Jump to content
IGNORED

Ayre QB-9 Twenty upgrade


Recommended Posts

1 hour ago, Ryan Berry said:

I think you're right, Barrows.  We bypass so much in the ESS chips beyond just the filters that we definitely do not use them as intended (but they sound MUCH better to us this way).  I've not noticed much of a difference between PCM and DSD with our implementation, but I also have to admit that I haven't sat down to really critically listen to each method for much time recently.  If anyone has spent some time with the Twenty version to experiment with this, I'm always interested in hearing what people think.

 

Cheers,

 

Ryan

I’ll chime in... In my setup, sample rate conversion to DSD 128 via Roon is the clear winner. DSD seemed to take some added time to burn in, but I’ve not looked back ever since.

 

More presence and emotion... as well as better pace, rhythm, and timing. It’s more akin to live music via DSD, whereas the more apparent aspect with PCM is the precision and detail. This is not to say that precision and detail is lacking with DSD; not in the least compared to PCM.

 

Hence my hope for Native DSD 256. :)

 

Thanks for being active in the forum Ryan!

Link to comment

Hey Des,

 

Thanks for the report.  It sounds like something I need to play with more.  Are you using any additional filtering in their DSP engine or just the conversion to DSD? 

 

DSD 256 is something the unit could technically support in the QB-9, but not with the improvements we made in getting the data from the USB microcontroller to the DAC chip in the QB-9 Twenty.  We did a LOT of listening to the more traditional method vs. our new way of handling the data and there was a pretty clear winner to us.  The tradeoff was that it caps us off at DSD 128 in the unit without getting into some really expensive oscillators like the QX-5 uses.  Theoretically, we could have had it simply revert to sending data the normal way when playing DSD 256, but it makes zero sense to have a higher sample rate that we know will sound worse than the lower rates.  We can likely get more out of it with other lines, but I doubt we'll be able to simply patch in some firmware that gets around this.  I hate to sound negative on something, as I love nothing more than to figure out solutions around issues, but I also hate to have you hoping for something we've done a few rounds on already.  You never know what we may come up with, however!

 

Glad to be a part of the community.  I'm a hobbiest myself and have a lot of similar interests to people here even outside of being a manufacturer, so this is probably the place I'm most active.

 

Cheers,


Ryan

President

Ayre Acoustics, Inc.

Link to comment

@barrowsThanks for the suggestion. That certainly fits with my SOP but I barely have any DSD, only one full album that I paid for and an assortment of tracks from 2L and, maybe, Linn. So, I was lazily looking for the easy answer 😬

Roon ROCK (Roon 1.7; NUC7i3) > Ayre QB-9 Twenty > Ayre AX-5 Twenty > Thiel CS2.4SE (crossovers rebuilt with Clarity CSA and Multicap RTX caps, Mills MRA-12 resistors; ERSE and Jantzen coils; Cardas binding posts and hookup wire); Cardas and OEM power cables, interconnects, and speaker cables

Link to comment
14 hours ago, Ryan Berry said:

Hey Des,

 

Thanks for the report.  It sounds like something I need to play with more.  Are you using any additional filtering in their DSP engine or just the conversion to DSD? 

 

DSD 256 is something the unit could technically support in the QB-9, but not with the improvements we made in getting the data from the USB microcontroller to the DAC chip in the QB-9 Twenty.  We did a LOT of listening to the more traditional method vs. our new way of handling the data and there was a pretty clear winner to us.  The tradeoff was that it caps us off at DSD 128 in the unit without getting into some really expensive oscillators like the QX-5 uses.  Theoretically, we could have had it simply revert to sending data the normal way when playing DSD 256, but it makes zero sense to have a higher sample rate that we know will sound worse than the lower rates.  We can likely get more out of it with other lines, but I doubt we'll be able to simply patch in some firmware that gets around this.  I hate to sound negative on something, as I love nothing more than to figure out solutions around issues, but I also hate to have you hoping for something we've done a few rounds on already.  You never know what we may come up with, however!

 

Glad to be a part of the community.  I'm a hobbiest myself and have a lot of similar interests to people here even outside of being a manufacturer, so this is probably the place I'm most active.

