[rmkjr]

A recent review I came across:

 

[Lépine]

@emilsoft Yes, you are correct. After maxing out every possible way to increase the buffers everywhere I can configure in the drivers and also in JRiver and not getting through a song without either a pop or sudden erratic behavior, I tried the exact opposite and went below every recommended buffer setting to absolutely zero or minimum possible. It did not make it worse. So buffering doesn't do anything, apparently and as should be expected, USB-C is plenty fast enough that it is not the cause of the jittery behavior.

I think I have found a successful combination. Finally, so I think the HiBy ASIO driver is junk. HiBy WASAPI is also not any better.
The settings that finally seem to play through multiple files without a glitch is by using the "Audio Output[Kernal Streaming]" and setting the bit depth to 32-bit DoP DSD64. At least this is one setting that is working for a half an hour without a problem. All other settings would not work for more than 2 or 3 minutes without some random pop/click/stumble/tempo change/jump. The music also sounds more normal without some sense of being a little inexplicably off. So far approaching 10 songs without any glitches.
This maybe the answer, either way, the music has taken on an appreciable amount of improvement that I wasn't aware I was missing until now. So further experimentation is likely needed to see if it is the Kernel Streaming, or 32-bit, or conversion to DSD or some combination of these settings that has remedied the problem. And at the moment, the buffer settings are set to absolute minimum possible and no problem. I can even open and close other applications without causing problems as should be expected from a way over specified laptop that normally unless running the most resource intensive HQPlayer settings, barely registers above 2%

Looks like I have found my solution, I am about to finish an entire album without a glitch and now I have all the radios turned back on as well as several Windows background processes. Nothing including notifications or scanning the SSD causes any change in the music. It is very stable.

Update: 16 tracks played now without a single glitch.

Update 10/31/2022: This is a working fix. Choosing the output driver interface "Audio Output[Kernal Streaming]" and setting the bit depth to 32-bit DoP DSD64.

Upate 10/31/2022 evening: After more experimentation, it appears that Audio Output[Kernal Streaming] is in fact using the Hiby Windows driver, but it is appearing to JRiver as a direct connection without any of the api interface layers is what I assume. To get the 32-bit mode to work, you must configure the option in the Hiby Audio Interface Control Panel. It sounds better than the 24-bit mode which sounds a bit sibilant. Changing the output from DoP to DSD seems to cause very unstable problems where I get very loud white noise so we won't be experimenting with that anymore.

 

[Lépine]

The RS6 sounds fantastic, by the way. I would buy this again. I love the sound and Darwin functionality gives me the ability to change the sound properties that my SP2000 really doesn't do. The SP2000 is great, but EQ and DAC filters are so subtle that depending on the earphone or headphones, I can't tell if it makes any difference.

I don't have any rolled off treble or bass heavy sound that I have read about. The first couple of hours did sound very bassy like a subwoofer sitting under my desk or treble turned way down. After upgrading the firmware to the current and a few hours somewhere during that period, the sound matured into a fairly neutral presentation that I am very pleased with.

I think HiBy should check if they have any issues with their Windows drivers. They are not working for me, but lucky there is a perfectly acceptable workaround.

Listening with the 64Audio U12t in balanced mode is really very enjoyable pairing.

"Ain't No Sunshine" by Karen Souza has great presence, atmosphere, and the naturalness of the acoustical instruments (piano and double bass). The rendering of the details in Karen's voice is among the best I have experienced with all of my gear. Now I maybe regretting not having waited for the RS8.

If anyone thinks the Cayin R-U6 is a good example of an R-2R, they need to hear an RS6.

 

[gLer]

If anyone thinks the Cayin R-U6 is a good example of an R-2R, they need to hear an RS6.

Very true this. RS8 is RS6 on steroids - same earthy sound but big jump up in technical ability. Significantly better battery life too, but it's also bigger and heavier.

 

[dsrk]

RU6 sounds like may be an R2R but RS6 sounds like an absolute R2R.

