[motberg]

Hahaha, yes. Confusing my acronyms here I guess.

To continue my impressions: I know burn-in is a controversial subject in our circles but god damn if BF2 couldn't make for a good study subject on the matter; I swear to God, the BF2's changed after leaving it on overnight. Mids are much more present, the highs have mellowed very nicely, and bass has deepened. All that resolution is still there. All my misgivings that had me thinking I may stick with the BFM are kind of melting away. This is just so clean, and so detailed, yet so musical.

Thanks - which input are you using ?

 

[Ripper2860]

Hahaha, yes. Confusing my acronyms here I guess.

To continue my impressions: I know burn-in is a controversial subject in our circles but god damn if BF2 couldn't make for a good study subject on the matter; I swear to God, the BF2's changed after leaving it on overnight. Mids are much more present, the highs have mellowed very nicely, and bass has deepened. All that resolution is still there. All my misgivings that had me thinking I may stick with the BFM are kind of melting away. This is just so clean, and so detailed, yet so musical.

I am a firm believer in electronics burn-in and am anxious to see if I experience the same. (My BF2 shipped today, so hopefully soon! Woohoo!!!).

 

[bigbenrfan99]

2 days until my Bifrost 2 arrives

 

[tincanear]

I am a firm believer in electronics burn-in and am anxious to see if I experience the same. (My BF2 shipped today, so hopefully soon! Woohoo!!!).

Overnight burn-in probably makes a difference for the electrolytic capacitors. Capacitor manufacturer's data sheets spec the leakage current, ESR, etc after 100 hours applied working (DC) voltage. If you measure brand-new caps often they have a small negative voltage (~100mV) across the terminals. the oxide layer needs some time with operating voltage applied to properly form.

Would be curious to know not only the input used (e.g. USB, coaxial SPDIF), but also the raw data rate (e.g. 16/44 or 24/96) supplied by the source. Bifrost 2 has 8x oversampling comboburrito filter vs the 4x on original Bifrost.

 

[Alcophone]

On Steve Hoffman was a post by a guy who used a BFM demo unit for awhile and then got a brand-new one and he even noticed the sound was different between the "burnt-in" unit and the brand new one. I don't personally subscribe to the idea of units needing 500 hours to sound their best but it doesn't seem completely unfeasible to me that straight from the assembly line their components might change slightly with use.

On another forum it's not just considered established fact that the Gungnir Multibit has received a silent Analog 2 update, but that the Bifrost Multibit also has seen a change that supposedly makes it sound significantly closer to a Gungnir Multibit (Analog 1, presumably).

Call me extreme, but to confirm burn-in as the reason for sound differences, I would buy two new units at the same time, open them up to confirm there are absolutely no visible differences (Schiit silently sold a few Yggdrasil Analog 2s at the old price before announcing the change, so it's conceivable that someone ordering two units may receive one older and one newer revision), confirm that the sound is the same initially (to account for good parts/bad parts, differences in the burn-in time at Schiit HQ, etc.), then use one for, say, 200+ hours with the other one sitting in a box, then turn the new one off and let it cool down for, say 4+ hours, then turn both units on at the same time, ideally with a Y splitter for power and identical power cables. If I then hear profound differences, going back and forth with the exact same cables, transducers, etc., I would be relatively certain it's burn-in (vs. warm up, silent revisions, parts differences, etc.)

 

[Rensek]

On another forum it's not just considered established fact that the Gungnir Multibit has received a silent Analog 2 update, but that the Bifrost Multibit also has seen a change that supposedly makes it sound significantly closer to a Gungnir Multibit (Analog 1, presumably).