 

Cheers,


Ryan

Hi Ryan,

 

I'll be off on a ski weekend starting today but wanted to acknowledge your response. I'll review the details later but wanted to share my setup details in Roon. I only use Sample Rate Conversion, with settings as per the pics attached. I experimented with Parallelize Sigma Delta Modulator but found it best turned off... It killed the life in the music and made it more forward sounding, almost loud/shouty... completely lost its sense of ease, and killed the subtleties within the soundstage. 

 

I also wanted to clarify - my enquiry was regarding Native DSD 265. You and I have been in email dialogue regarding it not working (yet) with my Antipodes EX. We had not yet conversed regarding potential resolution.

 

Cheers,

Des

 

image.thumb.png.f75ce78fd501cc27e9d43fd25bc60897.png

image.png

Link to comment
1 hour ago, desbiss said:

I also wanted to clarify - my enquiry was regarding Native DSD 265. You and I have been in email dialogue regarding it not working (yet) with my Antipodes EX. We had not yet conversed regarding potential resolution.

 

Ah, right you are!  I had DSD 512 in my head from a conversation I had with another Ayre user yesterday.  Sorry for the confusion.  Yes, DSD 256 SHOULD be a simple fix.  We'll try and follow up with Antipodes to see where we're at on this one.  On a positively note, Melco recently sent us an update to try on our Melco unit here and we were able to play DSD 256 successfully, so I'm quite confident that it's possible.

 

Cheers,

 

Ryan

President

Ayre Acoustics, Inc.

Link to comment
3 hours ago, Ryan Berry said:

 

Ah, right you are!  I had DSD 512 in my head from a conversation I had with another Ayre user yesterday.  Sorry for the confusion.  Yes, DSD 256 SHOULD be a simple fix.  We'll try and follow up with Antipodes to see where we're at on this one.  On a positively note, Melco recently sent us an update to try on our Melco unit here and we were able to play DSD 256 successfully, so I'm quite confident that it's possible.

 

Cheers,

 

Ryan

That’s great news! Thanks again Ryan. :)

 

As an update: when I last checked with Antipodes they had connected Ariel with an individual who is familiar with the automatic Native DSD detection in the Linux software used by Small Green Computer and Sonore, (as well as Antipodes).

 

I understand the problem is not in using the QB-9 with Linux kernels, but the QB-9 Twenty firmware’s ability to be detected as Native DSD capable in the aforementioned Linux application.

 

Thanks again for the cross-company collaboration. Cheers!

Link to comment

I've never seen an advantage in handling files for DSD conversion and applying particular filters/algorithms when DAC is not no-oversampling. If QB-9 "Twenty" continues to apply its own filters or its own upsampling/oversampling, even when using external conversion, there should be no advantage. Personally, I prefer to always work with files in their original format. But I think this is very relative from system to system. Sabre chips usually deal better with DSD, and I believe bit-by-bit data transmission is less dependent on clock synchronization between devices compared to 16 or 24-bit closed packet transmission. There are a lot of factors that determine the outcome, and perhaps that is why DSD works better. But, with good transport, the formats should play with equal quality.

 

Another determining factor of the result is the driver. The driver provided by Ayre, although a bit outdated, in my opinion, is a key to getting the best possible result from DAC. So, I think QB-9 works at its maximum potential with Windows (and the specific driver) than with Linux. I've tried over ten Linux operating systems (including AudioLinux and Euphony), and QB-9 always performs better with Windows, among other factors, also because of its specific Thesycon driver. It is critical to the best result.

Link to comment
2 hours ago, desbiss said:

That’s great news! Thanks again Ryan. :)

 

As an update: when I last checked with Antipodes they had connected Ariel with an individual who is familiar with the automatic Native DSD detection in the Linux software used by Small Green Computer and Sonore, (as well as Antipodes).

 

I understand the problem is not in using the QB-9 with Linux kernels, but the QB-9 Twenty firmware’s ability to be detected as Native DSD capable in the aforementioned Linux application.

 

Thanks again for the cross-company collaboration. Cheers!

