[barbz127]

I see the website says 6-9, is this week's? It's not showing anything more for me.

And would anyone have any issues buying a second hand bifrost 2 in lieu of available new stock?

Thankyou

 

[rkw]

I see the website says 6-9, is this week's? It's not showing anything more for me.

I think they made a mistake while updating the website. You should email info@schiit.com and ask for a clarification.

 

[rsnblmn]

I see the website says 6-9, is this week's? It's not showing anything more for me.

I'm pretty sure this means that they hope to have Bifrost 2 in stock by June 9th, so about three weeks from now, I guess.

Fingers crossed that it can be a Father's Day present to myself, depending on how many backorders deep I am.

 

[Sean_MR]

Trying to understand the technology a little better! Since this has 18 bits of converting power, does that mean it “works best” with 16-bit songs? When you play a 24-bit song, what “happens” to the other 6 bits? Probably dumb questions but trying to learn

 

[XERO1]

Trying to understand the technology a little better! Since this has 18 bits of converting power, does that mean it “works best” with 16-bit songs? When you play a 24-bit song, what “happens” to the other 6 bits? Probably dumb questions but trying to learn

Bit depth is a very misunderstood part of digital encoding/decoding. The primary reason for needing more than 16-bit word-lengths was that professional recording studios often use dozens of (sometimes well over 100) individual tracks to create a mix, and as those tracks begin to be layered on top of each other, the combined quantization noise of all of the individual 16-bit tracks becomes more and more audible. With 24-bit depth, the quatization noise floor is reduced by 48 dBs, so this gives a huge improvement in the noise floor of the final mix if the track count is really high. But once the final mix has been created, it can then be truncated down to 16 bits with almost zero audible loss when compared to the original 24-bit master.

Between sample rate and bit depth, sample rate has by far the most siginicant impact on overall SQ.

Here is a great, easy to understand video that helps explain the audible significance of bit depth.



As far as what happens to the extra bits, IIRC, Make Moffat was asked this same question on Head-Fi and I believe he said that the lowest LSB’s (Least Significant Bit) of a 24-bit track were not just thrown away, but were averaged together. I'm fuzzy on the details about how this is done, so maybe someone else can comment about this.

 

[Sean_MR]

Bit depth is a very misunderstood part of digital encoding/decoding. The main reason for needing more than 16 bits was that professional recording studios often use dozens (sometimes well over 100) of individual tracks to create a mix, and as those tracks begin to be layered on top of each other, the quantization noise floor of each of the 16-bit tracks becomes more and more audible. With 24-bit depth, the quatization noise floor is reduced by 48 dBs, so this gives a huge improvement in the noise floor of the final mix if the track count is really high. But once the final mix has been created, it can be down-sampled to 16 bits with almost zero audible loss when compared to the 24-bit master mix.

Between sample rate and bit depth, sample rate has by far the most siginicant impact on overall SQ.

Here is a great, easy to understand video that helps explain the audible significance of bit depth.



As far as what happens to the extra bits, IIRC, Make Moffat was asked this same question on Head-Fi and I believe he said that the bottom LSBs (Least Significant Bit) of a 24-bit track were not just thrown away, but were averaged out. I'm fuzzy on the details about this so maybe someone else can comment about this.

Thank you for all the information! I’ve added that video to my watch list and am definitely interested in getting to it. So if I interpreted correctly, the Bifrost 2 won’t necessarily “work better” with 16-bit songs, since 24-bits is hardly audibly better anyways?

 

[XERO1]

Thank you for all the information! I’ve added that video to my watch list and am definitely interested in getting to it. So if I interpreted correctly, the Bifrost 2 won’t necessarily “work better” with 16-bit songs, since 24-bits is hardly audibly better anyways?

Pretty much.

 

[ev666il]

Bit depth is a very misunderstood part of digital encoding/decoding.

[...]

Between sample rate and bit depth, sample rate has by far the most siginicant impact on overall SQ.

I agree 100% about bit depth. When it comes to digital audio, the word "resolution" is not a synonym to—nor does it imply the concept of—"definition" in the same way it does in the video world. This is a common misconception, and one "high-resolution" music vendors are happy to perpetuate and exploit.

Here's another easy-to-understand video that explains bit depth and performs a classic null test between the 24-bit and 16-bit version of the same audio file:



I do not, however, fully agree with your sentiment on sample rate. 44.1Khz is more than enough to capture the full band of frequencies the human ear can perceive—and indeed more, depending on your age and general hearing. I, for example, can't hear beyond 16Khz (and my hearing between 15.5Khz and 16Khz is such that those frequencies are easily masked in the presence of literally any other sound.) Higher sampling rates can be useful during mixing and mastering, but they won't make an audible difference when it comes to playback. Here's a video that explores the subject:




Either way, I own 24-bit FLAC files and the Bifrost 2 handles them with ease

 

[XERO1]

The reason why I said that sample rate is far more important than bit depth was not because of it's effect on a recording's frequency response (although this is important too, but to a much lesser degree). The main benefit of a higher sample rate is that it can track an acoustic waveform's impulse response much more accurately than a lower sample rate can.

This is the main reason why 2.82MHz DSD is considered to sound closer to the sound of analog tape than 192KHz PCM (DSD has a nearly 15x higher SR than 192K PCM), and why DSD is often considered the most 'analog sounding' of the popular digital recording formats. The human ear (and analog tape) effectively has a near infinite impluse response capacity, and the closer a digital format can come to emulating that, the more natural it will sound to us.

And the higher frequency response capability that you also get with a higher sample rate, I see as just a 'bonus feature' that you get in addition to the main benefit of having a high impulse response.

