Schiit Happened: The Story of the World's Most Improbable Start-Up

[bagwell359]

Pages and pages of incomprehensible verbosity. Seriously this is so pretentious and affected it is impossible to read if you dont drink the kool aid.

Originally I thought this was a way to keep Schiit people from leaving its biosphere, and I wasn't going to get any FB.

But, the functions at finger tip on the Gungnir 2 ended up being worth it to me. Now with the 5 band PEQ being added (for free) to my Gungnir 2. No way its remote is going to allow manipulation of PEQ so another win, not to mention the loudness option.

The Wotan on its way will extend that for me. Kara too.

I think I'll be putting all of my Schiit remotes (except Rag 2) in a drawer.

 

[cspirou]

Wouldn’t this require clock sync between DACs, not just volume? Not sure how could this be implemented with two independent DACs instead of a multi-channel one.

Yeah that’s true. There would need to be some sort of master clock to sync everything up

 

[sickmonkey]

Go rip a CD to FLAC and you’ll also see it’s not 1411 kbps. That’s just a worse case figure. The rest is the magic of compression.

Now I may need some education on this point. I want to learn. I am looking right now at Aimee Mann's 31 Today. Ripped via DBpoweramp FLAC no compression. Bit Rate is 1413kbps. File Size is 51,104 KB. If I go home and play this from Qobuz I think it will be about ~900 kbps.

Maybe a better example. I just remember I have both here at office. The Cure Mixed Up (discussed earlier). I have two copies. One I ripped via DBpoweramp FLAC no compression. 1415kbps file size 56,789kb. The same album downloaded from HD tracks as "CD Quality 44/16". 919kbps file size 36,804kb. Where did the 19,985kb go? That's roughly 36%. Is it all just packed zeros? H2G, I am just trying to understand.

 

[Orange5o]

Cross posting a bit, but in further prep for operation Valhalla 3 bedside stack, just got a new ZMF Bokeh. I think it will look cool with a tube amp, and also the right cup reminds me of one of my favorite book covers...

 

[roderickvd]

Now I may need some education on this point. I want to learn. I am looking right now at Aimee Mann's 31 Today. Ripped via DBpoweramp FLAC no compression. Bit Rate is 1413kbps. File Size is 51,104 KB. If I go home and play this from Qobuz I think it will be about ~900 kbps.

Maybe a better example. I just remember I have both here at office. The Cure Mixed Up (discussed earlier). I have two copies. One I ripped via DBpoweramp FLAC no compression. 1415kbps file size 56,789kb. The same album downloaded from HD tracks as "CD Quality 44/16". 919kbps file size 36,804kb. Where did the 19,985kb go? That's roughly 36%. Is it all just packed zeros? H2G, I am just trying to understand.

When you ZIP a Word document, then unzip it, you will also see that you got back all the content

This is how FLAC works: it looks at a set of samples. It’s a wave. It tries its best to approximate the shape of this wave with a mathematical expression. Storing this expression costs way less than all the data points the wave represents. How hard it tries, is determined by the compression setting.

Now it compares the approximate waveform with the actual set of samples. What remains is the “delta” it has to encode. This is less efficient than the mathematical curve, but still way less than encoding the entire sample.

When decoding, it reconstructs the original from the approximated wave plus the delta to the original samples.

 

[Syan25]

Schiit prices for UK are in Pounds and that is the worst currency for me + deliveries from UK to EU are very expensive and custom formalities are nightmare.
Purchase from USA is better but again it is not like having it from EU.

Yes. It costs me more buying in the UK.

 

[sixergixer]

This is one of my go to CDs. Fascination street really shines along with The Walk.

Thanks for the heads up. I just ordered Lost World remix. June 13 release.

Take Mixed Up and Staring at the Sea in comparison, What? IMHO both are just awesome, but incredibly different.

That huge variation across The Cure's timeline is what fascinates me. Even their bouncier pop stuff shines.

 

[sickmonkey]

It tries its best to approximate the shape of this wave with a mathematical expression. Storing this expression costs way less than all the data points the wave represents.

Thank You. This makes more sense. Yes, of course, you can compress an exe and uncompress have complete program, just like a word doc. I guess I was thinking more along lossy. FLAC is lossless and I should be good and accept all the data is there unaltered. I guess I have trust issues! So much BS gets played out today.

Thanks Again for clearing this up.

 

[Adam Balm]

Sometimes it takes a teeny bit of drama to remember and fully appreciate how good we have it here compared to other forums/threads.