 

Hi Des,

 

I'll let Ariel and handle that.  After further looking into things, I don't believe this is the final solution, and as the QB-9 is working flawlessly with other platforms like the Melco, I suspect it's just a matter of tweaking on the Antipodes end.  Again, I don't want to say too much on the subject as I'm not involved with the firmware end outside of the conceptual level, but the goods news is we know that it can work with the firmware in the QB-9 as-is.  There's been a few emails that have gone back and forth on this, so I'll want to check with Ariel once he's back from his trip for the latest update before saying more.

 

Cheers,

 

Ryan

President

Ayre Acoustics, Inc.

Link to comment
21 hours ago, FelipeRolim said:

I've never seen an advantage in handling files for DSD conversion and applying particular filters/algorithms when DAC is not no-oversampling. If QB-9 "Twenty" continues to apply its own filters or its own upsampling/oversampling, even when using external conversion, there should be no advantage. Personally, I prefer to always work with files in their original format. But I think this is very relative from system to system. Sabre chips usually deal better with DSD, and I believe bit-by-bit data transmission is less dependent on clock synchronization between devices compared to 16 or 24-bit closed packet transmission. There are a lot of factors that determine the outcome, and perhaps that is why DSD works better. But, with good transport, the formats should play with equal quality.

 

Another determining factor of the result is the driver. The driver provided by Ayre, although a bit outdated, in my opinion, is a key to getting the best possible result from DAC. So, I think QB-9 works at its maximum potential with Windows (and the specific driver) than with Linux. I've tried over ten Linux operating systems (including AudioLinux and Euphony), and QB-9 always performs better with Windows, among other factors, also because of its specific Thesycon driver. It is critical to the best result.

 

Hi Felipe,

 

I tend to agree, and I know Charley was fairly vocal on his feelings on DSD/PCM differences (or lack thereof) in the past, but I'm always open to an excuse to go listen to some music some more to test!  Computer audio can be tricky at times when it comes to nailing down what is making the real difference, especially when different players enter the mix.  Testing every combination can become a huge time commitment once you start comparing Windows vs. Apple, vs. Linux, fanless systems, dedicated players vs. general use computers, Roon vs. Decibel vs. Audirvana, etc...it gets overwhelming quick.  Similar to you, I always like to play it as recorded.  It sounds best to me in the tests I've done in the past, but not so much that I would discount the possibility of bias of wanting to keep it simple.

 

It's interesting you should mention the driver.  We actually have a brand new one this year that's completely updated and necessary for raw/native DSD playback on Windows.  You can find it on our page at https://www.ayre.com/support/, select Firmware, and then select the USB Driver tab.  Let me know if you have any trouble with it.

 

Cheers,

 

Ryan

President

Ayre Acoustics, Inc.

Link to comment
6 hours ago, Ryan Berry said:

 

Computer audio can be tricky at times when it comes to nailing down what is making the real difference, especially when different players enter the mix.  Testing every combination can become a huge time commitment once you start comparing Windows vs. Apple, vs. Linux, fanless systems, dedicated players vs. general use computers, Roon vs. Decibel vs. Audirvana, etc...it gets overwhelming quick.  Similar to you, I always like to play it as recorded.  It sounds best to me in the tests I've done in the past, but not so much that I would discount the possibility of bias of wanting to keep it simple.

Hi Ryan, thanks for participating here.

 

Yep, I’m a fan of KISS, keep it simple, stupid. Years ago on this forum, I wrote that computer audio combines audiophile nervosa with computer geek obsession. Not a good recipe!

 

When I first got my QB-9 (192 kHz version) I experimented with OS, music players and one or two other computer-side tweaks. Most things made zero audible difference to my ears, in my system. Maybe because of Ayre’s galvanic isolation? And nothing made a significant difference. I *am* able to distinguish  16 and 24 bit files, preferring the latter but I can’t distinguish higher sample rates. DSD sounds fine to me but nothing special over PCM. At this point, I buy 24 bit files when available but otherwise don’t concern myself over such matters, preferring to listen to music rather than formats and esoteric computer tweaks. Maybe that’s why I love the QB-9 - it does one thing and it does it to the nth degree. 

 

The Twenty upgrade is special (although its ability to play 24/352 is pretty much lost on me!). I’m approaching the 100 hour mark and will have more to write about in the coming days/weeks.