 

[emilsoft]

@emilsoft Yes, you are correct. After maxing out every possible way to increase the buffers everywhere I can configure in the drivers and also in JRiver and not getting through a song without either a pop or sudden erratic behavior, I tried the exact opposite and went below every recommended buffer setting to absolutely zero or minimum possible. It did not make it worse. So buffering doesn't do anything, apparently and as should be expected, USB-C is plenty fast enough that it is not the cause of the jittery behavior.

I think I have found a successful combination. Finally, so I think the HiBy ASIO driver is junk. HiBy WASAPI is also not any better.
The settings that finally seem to play through multiple files without a glitch is by using the "Audio Output[Kernal Streaming]" and setting the bit depth to 32-bit DoP DSD64. At least this is one setting that is working for a half an hour without a problem. All other settings would not work for more than 2 or 3 minutes without some random pop/click/stumble/tempo change/jump. The music also sounds more normal without some sense of being a little inexplicably off. So far approaching 10 songs without any glitches.
This maybe the answer, either way, the music has taken on an appreciable amount of improvement that I wasn't aware I was missing until now. So further experimentation is likely needed to see if it is the Kernel Streaming, or 32-bit, or conversion to DSD or some combination of these settings that has remedied the problem. And at the moment, the buffer settings are set to absolute minimum possible and no problem. I can even open and close other applications without causing problems as should be expected from a way over specified laptop that normally unless running the most resource intensive HQPlayer settings, barely registers above 2%

Looks like I have found my solution, I am about to finish an entire album without a glitch and now I have all the radios turned back on as well as several Windows background processes. Nothing including notifications or scanning the SSD causes any change in the music. It is very stable.

Update: 16 tracks played now without a single glitch.

DSD apparently takes a different pathway in Hiby vs standard PCM, there was talk somewhere in this form but i don't remember and too lazy to look now, maybe why it's working fine.. Good to know there is something, but I don't have this option as I'm using Apple Music/Qobuz. Also - the sound changes/improvements you are sensing might not be related to this at all but due to R2R burn in (hundreds of hours, and also pre-warmup in hours before listening sessions even post burn in).

 

[emilsoft]

One thing I noted , the 3.5mm line out doesn't sound great to me (bit rough), but i haven't burned it in properly. Mojo 2 treated as line out sounds much better, smoother, wider, more dynamic, you name it, going into my desktop amplifier. I haven't been able to test the 4.4mm line out properly yet.

 

[emilsoft]

If anyone thinks the Cayin R-U6 is a good example of an R-2R, they need to hear an RS6.

Cayin RU-6 was an interesting and enjoyable listen at first, but after some time I realised it pollutes the signal too much - it's R2R implementation is basic to fit in a dongle, basic R2R can be a nasty thing - i still think the RU6 does some things nice like mids, but it measures horribly and once hearing something like RS6 one hears what a R2R should really be.. still RS6 is not perfect, they use cheap 0.1% resistors which is a bit disappointing considering the cost. See what Luxury & Precision use in their DAPs

 

[Lépine]

DSD apparently takes a different pathway in Hiby vs standard PCM, there was talk somewhere in this form but i don't remember and too lazy to look now, maybe why it's working fine.. Good to know there is something, but I don't have this option as I'm using Apple Music/Qobuz. Also - the sound changes/improvements you are sensing might not be related to this at all but due to R2R burn in (hundreds of hours, and also pre-warmup in hours before listening sessions even post burn in).

Just to clarify, I realize I omitted an important detail. I am transferring realtime conversion to DSD by DoP protocol. So isn't that also PCM encapsulated, therefore using the same data path as other "native" PCM data? I will do more testing later to see if it is just the Kernal Streaming interface that is the real solution. I suspect JRiver, HQP, FB2K do not work well with HiBy ASIO or HiBy WASAPI interface drivers.

Once I found a way to trace the data path with JRiver, I looked for a direct connection data path. The KS interface is the first one I came across that was directly connected (and did not have thousands of milliseconds delay between played and heard). So it is not running through another separate executable that maybe either doesn't work well or is having priority issues even after promoting to realtime. The ASIO executable seems to be the same software as the one used by Astell Kerns SP2000. I haven't checked if my SP2000 is now not working well with USB DAC mode after installing the Hiby version of the ASIO runtime executable.