 

[ArmchairPhilosopher]

Now I may need some education on this point. I want to learn. I am looking right now at Aimee Mann's 31 Today. Ripped via DBpoweramp FLAC no compression. Bit Rate is 1413kbps. File Size is 51,104 KB. If I go home and play this from Qobuz I think it will be about ~900 kbps.

Maybe a better example. I just remember I have both here at office. The Cure Mixed Up (discussed earlier). I have two copies. One I ripped via DBpoweramp FLAC no compression. 1415kbps file size 56,789kb. The same album downloaded from HD tracks as "CD Quality 44/16". 919kbps file size 36,804kb. Where did the 19,985kb go? That's roughly 36%. Is it all just packed zeros? H2G, I am just trying to understand.

The file itself doesn't matter. What you're comparing is different ways to interpret the original PCM data and different ways to store that data on your drives to optimize for space, not the actual audio data contained within the files after they've been uncompressed and processed back to their original state.
(This obviously only applies to actually lossless formats like WAV and FLAC, not lossy ones or those now-obsolete formats who shall remain unnamed that liked to pretend to be lossless.)

I've spent quite a few weekends recently doing a real world deep dive into figuring out why an audio stream from an Urd via Unison into a headphone setup sounds different than a bit-perfect rip of that very same disc streamed from my Mac through Unison to the exact same DAC and headphone setup.

The results were interesting, and what I've learned will likely find its way into future products.

But for now, what I'll say is this:
At least for the way our DACs are implemented, the PCM data that the DAC ultimately "sees" is all that actually matters. And depending on what's happening in the player software or the operating system's or a third party's system-wide sound mixer (if not bypassed in some way or another) you can get a different result from the very same audio data, no matter in what format that data was stored, depending on how that data is being played back (in the case of your computer) or the medium that the data was stored on (in the case of a computer versus a transport, for example). What counts is that the decoded, uncompressed PCM data that eventually enters the DAC is the same in all cases. If it is, there's no appreciable difference—audible or measurable—in the resulting analog output of your DAC.

(And before someone complains: With Unison, even the "cleanliness" of the signal (i.e. lack of jitter or ringing in the actual, physical signal) doesn't matter, at least not up to the point where you run into buffer underflows due to dropped or otherwise unreadable frames. And if you get to that point, something's seriously broken anyway. With SPDIF, the cleanliness of the source can matter to a degree under certain circumstances and might make a difference, even if the PCM data being transmitted remains precisely the same.)

 

[JanMc]

So when you look at the actual bit stream hitting the DAC from those two sources, do you see differences?

 

[ArmchairPhilosopher]

So when you look at the actual bit stream hitting the DAC from those two sources, do you see differences?

Yes, there are differences if there's something in the path that the audio data is taking that "screws" with it before it comes out of your USB port. That can be a player software with integrated resamplers, EQ, or mixers, an operating system's own or a third party's system-wide audio mixer that you installed knowingly or unknowingly, etc.

If there isn't, though, the data that comes into the DAC is exactly the same. And in those cases, the analog output sounds and measures precisely the same, regardless of the source or the storage medium.

And just to be abundantly clear: This doesn't necessarily have to apply to a 16 bit 44.1 stream from a streaming service when compared to that same album's CD, for example. It can be the same, but depending on whether any processing happened on the label's or streaming services' end, it doesn't necessarily have to be the same. My tests and the results I saw only apply to comparisons between CDs I own and bit-perfect rips I personally made of those very CDs.

 

[theveterans]

The file itself doesn't matter. What you're comparing is different ways to interpret the original PCM data and different ways to store that data on your drives to optimize for space, not the actual audio data contained within the files after they've been uncompressed and processed back to their original state.
(This obviously only applies to actually lossless formats like WAV and FLAC, not lossy ones or those now-obsolete formats who shall remain unnamed that liked to pretend to be lossless.)

I've spent quite a few weekends recently doing a real world deep dive into figuring out why an audio stream from an Urd via Unison into a headphone setup sounds different than a bit-perfect rip of that very same disc streamed from my Mac through Unison to the exact same DAC and headphone setup.

The results were interesting, and what I've learned will likely find its way into future products.

