Schiit (MultiBit) Bifrost

Between subtle things. Between obvious night-and-day things, it's immediately super-clear exactly what differences you're hearing. If you hot-swap a DT-880 and HD-650 in quick succession, there'll never be the slightest doubt in your mind that you're hearing differences.

 

Agreed, but it can sure be hard even then to capture the differences with a hot swap.

But yes, transducers should be the ones that sound the most different from each other.

 

I'm going to jump right in between with some impressions. I had already auditioned Yggdrasil (it might be a bit too much for me listening laid back) and Gungnir Multibit before anything Bifrost Multibit and it was actually @Bill-P's impressions that made me jump on the Bifrost Multibit (figured I can send it back if it doesn't fit the bill).

Currently Bifrost Multibit > HDVA600 > HD800. This is absolutely insane, I'm hearing so many incredible nuances I didn't even know the HDVA600 was capable of showing. The amount of details paired with the articulation of each note is just superb... I've been going through my entire library since.

I sold my Hugo this weekend and honestly I don't think I need anything more for the time being. This right here is what I'm looking for, highly detailed, great separation but still having that almost romantic sense of musical experience. It's a bit vinyl like to be honest. In fact I directly compared it to my BeoCenter 4600 and SL1200 rigs and feel it fits the bat, with more micro details.

I've done a few direct DAC comparisons with a friend, the results so far:
Hugo vs. Bifrost Multibit: Bifrost Multibit, blind and sighted for both of us. Also 7/10 for my friend and 8/10 for myself in AB testing.
DAC1 vs. Bifrost Multibit: Bifrost Multibit, again no doubt. 8/10 for my friend and 8/10 for myself in the AB.
ODAC vs. Bifrost Multibit: Bifrost Multibit, no doubts here. 7/10 for my friend and 9/10 for myself in the AB.
Modi 2 vs. Bifrost Multibit: Bifrost Multibit, again this was clear cut. 7/10 for my and 7/10 for myself in the AB.

Next up on the comparison list: Lynx, QB9, La Voce and the Devialet 200 with SAM unit (currently not plugged in due to housing circumstances, will be back up when I move out to a new place in a few months). Side the Hugo, all other DACs will be used in the tests as well, DAC1 and Modi2 are loaners of friends so we can do some direct comparisons. Tonight is actually LA Voce's turn, I'll be sure to update, especially since the La Voce is R-2R based (Burr, Philips and another DAC that you can pick from there) and we will use both the Cizek KA-I's with the Son of Beast amp (exceptionally musical pairing might I add for speaker lovers) and my HD800 + HDVA600.

Also thanks to Bifrost Multibit, I'm gonna get a Valhalla 2 just for some extra tube joy and be done with it for a while. Mabye I will sell the HDVA600 and get the EC BW when it's out but nothing more. It's been an expensive year as it is haha.

 

Mine's only been warming up for 2 hours so far, but ten minutes into my Bifrost Multibit > Project Polaris > HD650 experience and I've got a big, shit-eating grin. I really didn't expect such a noticeable upgrade from my GOV2.

 

Volume matched? That's really impressive.

 

I measured the Bifrost Multibit DAC out to be about 0.8 db louder than my ODAC at the analog output. I found this out by easily AB identifying the Bifrost Multibit as sounding better until I figured out this mismatch. So yea, careful volume matching is necessary. Now that I have it matched (in analog domain to not touch the bits), I can tell this is going to be *very* difficult.

 

We did as far as I know, however @paranoidroid is giving me a bit of doubt on the ODAC part. I always try to do multiple series of tests, sighted, blind and lastly volume matched. Also first off extended periods of listening, as @nedifer already noted, I too agree with Mike on this. Don't just swap around before getting inside the gear, listen to it, experience it and it will be much easier to identify, IME. Also I find most D/S DACs very hard to distinguish between, much harder than this.

This I did not know, I will try again soon. Thanks for that heads-up! I found it the easiest to distinguish between but this might be the reason. I'm not planning on selling my ODAC so that's always testable.