Call me extreme, but to confirm burn-in as the reason for sound differences, I would buy two new units at the same time, open them up to confirm there are absolutely no visible differences (Schiit silently sold a few Yggdrasil Analog 2s at the old price before announcing the change, so it's conceivable that someone ordering two units may receive one older and one newer revision), confirm that the sound is the same initially (to account for good parts/bad parts, differences in the burn-in time at Schiit HQ, etc.), then use one for, say, 200+ hours with the other one sitting in a box, then turn the new one off and let it cool down for, say 4+ hours, then turn both units on at the same time, ideally with a Y splitter for power and identical power cables. If I then hear profound differences, going back and forth with the exact same cables, transducers, etc., I would be relatively certain it's burn-in (vs. warm up, silent revisions, parts differences, etc.)

I believe there is only one version of Bifrost 2 sold and out in the wild. I believe you are referring to the posts that said Bifrost 2 sounds very close to Gungnir Multibit analogue 1 (Gungnir analogue 2 is an unconfirmed rumor, but taken as gospel elsewhere). Bifrost 2 is apparently a huge change from bimby (Bifrost multibit v1). Relatively speaking in audio terms, of course.

 

[tincanear]

On the main Schiit thread here on head-Fi, Jason talked about the design discussion he had with Mike, and the following items were detailed about Bifrost 2 (using some tech borrowed from gungnir) :
uses 18-bit AD converters AD5781A (spec'd linearity to 4 LSB, typ 2 LSB) vs 16-bit parts used on Bifrost Multibit (1st gen)
uses differential amp LME49724 for both SE and XLR outputs (Bifrost Multibit OG uses AD8512 amp chip or something like that),
uses one 18-bit converter per channel (gungnir uses two of higher-linearity-spec AD5781 parts per channel)
uses 8X oversampling digital filter on SHARC DSP (OG Bifrost Multibit is 4X)
Unison USB chip with added electrostatic and electromagnetic isolation.

 

[Alcophone]

I believe there is only one version of Bifrost 2 sold and out in the wild. I believe you are referring to the posts that said Bifrost 2 sounds very close to Gungnir Multibit analogue 1 (Gungnir analogue 2 is an unconfirmed rumor, but taken as gospel elsewhere). Bifrost 2 is apparently a huge change from bimby (Bifrost multibit v1). Relatively speaking in audio terms, of course.

I understood the post I was quoting to mean that someone bought a demo Bifrost Multibit (Bifrost 1) and compared it to a brand new Bifrost Multibit (also Bifrost 1).
I didn't mean to suggest there's more than one (working) version of the Bifrost 2 out there.

Edit: And the comparisons to Gungnir Multibit I saw long before Bifrost 2 was out.

 

[Rensek]

I understood the post I was quoting to mean that someone bought a demo Bifrost Multibit (Bifrost 1) and compared it to a brand new Bifrost Multibit (also Bifrost 1).
I didn't mean to suggest there's more than one (working) version of the Bifrost 2 out there.

Edit: And the comparisons to Gungnir Multibit I saw long before Bifrost 2 was out.

Ahh I guess I don't know about that. I haven't followed Bifrost 1 multibit very closely. Schiit has been known to do silent revisions previously, see Jason's slouching towards kaizen chapter. It's also entirely possible the threads talking about a Bifrost 1 being close to Gungnir were Bifrost 2 beta testers having private conversations in public, and not being able to talk about it. Or maybe the DAC/output stages used in Bifrost 2 was originally slated as Bifrost 1 upgrade, before it became the first autonomy product. Someone knows, but I doubt we will ever find out.

Has Anyone purchased a Bifrost multibit upgrade recently, that would be willing to open the case up and take a look? Probably the only way to find out.

 

[Currawong]

On another forum it's not just considered established fact that the Gungnir Multibit has received a silent Analog 2 update, but that the Bifrost Multibit also has seen a change that supposedly makes it sound significantly closer to a Gungnir Multibit (Analog 1, presumably).

I've not asked, but it wouldn't surprise me, considering that the original Gungnir MB would have had the same glitch error as the Yggdrasil.

