Schiit (MultiBit) Bifrost

I also feel you too. Life is hard when you go back to sigma delta. It's like waking up from a wet dream.

 

Aye, and since I had the loaner before you, we're mourning the loss of the exact same piece of equipment!

 

I've been reading about these Blind DAC tests and how some DACS are better than others. SO maybe I am a newbiphile, but what makes one DAC better than another? The chip? Jitter? Convert your digital audio files with greater precision and accuracy? Better clarity, finer detail, and more sonic realism?

 

All of the above.

 

Good points about statistics and science in the last 2-3 pages. Some even arose from disagreeing posts. Here are a few somewhat unfocused (pre-coffee sorry!) comments based on my day job.

Don't get hung up on significance or especially - to give it its full handle - 'null hypothesis significance testing'. Many fields - my own (Psychology) among them - took a wrong turn with the idea that the likelihood or not of a sample finding can be used to confirm or disconfirm a hypothesis. This is your typical "5% threshold". Advice for the last 20 years has been to favor reporting in terms of other statistical measures, so downplaying the 'significance of significance'.

One of these measures was mentioned by Purrin. Simply, how big is the difference? Called "effect size", it is critical to determining sample size required to achieve a given 'statistical power', which is the measure of how 'likely' it is that a difference found is 'true'.

Unfortunately, (inferential) statistics is likelihood-based. Any adequately powered sample-based statistical finding is only likely to be true, which means possibly wrong. The one sure-fire way to know the truth (technically speaking, to "accurately measure the parameters of a population of interest") is to measure the entire population: 7 billion individuals indeed!

However, provided certain assumptions can be met it is likely we can estimate the population with a considerably smaller sample. With large differences and the badly named 'normal' population, a few score individuals/trials/whatever might suffice. With small differences sample sizes in the thousands might be needed - which incidentally has shut down some proposed studies I've advised about. (A pity because most proposed studies do ask scientifically useful questions).

Yet - circling back - the problem remains that scientific questions answered with statistical inferential methods are inherently uncertain and hence other epistemological and methodological principles are needed: the strength of the theoretical foundation which led to any particular study (but note that useful science never arises from safe, obvious questions!); study replication; and use of techniques like meta-analysis among them.

 

Absence of evidence isn't evidence of absence?

 

f**k me, just plugged in my Bifrost Multibit. I am convinced. I don't think I'll ever ABX these as I want to preserve my enjoyment (even if it's imaginary).

 

Good stuff! What DAC are you coming from?

 

Coming from a Modi 2 Uber.

 

There are some people who are very emotionally attached to the idea that $200 audio equipment and $2000 audio equipment "sound the same".and so they have "expectation bias" of sameness.

Despite that, you should find that the Multibit Schiit DACs have significantly less effect on the sound than sigma-delta DACs, and what remains is mostly the sound of the amp, speakers/headphones, and the recording.

"IMHO and I have not heard all other DACs"

 

Well it may in fact sound the same, but I'd really rather not know.

 

As I mentioned above, even the exact same DAC never sounds the same twice - due to air currents, temperature differentials, different placement of your head in the room, etc. etc.

The only question is "audibility". If you look at the experiences of people who became blind, it becomes clear that audibility is hugely variable.

 

Not sure what the point of this statement is.

The brain is very good at accomodating for those kinds of differences. Otherwise it would be impossible to recognize speech or faces, say if the sun were at a different angle or it was a chilly evening. In other words, it's possible to recognize something as the same (or different) even in the presence of apparently confounding factors.

 

The brain has evolved to be very good at recognizing speech or faces, so much so sometimes we see them where they don't exist like that thing on Mars or hear them when they aren't there like "satanic messages" on vinyl when you play it backwards. I don't think it pays to get that fixated on the kinds of minor differences mentioned but I wouldn't discount all environmental variables, e.g. I think the lower ambient noise level at night can make a pretty big difference.

 

Enjoy the Bifrost Multibit man! What are you running it with?

For me there were more than enough differences, both sighted and blind and whether or not I find my sample size statistically significant enough to be counted scientifically, is irrelevant. It's just amazing how much fun a DAC is bringing to the chain, something I never imagined.

 

Line up the tunes and just enjoy the music.

 

Anyone running a good USB - Spdif bridge into the Bifrost Multibit?

Got my bifrost in the house. Impressed so far and even more so running a little Halide bridge into the coax input. Not as thick sounding as straight USB. Could be my WW Starlight .3 meter cable is not up to snuff, but I think not. No glare to be found overall, yet some vocals come across a bit heavier than I would like. Only have 100 hours or so but on the usb input, snare drums, guitar strings have a bit of rounding off, compared to the bridge.

Nice overall.

 

What is the consensus on the break-in time for the Bifrost Multibit ?
The reason I ask is, when I first got my Bifrost Multibit I did some A/B testing with my existing DAC and I didn't perceive any significant difference. I pretty much declared myself as DAC deaf at that point. But in the last couple days the Bifrost Multibit has blossomed and it's absolute treat to listen to, I don't even need to A/B again (I will anyway, but for now just enjoying my new toy).

Note:
1. I've no record of how long I've had the Bifrost Multibit running.
2. I only play music on them during the time I am at home. Which during the weekdays is at best 4-5 hours.

 

For me I noticed a difference after 24 hours.
YMMV

 

4 hours+ I find it very thin and almost boring in the first few hours. This was actually funny as my friend almost turned it off when we were doing the blind tests because he thought I was joking that this was a masterpiece. Gradually but surely, the sound came to be. Mike commented that 4 hour is the warm-up time, I feel that this was not in any way an exaggeration. YMMV.

24 hours I feel are the best results but this I can't back up with anything more than a hunch. This is not break-in time however, literally just keeping the DAC on.