As promised, the next match-up: Aqua Hifi La Voce vs. Bifrost Multibit. This was very interesting! We did blind tests and sighted tests but not volume matched (time constraints, it's a weekday after all).
First off simply listening to the La Voce is a real joy, it's smooth, it's warm and intimate. What it lacks in micro details, it makes up with presenting music very cohesively, everything feels like it's one and very natural, "fuller" if that makes sense. On the other hand Bifrost Multibit, gives much better articulation on every note, soundstage and resolution are much bigger, the separation and presentation are much wider and a certain satisfying slam. This was a very hard match-up to pick from, each has its own special traits.
For me personally the Bifrost Multibit would win, I only had 1 song and a few parts of some songs that the La Voce would win. My friend was first leaning towards the La Voce but in the end mostly leaning towards the Bifrost Multibit (warm-up time (it was off for 30 minutes from my house to his)? songs used? getting used to the sound? many factors here to be fair), especially on the headphone rig. Blind was much harder to tell, it was mostly sense of stage, micro details etc. I'm gonna bet volume matched will be a different beast entirely and very hard picking out the minor details. One thing that was great to see is this friend of mine, just like the other one started with being wowed and ended with being wowed. This I think is the best part here, we're talking about an MSRP €2180 DAC vs. €630 DAC. @schiit and @baldr, honestly kudos, the value presented here is just amazing on all counts.

Another note that was interesting, both pairings made the HD800, without mods (I took them out for the test), much, much more enjoyable. Seems that R-2R DACs work well in general in that area...

So far Bifrost Multibit has, in my humble opinion, outclassed every match-up, while only 2 out of 5 other DACs were below its price point. When more match-ups have taken place I will update accordingly for anyone interested in these impressions.

Sources used:
Qobuz through Wifi via iPad + Apple streamer
Deezer through USB
Lossless tracks through USB via laptop
Gear used:
Cizek KA-Is paired with Transcendent Sound Son of Beast
HD800 (modded and without mods for testing) and Shure 1540 paired with the HDVA600

Prominent songs that we used for the blind and critical tests:
Michael Jackson - Baby Be Mine
Andreas Vollenweider - The Glass Hall
Fabrizio de Andre ft. Mina - La Canzone di Marinella
Dire Straits - Sultans of Swing
Dire Straits - Your Latest Trick
Mahsa Vahdat & Mighty Sam McClain - A Deeper Sense of Longing
Mahsa Vahdat & Mighty Sam McClain - Silent Song
Raul Midon - State of Mind
Marcus Miller - Blast
Tracy Chapman - Fast Car

Sorry for the long wind up... I'm just having a ton of fun with a piece of gear that I never expected to give me such an experience. Especially since I paid € 870 more for the Hugo which I expected to be "never needing to upgrade" level...

 

Cizek speakers, oh my! I used to sell his larger bookshelf back in the late 70s. Very well done.

 

I graph a frequency sweep through both dacs to volume match w/in 0.1db, but an easy way to do it is also to play some tone through both dacs, then put the right channel of one dac and left channel of another dac through the amp and volume match until the tone sounds perfectly centered through both speakers. It's interesting that the default 0.8db louder of the Bifrost was not detectable as 'louder' but simply sounding better (better imaging, attack, etc).

 

Oh yes and to think I never had heard these until recently. This friend of mine has two sets one Italian and one American one. Funnily I prefer the American one and he prefers the Italian one (him actually being Italian should give a slight bias though haha). But man the resolution on these things, holy damn these scale well with gear. Quite the exceptional pair, well worth the amount of hype and writings on them.

I think this was my main thing at some point it seemed perfectly matched but in my mind the Bifrost Multibit simply excelled at everything, very easily distinguishable. Minor mistakes in this can have serious impact which is why it's hard to conduct perfect tests without measuring. Looking forward to your impressions.

 

In the whole world of people who listen to music, there is the conventional wisdom that " I should be able to listen as well as anyone ", which like almost all conventional wisdom, is false.

No one says " I should be able to run 100 yards as well as Usain Bolt" or "I should be able to do open heart surgery".

One's ability to discern audio differences is proportional to the amount of time one has spent with critical listening. It's a skill that requires training.