Roon ROCK (Roon 1.7; NUC7i3) > Ayre QB-9 Twenty > Ayre AX-5 Twenty > Thiel CS2.4SE (crossovers rebuilt with Clarity CSA and Multicap RTX caps, Mills MRA-12 resistors; ERSE and Jantzen coils; Cardas binding posts and hookup wire); Cardas and OEM power cables, interconnects, and speaker cables

Link to comment

Just a quick update.

 

Somehow, when we shipped our QB-9 DSD out to Ayre, it was delayed due to a train derailment.  Figures, doesn't it?  But, Ayre turned the QB-9 DSD around in a couple days.  I asked them to not ship it back until after the Thanksgiving holiday break, in order to make certain that somebody was home to sign for it on delivery.  Gary Mulder at Ayre made all this work seamlessly.

 

It was pretty cold here when the QB-9 Twenty arrived.  The cabinet was mighty cold after sitting in a UPS truck all day.  But, I plugged it in to make sure it worked ok.

 

It did.  That is, it made sound when fed digital files through the USB port.  Not really good sound.  That was true even after an hour of stabilizing to room temperature.  I wondered if I'd made a huge mistake.

 

So, I let it run all day and all of the night with a continual feed of various music files.  Including material from The Kinks.

 

Yesterday was the 180 hour mark, or thereabouts.

 

What a difference!

 

The QB-9 Twenty is way better than the olde QB-9 DSD, in just about every way.  I won't give a reviewer's list of audiophile cliches, but instead will quote my wife, who sat down and listened.  "This is way better!  It's like a whole bunch of [actual word not used because this is a family forum] has just been taken away.  Now I can hear [a whole bunch of musician's terms that I zoned out of about 30 seconds into]."

 

Moral of the story:  Despite what some people say about warm-up and break-in, it matters with this DAC.  A lot.  It's now back to breaking in with a constant diet of digital files before I listen again.  If it improves at all from this point, which I'll guess it will some based on past experience, this will be really good.  There are a couple areas where the QB-9 Twenty isn't perfect.  Then again, what audio component is?  But, it's more than acceptably close.  Some additional experimenting is probably in order here, once the DAC has some more break-in hours on it.

 

Second point...  With the QB-9 DSD, I'd used an UpTone ISO Regen with the short USPCB adaptor.  It made a substantial difference in performance.  For some reason, I cannot make the system work now with the ISO Regen.  No communication at all, no matter what order I plug things in and turn things on.  That's the bad news.  The good news is that I'm not sure the ISO Regen is needed now or would even help.  The aspects of the sound that the ISO Regen previously improved are already much better with the streaming computer now connected directly to the QB-9 Twenty through the very same cable as used previously.  Yeah, I'd like to solve the mystery of why the ISO Regen won't work now (I did make certain the switch on the USPCB adaptor was set in the right position to power the QB-9 Twenty's USB section) but, maybe in the end it won't matter.  Sometimes less can be more.

 

Finally, one reviewer of the QX-5 compared it to his own QB-9 DSD.  If I understood his description right, he didn't find much difference between the QX-5 and the QB-9 DSD.  Nuanced and slightly were the words he used.  

 

This confuses me.

 

Both my wife and I found that the QB-9 Twenty version is a ton better than the QB-9 DSD was.  Does that make the QB-9 Twenty now way better than the QX-5?  I doubt it.  This just shows how different people hear things differently and how different various systems are.

 

It's great to have the upgrade option instead of needing to re-enter the DAC purchasing market.  Bad for Ayre.  Great for Ayre!

Link to comment
1 hour ago, CG said:

For some reason, I cannot make the system work now with the ISO Regen.  No communication at all, no matter what order I plug things in and turn things on.  That's the bad news.  The good news is that I'm not sure the ISO Regen is needed now or would even help.  The aspects of the sound that the ISO Regen previously improved are already much better with the streaming computer now connected directly to the QB-9 Twenty through the very same cable as used previously.  Yeah, I'd like to solve the mystery of why the ISO Regen won't work now (I did make certain the switch on the USPCB adaptor was set in the right position to power the QB-9 Twenty's USB section) but, maybe in the end it won't matter.  Sometimes less can be more.

 

This is curious.  FYI - I have no problems using the ISO Regen with mine.  Still think it makes a difference in my system.