But for now, what I'll say is this:
At least for the way our DACs are implemented, the PCM data that the DAC ultimately "sees" is all that actually matters. And depending on what's happening in the player software or the operating system's global sound mixer (if not bypassed in some way or another) you can get a different result from the very same audio data, no matter in what format that data was stored, depending on how that data is being played back (in the case of your computer) or the medium that the data was stored on (in the case of a computer versus a transport, for example). What counts is that the decoded, uncompressed PCM data that eventually enters the DAC is the same in all cases. If it is, there's no appreciable difference—audible or measurable—in the resulting analog output of your DAC.

(And before someone complains: With Unison, even the "cleanliness" of the signal (i.e. lack of jitter or ringing in the actual, physical signal) doesn't matter, at least not up to the point where you run into buffer underflows due to dropped or otherwise unreadable packages. And if you get to that point, something's seriously broken anyway. With SPDIF, the cleanliness of the source can matter to a degree under certain circumstances and might make a difference, even if the PCM data being transmitted remains precisely the same.)

There's more into the pandoras box. Now if you're streaming using a streamer on Ethernet, you'll find Ethernet filtering, Ethernet switches and Ethernet cables which delivers packets do affect the sound as well. My educated guess with the Urd vs Mac is through the theory that digital signals are analog in nature in the form of square waves, i.e. there is a huge tolerance that the voltage swings on the 1's and 0's that the signal would still be intact despite the imperfections in the square wave signal. Noise can affect the timing or the amplitude of the square wave, but despite all those variations, the DAC is still receiving the same digital information, but the analog noise contamination does spread out and transmit through the transducers hence we hear a difference between Mac and Urd (as these two have varying amount of jitter that is measureable and noise generated by CPUs, no shielding/isolation of USB outputs in the MOBO, mains leakeage, etc.) Unison can clean some of those noises from the Mac but never completely so if you have a cleaner digital signal going into the Unison, you will hear better sound overall.

If you look at Pi2AES Mercury, there is isolation transformer before the signal goes out to their SPDIF ports which means the designer knows that the RPi itself generates some noise and isolation is needed to reduce significantly/not totally eliminate the noise which to me is impossible by physics.

 

[sickmonkey]

The file itself doesn't matter. What you're comparing is different ways to interpret the original PCM data and different ways to store that data on your drives to optimize for space, not the actual audio data contained within the files after they've been uncompressed and processed back to their original state.

I've spent quite a few weekends recently doing a real world deep dive into figuring out why an audio stream from an Urd via Unison into a headphone setup sounds different than a bit-perfect rip of that very same disc streamed from my Mac through Unison to the exact same DAC and headphone setup.

The results were interesting, and what I've learned will likely find its way into future products.

But for now, what I'll say is this:
At least for the way our DACs are implemented, the PCM data that the DAC ultimately "sees" is all that actually matters. And depending on what's happening in the player software or the operating system's global sound mixer (if not bypassed in some way or another) you can get a different result from the very same audio data, no matter in what format that data was stored, depending on how that data is being played back (in the case of your computer) or the medium that the data was stored on (in the case of a computer versus a transport, for example). What counts is that the decoded, uncompressed PCM data that eventually enters the DAC is the same in all cases. If it is, there's no appreciable difference—audible or measurable—in the resulting analog output of your DAC.

(And before someone complains: With Unison, even the "cleanliness" of the signal (i.e. lack of jitter or ringing in the actual, physical signal) doesn't matter, at least not up to the point where you run into buffer underflows due to dropped or otherwise unreadable packages. And if you get to that point, something's seriously broken anyway. With SPDIF, the cleanliness of the source can matter to a degree under certain circumstances and might make a difference, even if the PCM data being transmitted remains precisely the same.)

This is great to read from the point of people smarter than I are thinking this deep into the process. Much appreciated.

I made a comment earlier about a Wiim Ultra sounding a bit shrilly and bright. . Not in a good way. The source files were from the same NAS going to the 384 input on Gungnir 2. The MAC mini with Audirvana did not exhibit this characteristic. Note: Audirvana was neutered as was the Wiim (no eq and such) and going straight to the Gungnir. Clearly, it's the same source and what should be the same digital presentation at the Gungnir 2 USB port. But to my ears and others Nope, Nada.

I can't explain it. I know it's different. My example above, Wiim/Mac, was very obvious between the two sources. Very apparent to others also.

I will note that when I had my CD transport connected to Bifrost 2/64 and the Mac mini above. I really could not find a difference. I tried to convince myself there was a difference, but no. I could not tell.

If only we had a digital, clear sight glass that we could see the data running through. We could see the assay and feel the viscosity of the digital stream. Can you put that into Forkbeard ?

 

[Ableza]

But does it sound good? It does? Then I don't care.