Call me extreme, but to confirm burn-in as the reason for sound differences, I would buy two new units at the same time, open them up to confirm there are absolutely no visible differences (Schiit silently sold a few Yggdrasil Analog 2s at the old price before announcing the change, so it's conceivable that someone ordering two units may receive one older and one newer revision), confirm that the sound is the same initially (to account for good parts/bad parts, differences in the burn-in time at Schiit HQ, etc.), then use one for, say, 200+ hours with the other one sitting in a box, then turn the new one off and let it cool down for, say 4+ hours, then turn both units on at the same time, ideally with a Y splitter for power and identical power cables. If I then hear profound differences, going back and forth with the exact same cables, transducers, etc., I would be relatively certain it's burn-in (vs. warm up, silent revisions, parts differences, etc.)

I was A/B'ing the Yggdrasil (V1) against a Chord Mojo when I first bought mine. At the very start, they sounded about the same, but the Yggdrasil, over the period of about a week, definitely pulled ahead. Given that resistors measurably change with use, as a result of heat, enough to cause audible changes in an analog circuit, and that the DAC chips are ultra-precise resistor arrays that are known to be affected by use, something similar isn't so surprising to me at least.

 

[Alcophone]

I was A/B'ing the Yggdrasil (V1) against a Chord Mojo when I first bought mine. At the very start, they sounded about the same, but the Yggdrasil, over the period of about a week, definitely pulled ahead. Given that resistors measurably change with use, as a result of heat, enough to cause audible changes in an analog circuit, and that the DAC chips are ultra-precise resistor arrays that are known to be affected by use, something similar isn't so surprising to me at least.

Definitely possible. Still, did you leave the Yggdrasil on continuously? If not, and it still pulled ahead, then burn-in is definitely a strong contender (as is you getting accustomed to the differences), otherwise it may just be the known needed warmup time (which I have experienced as well with the A2).

Both Bifrost 2s I've had (third replacement due today as of 25 minutes ago) definitely sounded noticeably better after the first few hours or so, but since the first one didn't survive the first night and since I only turned off the second one for the firmware update and it started showing issues the next day, I can't say whether that's burn-in or warmup time.

I'm not ruling out burn-in as the reason for the new Bifrost Multibit sounding different than the older demo unit. Just saying that it could be burn-in, warmup, silent revisions, different quality parts, etc. We can't be sure burn-in of the existing unit is the (only) reason unless we learn a whole lot more about the units at hand. It's just an easy guess if you assume a unit built much later is identical, which is definitely something I naively assumed when I got into the hobby.

But given what I learned in the meantime, I would hesitate to get a second Vidar for monoblock use, for instance. I'd probably get two new ones instead and repurpose or sell the two year old one. I even asked Schiit whether I should expect a new one to sound the same given that silent revisions are (were) a thing, and was told "I cannot speculate on differences between products or product runs", followed by a pointer to the "Slouching Towards Kaizen" chapter.

 

[Currawong]

I left the Yggdrasil on continuously. I've just switched the Bifrost 2 on again now and had a bit of a listen. I'll see if I notice anything after a few hours. It's plugged into an amp though that is also affected by being left on a few hours though.

 

[schneller]

So Bifrost2 sounds like Gungnir Multibit (v1). The unofficial Gungnir Multibit v2 sounds like Yiggy v1. Yiggy A2 has evolved beyond v1. What will the real Gungnir2 sound like? You know it's coming...

 

[knivek]

So I am just curious, maybe it's just me... but I was watching the Transformers movies again now that my daughters being really into it lately. When the orchestra and the strings pick up in some of the sappy areas of the movies, it sounds really really artificial. So I tested my headphones (AFC's) on another source. Sounds totally different.

This made me try some classical music.. same kind of thing. I have 0 EQ settings. Windows settings are set to 24 bit/48000.

Is this something I should worry about?

 

[Rensek]

It's all one big circle of dogs chasing their tail. I read that Gungnir Multibit v1 sounded so good they made yggy A2 sound like gumby A1, and Gumby A2 sounds closer to yggy A1, which would mean Bifrost 2 is close to Gumby A1 and Yggy A2??? I'm sure they all sound great. Buy the DAC you want/can afford, and if you like how it sounds, be happy!