 

Sorta On-Topic:
After reading Tyll's Big Sound articles and listener experiences, I had a chance to blind test with & without the USB Regen in a system loop.
What I stumbled upon was a method of concentrating on differences only at first..not trying to make a value judgment simultaneously.
Step One: Simply trying to perceive any differences.
I can always make a subjective determination ("like it or don't like it") later as a second step.

On-Topic:
I'm here because Schiit Multi-bit DACs are so good with my tube amps.
My PWD is great with my BW. But I am listening to my Original ZD with 800 and thinking about Bifrost Multibit again.
Issues? Treble grain, roughness.
Want to keep the 800 treble with the BillMods but need to work on the treble or keep the 800 on the BW.
ZD headstage is more expansive so that's the pull...however, PWD is more forward than the Bifrost Multibit in the vocals/center stage.
Wish I tried this pairing when I had the loaner!

 

I spent four full days with the Bifrost Multibit and got to do some careful comparisons with my ODAC. The setup is OSX->USB->Bifrost Multibit->Mjonlir 2 (on loan)->HD800/HE1000. I work from home so I get to spend all day and night listening to music through my headphones. I spent the first afternoon mostly listening to the setup and getting used to the sound and letting it get warmed up. Late evening, I did some volume matching between the ODAC & Bifrost Multibit by measuring frequency response sweeps and getting volume match roughly within 0.1db using a pair of analog attenuators. By default the Bifrost Multibit was 0.8db louder than the ODAC and this had the effect of easily identifying which was which by simply what sounded better overall so this volume matching was critical to a fair comparison.

In OSX I aggregate both DACs into one output device so both DACS would play whatever I was listening to. At this point it took me a second or two to switch between DACs via cables, and I was feeling confident I could tell them apart. The Bifrost Multibit sounded clearly different. The impressions held over into the next day. I wrote a few days ago:

"So far, the Bifrost Multibit seems to have a bit more lush bass, a bit cleaner and more incisive attack in the sound as well as more depth. The most noticeable difference is the hint of extra bass / warmth."

I'm at my personal end-game when it comes to headphones with the HD800 & HE1000 so I was really looking forward to find improvements in the rest of my chain. I also wanted to see if I could blind A/B them apart so I ordered a switch box feeling decently confident I could pick out the Bifrost Multibit. I had a few sections of certain tracks that brought out the differences so I could compare. My listening tastes run the full gamut, and for testing I used a selection of high production electronic, acoustic/instrumental, high dynamic range orchestral and female vocals.

The next day the switch box came, hooked things up and started testing. By now the Bifrost Multibit has been on for over 2 full days. The switch box gave me a near instant switching time, and I was a honestly a bit let down to discover all the audible differences I noticed before were gone - this was going to be very tough to pick out which was which blind. I spent that evening and the next afternoon A/Bing at my leisure. I used both the HD800 & HE1000 and mostly focused on listening to the whole of the sound and if there were any differences - not which was which DAC or which was better. Bass, detail, imaging/depth/width, texture, tone was just indistinguishable to the point I had to check that the switch box was working correctly. I removed the switch box to make sure that wasn't effecting the sound and it didn't - both DACs just sounded great either way through the Mjolnir and HD800/HE1000. I did 20 runs over many hours where I'd spend as much time as I want going between DACs and different tracks and pick out the Bifrost Multibit. Didn't score better than random. In fact I noticed a funny bias - once I did identify the Bifrost Multibit, it actually did seem to sound better. Almost like there's some quantum mechanical effect of the DACs being unidentifiable in sound until the probability function of which is which collapses. Then, characteristics of how I noted the Bifrost Multibit to sound show up. But once randomized, they seemed indistinguishable - even paying particular focus to imaging (timing advantage of Bimbo filter). From a lot of experience trying to identify high bitrate mp3/vs. FLAC I knew sometimes you have to just turn the logical part of the brain off and go by how you felt about the sound overall. That seemed to fail me as well. I would definitely note one side of the A/B feeling better in sound sometimes, but it'd end up no better than random on which DAC it was.