 

[Lépine]

Very true this. RS8 is RS6 on steroids - same earthy sound but big jump up in technical ability. Significantly better battery life too, but it's also bigger and heavier.

At those price levels, I would need to seriously compare with a SP3000. Thank you for the note, I will envy all of you that got an RS8. But, I am perfectly happy with the RS6 now. And it isn't too much investment that I wonder if I shouldn't get a Holo May or Terminator instead.

For $1200 delivered, I think RS6 was a smart buy.

And now that my unit is fully working the ways I use it, I can also admire the beautiful rose colored casing. My SP2000 is also Copper.

 

[Lépine]

The sound stage is very impressive, very strong sense of realism/holographic, listing to Barbara Lusch "Every Night" (16-bit/1411 bitrate WAV). I am using at the moment an old set of Noble Audio Kaiser K10U. It is nice to get desktop class soundstage even if I am cheating and using a laptop to convert to DSD128 over DoP without any upscaling. It may sound better from the laptop than directly off the internal memory card.

I had to just compare against my SP2000.... I think the SP2000 has the edge, very different sounds and both very enjoyable. The RS6 is brighter closer to live music.

 

[dsrk]

The sound stage is very impressive, very strong sense of realism/holographic, listing to Barbara Lusch "Every Night" (16-bit/1411 bitrate WAV). I am using at the moment an old set of Noble Audio Kaiser K10U. It is nice to get desktop class soundstage even if I am cheating and using a laptop to convert to DSD128 over DoP without any upscaling. It may sound better from the laptop than directly off the internal memory card.

That's the beauty of RS6, most other DAPs which have greater soundstage sounds brighter.
RS6 has better separation and soundstage compared my desktop DAC/AMP EF400.

I had to just compare against my SP2000.... I think the SP2000 has the edge, very different sounds and both very enjoyable. The RS6 is brighter closer to live music.

It's been a while since I heard the SP2000.
SP2000 sure sounds technically better than RS6 but I would still take RS6 over it for pure sense naturalness.

 

[Lépine]

It's been a while since I heard the SP2000.
SP2000 sure sounds technically better than RS6 but I would still take RS6 over it for pure sense naturalness.

I am not sure what technically better means, but if we are discussing which player sounds closer to reality, the RS6 is the clear winner in NOS mode. However, if we are discussing any harshness, again it would be the RS6 in NOS mode with all DARWIN adjustments turned off.

The SP2000 sounds more attenuated in certain frequencies so there is a very apparent Astell&Kerns designed playback rendition as if you have a sound engineer that insists on dampening the concert hall acoustics to his preference rather than unveiling everything inherently present in the venue and amplification as well as the performers.

The RS6 definitely can be a more unprocessed naked sound. Everything right or wrong can come through to the earphones.

Which is technically superior, I do not know, but the fact that I am not sure is another example of Moore's/Neven's Law.

An M-scaler...technically superior. The sound is transparent and slight flaws are corrected during reconstruction and filtering.

 

[dsrk]

One thing I noted , the 3.5mm line out doesn't sound great to me (bit rough), but i haven't burned it in properly. Mojo 2 treated as line out sounds much better, smoother, wider, more dynamic, you name it, going into my desktop amplifier. I haven't been able to test the 4.4mm line out properly yet.

You are right about the 3.5mm line out, it's a subpar implementation. I was using NX7+RS6 for a while now, I thought NX7 was smoothing out the higher frequencies. But it's the 3.5mm line out that is the culprit. I have tested this today and I don't see that issue when connected it using the headphone out, wow!

I am very surprised and disappointed to see that line out not up to the headphone out on such a high quality device. I can now insert the dust plug into the 3.5mm line out too, what a waste.

 

[Lépine]

Hopefully, a week after November 8th, I will be able to try the Hiby RS6 with a new IEM.
Anyone guess what IEM I am going to try?