But I agree with you, that having a FR higher than 20kHz has much less audible impact than all of the marking fluff would have us believe.

 

[jkpenrose]

Thank you for all the information! I’ve added that video to my watch list and am definitely interested in getting to it. So if I interpreted correctly, the Bifrost 2 won’t necessarily “work better” with 16-bit songs, since 24-bits is hardly audibly better anyways?

I am not an audio engineer and may not understand this fully, and this is from many years ago, but my understanding is that greater bit depth translates to greater dynamic range. With digital volume control the signals dynamic range is compressed in the digital domain (effectively reducing the bit depth of the source) and brings the audible information closer to the noise floor - when you run this signal to an amplifier after the conversion the noise is also amplified and can become audible.

This is why, as I understand it, that doing digital volume control (back in the day) could degrade the sound and that you should do volume control with high quality analog gain after the signal was converted, and why you should feed your pre-amp/amp at full line level.

One way around this, is for the dac to upsamole the incoming signal to a higher bit rate before applying digital volume control and converting the signal to analog - this helps preserve the separation of the sound we want (music) from the noise we don't want - when processing in the digital domain.

But if you are not doing digital volume control - the extra bits don't matter (at least for this reason) because 16 bits provides plenty of dynamic range to cover what we can hear, while keeping the noise below audible levels.

If I've got this wrong, someone with more knowledge and understanding please clarify!

 

[rkw]

my understanding is that greater bit depth translates to greater dynamic range

No, that is not correct. Greater bit depth translates to higher resolution of dynamic level, but the dynamic range is the same. As an analogy, imagine a 1 foot ruler. One ruler is marked only in inches, and another ruler has 1/16 inch markings. They both measure the same minimum to maximum (1 foot) which is like dynamic range, but the one with 1/16 markings can measure more accurately, which is like having greater bit depth.

 

[roskodan]

I always wondered how would a discussion about audio, between beings from different universes, look like. This must be it!

 

[aGloveSupreme]

I always wondered how would a discussion about audio, between beings from different universes, look like. This must be it!

Audio is a bit more difficult to get a handle on for folks that anything visual. We experience things through our vision first and foremost IMO.

I find it interesting that if you are to do a blind tasting or say, jolly ranchers, it would be very difficult to actually pick out what flavour is what, but if you see them first it’s very obvious. It’s the same with great food, say Michelin star restaurant food. If you see it first, then eat it, well.. you can taste the nuances. If you were to go blindfolded that wouldn’t be the case.
I
I don’t think that it’s the case or seeing is believing, I think it’s a case of the brain knowing what to focus on. Take wine or scotch tasting for example.. if the tasting notes are given to you it gives you the ability to concentrate on trying to taste those notes yourself. Without such guidance you may very well miss out.

it’s not placebo, and it’s not BS. people actually need a heads up on what to pay attention to in order to fully appreciate what it has to offer.

still waiting on my Bifrost 2. Hopefully just another couple of weeks.

 

[jkpenrose]

No, that is not correct. Greater bit depth translates to higher resolution of dynamic level, but the dynamic range is the same. As an analogy, imagine a 1 foot ruler. One ruler is marked only in inches, and another ruler has 1/16 inch markings. They both measure the same minimum to maximum (1 foot) which is like dynamic range, but the one with 1/16 markings can measure more accurately, which is like having greater bit depth.

No, that is not correct. Greater bit depth translates to higher resolution of dynamic level, but the dynamic range is the same. As an analogy, imagine a 1 foot ruler. One ruler is marked only in inches, and another ruler has 1/16 inch markings. They both measure the same minimum to maximum (1 foot) which is like dynamic range, but the one with 1/16 markings can measure more accurately, which is like having greater bit depth.

Thanks rkw... I didn't mean to imply that upspampling to more bits for playback adds dynamic range to the original signal.

Though as I understand it, recording at higher bit rates could capture a greater dynamic range - which at 16 bits already covers the difference between the loudest and softest passages of almost all music. The benefit of doing so would be, as you say, smaller, more precise detail of volume differences (whether we can discern those differences is a different question but also more discrete steps between the noise we don't want and the sound we do - which is beneficial in mixing, mastering and filtering - as I understand it - which granted, could be wrong

But for playback, discerning the volume differences between 16bit and 24bit is theoretically beyond our ability to hear - certainly it is beyond mine - or at least I think so!

I think it was an old article on ESS Sabre 32 chip that described the benefits of bit rate upspampling for digital volume control (but this was many, many years ago) and while it made sense to me, maybe was just marketing propaganda.

I use to be more concerned about the technical stuff, which is still interesting, but now it is way more about enjoying the experience - even if what I hear is placebo or snake oil - if I LIKE it better - that's enough these days.

For example, last night I listened to flac files from Amazon hd vs the same files on YouTube music... And despite the blind tests I have seen that demonstrate our (or most people's) inability to distinguish between lossy and lossless formats - the flac files were a little more precise in imaging, a little clearer on vocal reproduction, a little more enjoyable and I liked them better - even though not a night and day difference on my es100 and meze 99s - nice, but certainly not top end.

All that said and to stay on topic, I'm inclined to trust Jason and torq and many others that the Bifrost 2 sounds great and is technically competent - whether I like it better than what I have, I'll only know for sure when I get one

 

[Sean_MR]

Pretty much.

Wow this sparked a lot more than I was intending. But I did watch the video, and wow it was fascinating! It honestly blew my mind it was really interesting to me to learn that adding bits doesn’t actually increase detail or resolution. I definitely learned something new thanks for the link!