At this point, I just un did everything and spent the rest of the night listening to the Bifrost Multibit only and enjoying my music while I worked. The next day, I was determined to give A/B testing another try and hooked things back up and volume matched. I spent a few hours working and just switching between DACs every 15 or 20 minutes. Finally I came across a section of a track where there seemed to be a very slight difference. It has some very strong high hats and one of the DACs just felt a tad sharper, distinctive & with more energy and the other a bit smoother. Identified the sharper one as the ODAC, then used that same section run a 10 trial A/B. The difference was so slight, the minute I actually focused on the high hat it was hard to distinguish between the two but if I just let myself go by gut on which felt a bit smoother I was able to identify it. Scored 10/10. A this point, despite how slight the difference was I was happy to be able to just reliably distinguish between the two. I went back and checked volume matching and it looked very good, about 0.1db or so. For good measure I redid the volume matching to try to get it even closer. With nothing other than that changed, I ran through another 10 trials and it was clear it was going to be much harder. That very slight difference I noticed seem to disappear, but I ran through the trials and now scored 2/10. In other words the DAC where the high hats on that particular section of track felt smoother seemed to switch places just by a slight tweak in volume matching. Sigh..

So here are my personal conclusions:

- I have tin ears / am DAC deaf. I love reading the implementation details of the Schiit Multibit DACs and appreciate the engineering behind them as well as the technical minutia of how the D/A conversion is done vs. various implementations of delta-sigma methods. However despite the true appreciation of the engineering behind it, the ODAC Sabre delta-sigma tech ends up being good enough / indistinguishable to me from the Bifrost Multibit multibit. Note that my hearing goes up to 16khz so any difference beyond that point in FR is lost.

- The A/B switch box or something else in the chain is degrading signal to where it erases audible differences between DACs. Very possible, but not one I felt is happening since the sound subjectively sounded just as great with the switch box in or out of the chain. Neither DACs ever felt handicapped, both sounded great that it's hard to imagine how it could be better. Still, I put this as a possibility.

- The smallest variations in volume matching can eventually lead you to identify one DAC (even 0.1db or less seemed to let you pick one if you happened on a particular section of music that'd bring it out) - but despite eventually picking one, it wasn't subjectively better.

- I also noticed the output rate of the DAC effected the frequency response graph a bit, so along with volume matching, making sure each DAC was set to output the same output rate was important.

- The Bifrost Multibit sounds freaking great. But so does the ODAC. If I didn't already have the ODAC I might pay extra for the Bifrost just for aesthetics and engineering implementation alone, DAC deaf be damned. Even if I can't hear it, I'm convinced the Schiit Multibit implementation is technically superior for rendering the digital to analog signal in a truer to source form.

Perhaps I'll need to go further up the DAC chain to really get a discernable difference with my ODAC and I look forward to trying them, but for now the ODAC sabre delta-sigma tech works for me

 

The differences are greatest with excellent recordings that are "minimally miked". People tend to think that audio electronics differ in sound of a particular instrument (which is what one might find with headphones or speakers), but in actuality, most of the differences are in the shapes and positions of the instruments, and things like the room sound (the room of the recording, not your room).

Atomicbob - who has extensively heard all the Schiit Multibit DACs describes the above in colorful terminology:
" 1. Best of D-S Dacs - an indistinct sound is barely audible in the recording that appears to not be musically related to the orchestral piece playing.
2. Bifrost MB - someone in the Second Violin section has flatulated
3. Gungnir MB - the flatulent sound originates with the 2nd chair in the 2nd row of the 2nd violins
4. Yggdrasil - that 2nd chair, 2nd row, 2nd violinist has let loose with a Db when clearly the piece being played is in the key of G."

There are many people who feel that the A/B process destroys perception of sound quality, because the latter is a mental process.

 

Wow - I've never come across anyone else who uses that track and album. I've been using it ever since it was released, I used to carry it around to audio salons. I use the first 3 tracks of the album. The part of the first track that has oars dipping in water ("The White Boat") is often a particularly good test.

 

I mean, at this point your conclusions are exactly the same as anyone else's -- you hear more warmth and resolution when listening sighted, and you can tell differences when listening non-level-matched blind. So it's really just that nobody else has tried a level-matched blind test.