Main System: Mac mini (Audirvana+, MMK, JS-2) -> ISO Regen (LPS-1) -> Icron 2201 (Rex LPS-1.2) -> ISO Regen (LPS-1.2) -> Ayre QB-9 Twenty -> Headamp GS-X Mk2 -> Classe CT-M600 -> KEF Reference 201/2

 

Link to comment
3 hours ago, CG said:

It was pretty cold here when the QB-9 Twenty arrived.  The cabinet was mighty cold after sitting in a UPS truck all day.  But, I plugged it in to make sure it worked ok.

 

It did.  That is, it made sound when fed digital files through the USB port.  Not really good sound.  That was true even after an hour of stabilizing to room temperature.  I wondered if I'd made a huge mistake.

 

So, I let it run all day and all of the night with a continual feed of various music files.  Including material from The Kinks.

 

Yesterday was the 180 hour mark, or thereabouts.

 

What a difference!

I had mine running within 10 minutes off the cold UPS truck. It sounded as if something was vaguely amiss at that point but it was abundantly apparent that it bested the DSD version in terms of clarity and expansiveness. One day later I thought it still sounded a bit rough in some respects. 
 

I’m now at around 150 hours. To my ears, it clearly bests the DSD version in some important ways and I don’t hear any aspect of performance that has slipped. I want to get past the 200 hour mark before playing my reference tracks. 

Roon ROCK (Roon 1.7; NUC7i3) > Ayre QB-9 Twenty > Ayre AX-5 Twenty > Thiel CS2.4SE (crossovers rebuilt with Clarity CSA and Multicap RTX caps, Mills MRA-12 resistors; ERSE and Jantzen coils; Cardas binding posts and hookup wire); Cardas and OEM power cables, interconnects, and speaker cables

Link to comment
4 hours ago, Tone Deaf said:

 

This is curious.  FYI - I have no problems using the ISO Regen with mine.  Still think it makes a difference in my system.

 

That is curious.  In my case, I just plugged everything in as before.  Nada.  Then I tried various combinations of powering this first or that.  Still nada.

 

There's still some things left for me to try on that, so I haven't given up.  I'd really love if ISO Regen wasn't needed.  I'd also love it if the ISO Regen improved the sound.  In the mean time, more burn-in.

Link to comment
2 hours ago, GAR said:

 

 

I could be wrong, but I'd guess that Ryan Berry fellow might have some insight.

 

Obviously, the QX-5 certainly has the potential to be substantially better, because they had a larger component budget and they used it for better parts.

 

In my own case, I only upgraded to the Twenty because I could and it was a compelling combination of a new, refined design and price.  I've never even seen a QX-5 in person.  Of course, I also don't visit audio stores, so that's not a surprise...

Link to comment

As far as I know, the QB-9 "Twenty" is superior to the Ayre DX-5 (I don't know if I can reveal the source of this information, but maybe Ryan can help us), which was an equipment that by Audio Magazin had 130 points for the XLR output and 125 points for de RCA output, compared to the 125/120 of the QB-9 192/24. The DSD version has never been evaluated, but it would be interesting if the latest version of QB-9 were reviewed again by the magazine, with good digital transport, excellent cables and the latest driver version. I'm sure we'll be surprised just as I'm impressed daily with this DAC. I currently use a Purist Audio Design Limited Edition Power Cable, a cable over $4,000, and it easily matches the change (that's why I always highlight the importance of cables, digital transports and drivers).

 

I don't consider any of these audio magazines 100% reliable, but at least that would be a parameter for us. If it were possible to reach 135 points through the QB-9's "Twenty" XLR outputs, it will surely be one of the best and cheapest DACs on the market. You can buy a used QB-9 from Audiogon for $1,200 and then upgrade for another $1,500. Nothing I know, at this price, has 135 points via XLR output. I'd like to trust these magazines more, but even without trusting, it would be a nice parameter...

 

 

Link to comment
15 hours ago, GAR said:

Actually, I'm in error. I think the proper comparator is the QX-8.

 

Here's the thing...

 

If you just use standard inflation adjustments - which may not be accurate for electronics today due to tariff adjustments - a QB-9 DSD (the old model) would sell for about $3900 in 2019 dollars.

 

A new QX-8 with the USB input option sells for $5450, if the information I found online is accurate.

