacted Schiit.
Anyone have any experience with that? I know Schiit is slammed right now so I’m worried it will take forever.
Yeah, they are pretty much paralyzed right now. I eventually gave up on ever receiving the Jot 2 I ordered and went elsewhere. You never know though, the protocols might be different for repairs. Sorry to hear you are having issues. How old is the unit? I ask because I have a Bifrost 2 as well. Thanks.
Schiit is slammed for new orders. They should always have stock and resources available to customer support. My BF2 outright failed on me, and they had one shipped to me next day. As long as you work with their support channels, you should be just fine.