 

Personally I always figured the difference would be subtle and in the realm of position/imaging/ambience, etc. I was surprised to hear a distinctly notable difference at first but I figured multibit rendered from the bits in enough of a different way that I accepted that difference. One thing I noticed when trying to do abx testing on http://abx.digitalfeed.net is that for the high bitrate stuff where it's very close to transparent from source, the harder you try to focus on distinguishing differences the less you're able to - because the differences are beyond the realm you're able to logically discern and into the realm where it's about feel. E.g. 'can you image the size of the room, subtle rendering of the reverb, how realistic does it feel, how separated are the layers'. It can come down to very emotive subjective qualities and in those cases I can do better by not listening for specific artifacts but by going by gut. It's also why I when doing my A/B of the DACs I also used the technique of just letting my mind wander and only switch every 15-20m to let any differences in the emotive qualities to accumulate before switching.

Funny, when I was listening to a bunch of orchestral works I had the exact description Atomicbob wrote in my thoughts, but if there were differences even 1/20th of what was described I'm personally confident I would have heard it.

 

I think one of the most interesting things I learned out of this was I could reliably (10/10 trials) discern the DACs if volume matching was off by just ~0.1db (in one particular section of a particular track and focused on one specific thing). Getting it matched better than that is hard when turning some analog potentiometers even when I'm objectively measuring the output.

 

The following post by Schiit DAC designer Mike Moffat (Baldr) is relevant:

"Elsewhere posted but very germain:

Back in the early 1970s, before I founded Theta Electronics, the tube audio products company, I had a busy part time biz rebuilding Dynaco Tube Amplifiers. At that time I had converted to the tube based practice for my own system, convinced that tubes sounded better than the solid state gear of that era. In my ramblings, I met John Koval, a man who had designed a modification for the old Qual ESL loudspeakers which made them sound much better. “The mod gets rid of a 5 db bump in the 200-400 Hz region which makes them much flatter” he explained. I told him that I was enchanted with the sound of tube amplifiers and preamplifiers. He explained that as long as the frequency response was the same and the levels were precisely matched, there was no way anyone could tell any amps/preamps apart in blind A/B tests. He had built a custom box that matched levels and randomized any two amplifiers or preamplifiers with a pushbutton to switch between them. Bullschiit, I thought, what about the solid state A/B box and its sonic signature.

Intrigued, I built a similar box with passive relays and a passive attenuator. Damn, if he wasn't right. It is really difficult to tell differences in an instantaneous blind A/B test between tube gear that I built versus some commercial gear that I was not particularly fond of. I used to bet John beers that I could tell the difference. Usually, I won at 7 out of 10 picks or so – the best I ever did was 9 out of ten. But it was really hard.

This whole deal made me wonder if I was crazy hearing differences between amps. If what John said was true, and many others have said in the passing 40 years or so, there is no point for an audio hobby involving anything other than transducers. WTF?

So I tried something new – I still did the A/B tests, matched levels, but allowed long-term listening to each; at least an hour or two with known recordings. Guess what! Suddenly I knew which was what. I tried it out on John B and Mike and Dave and all my other audio buddies. They called it too – tubes vs a bad solid state preamp. Every friggin' time. My enthusiasm had returned. This taught me that the human ear is an integral, NOT differential device.

So much for the blind A/B instantaneous naysayers. All that matters is frequency response, they say. People can't hear anything much above 20KHz in their prime, less later. The ear has a short memory, it is all bias, blah, blah. They should take up a different hobby, say stamp collecting.

Thanks to Dr. Heil, the inventor of the Heil AMT speaker who shared this experiment with me over 40 years ago, Consider this: I am 67 years old – my high end extends to just under 15KHz (not bad for and old fart). I can play back two pulses 200 microseconds in length separated by 20 microseconds and clearly hear two pulses. Not unusual until one considers that 20 microseconds corresponds to a square wave of 50KHz. And then, there is the time domain – home of spatial cues which audio measurement traditionalists ignore. I believe that in the quest for the best sound, an open mind is the most important asset. I will even listen to cables, even though I believe in my heart that all technology about cables is well known. Who knows, even an old fart like me could be surprised.

Until then, yet another retelling of my old John Koval saga is 40 year old news to me."

 

I can relate to that post a lot. I found that the differences were far more discernable through extensive listening (and then confirmed by switching back and listening extensively again via the other device). Quick A/B testing just doesn't work for my ears.

But that's just my own experience.