 

So, for roughly $1500 you get a QX-8 that has a headphone system built in, complete with volume control, and a wider range of options.  Plus, obviously, the new technology that has migrated into the QB-9 Twenty.

 

So, you're right, I think.  But, it may also be that with the advancements, the QB-9 Twenty is closer to the QX-5 than we might expect.  Due to the QB-9's simplicity, it might actually be better in some ways.  The QX-5 is about three years old now, I think, so things may have advanced between then and now.

Link to comment
On 12/12/2019 at 4:20 PM, Ryan Berry said:

 

Hi Des,

 

I'll let Ariel and handle that.  After further looking into things, I don't believe this is the final solution, and as the QB-9 is working flawlessly with other platforms like the Melco, I suspect it's just a matter of tweaking on the Antipodes end.  Again, I don't want to say too much on the subject as I'm not involved with the firmware end outside of the conceptual level, but the goods news is we know that it can work with the firmware in the QB-9 as-is.  There's been a few emails that have gone back and forth on this, so I'll want to check with Ariel once he's back from his trip for the latest update before saying more.

 

Cheers,

 

Ryan

Thanks for your reply Ryan. I look forward to an update and will wait to hear further from you upon Ariel’s return. Much appreciated!

Des

Link to comment

It's been 2+ years since I've heard a QX-5 (and I've never heard a QX-8) but I think the QB-9 Twenty is good enough that a head-to-head comparison in the same system/room would be needed for any meaningful comparison. The QX-5 *probably* sounds better due to higher quality parts ("better" DAC chip, Morion oscillator, etc) but, as CG wrote, the QB-9 should benefit from it's singular purpose.

 

I have been lusting after a QX-5 but it remained beyond my budget (it also has more features than I would likely ever use). So, I was stoked when I heard about the QB-9 upgrade. I can't say how how close it gets to the QX-5 but it certainly narrowed the gap!

 

@FelipeRolim Did Audio Magazine do a head-to-head of the DX-5 and QB-9? I can tell you this much: the QB-9 DSD version is "notably" better than the 192 kHz version and the Twenty version is "notably" better than the DSD version.

Roon ROCK (Roon 1.7; NUC7i3) > Ayre QB-9 Twenty > Ayre AX-5 Twenty > Thiel CS2.4SE (crossovers rebuilt with Clarity CSA and Multicap RTX caps, Mills MRA-12 resistors; ERSE and Jantzen coils; Cardas binding posts and hookup wire); Cardas and OEM power cables, interconnects, and speaker cables

Link to comment

In fact, the version of QB-9 reviewed at Audio was 96/24, as the review was conducted in September 2009. The 192/24 version only came in 2011, the F revision, and the DSD version in 2014, the J revision. I didn't see any direct comparisons between QX-5 and QB-9, just saw the notes on the Audio Magazin list (I attached a photo from the magazine).

 

I'd like to see QB-9 "Twenty" being reviewed again. I believe we'll have a nice surprise.

Screenshot_2019-12-16 de downmagaz com - AUDIO 2020 01 pdf.png

Link to comment

CG

I have had issues with my ISO REGEN  connected to my QX-5 20 as you described earlier, original USB REGEN functions normally. I installed a simple on/off switch in the power line providing power to the ISO REGEN. Normally power to the ISO REGEN is always on. If there is a problem I cycle the power off and back on to the ISO REGEN and it then works fine till the next time. What is so weird is that I have not had any issues with either REGEN when connected to my QB-9 20!

A flaw in reasoning is a mistake in how conclusions are derived from assumptions, not a mistake in assumptions.

 

AB835

Link to comment
Quote

 

Quoting from this article: https://audiophilestyle.com/ca/reviews/my-quest-for-a-new-dac-part-2-ayre-qx-8-r737/

 

Everything I like about the Codex, the QX-8 does similarly, but better. I've described the Codex as having a lovely midrange, a satisfyingly solid mid-bass, and a smooth, refined treble. Dynamics-wise, it can be a bit polite-sounding, which I attribute to its necessarily small PSU, to fit its small case.

 

Ryan, do you agree with this statement, in particular, the PSU? Does the same thing apply to the QB-9 Twenty? The size and quality of the PSU seems to be SO important to all audio gear.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...