Is there any sound quality improvement with the new Roon 1.8? What are the improvements, if any?

[Bill_Janssen (Wigwam wool socks now on asymmetrical isolation feet!)]

@ipeverywhere

Not everyone agrees vanilla is the best ice creme.

We don’t even agree on how to spell “ice cream”, apparently.

Clearly the right spelling is “frozen custard”, and the best flavor is red raspberry.

If you’re stuck on the benighted West Coast, this vanilla isn’t a bad option.

[Bill_Janssen (Wigwam wool socks now on asymmetrical isolation feet!)]

@ipeverywhere

I owning it though.

I thought that might be a British spelling? Like “colour” or “theatre” – words which my teachers used to flag when I used them.

[HWZ]

If Roon 1.8 looks better, it will automatically sound better, too.

[rrwwss52 (Robert)]

Any SQ program improvements received in 1.8 would be a great bonus.

[CrystalGipsy (Resisting the flow!)]

I commented recently on the difference between LMS and Roon on a raspberry pi. I didnt expect there to be a difference and have always been in how can it be different it’s bit perfect.

So I tested myself rather than rely on hersay and denial, and yes there is a subtle difference. I found no difference using Squeezelite or RAAT via Roon itself but Roon Vs LMS there was.

Now I am not unhappy with Roon or demanding they change anything nor am I changing away from RAAT but there is a difference, whether you find this of importance is up to you, its not like ones horrid the other isn’t, it’s subtle like changing a DAC filter, some people can hear these more than others.

If I get the opportunity to get a pcm audio recorder I’ll capture the output and compare them but I don’t have access to any right now.

This is all just using a raspberry pi 4 and Ropieee XL, SMSL SU9 DAC and Meze 99 Classics only difference being the server used as I can’t run LMS on ROCK. So thats my tuppence worth, I don’t expect it will change any blinkerd thinking but there you have it. Those who deny there is a difference without listening try it out its free to do so and doesn’t take long to set up and is easy to switch playback between them on the same device. Then feel free to discount it if you wish to cognitive behaviour if that’s your belief.

[MusicFidelity (Guy)]

Would it possible to defer the major upgrade to 1.8 as I know there will be some bugs to iron out? I don’t want the upgrade to disrupt my music listening. SQ is still an unknown and probably I want to stick to 1.7 for a while.

Until then when the new release goes through a few round of bugs fix, then I would like to upgrade.

[PeterD]

Everybody is rolling over each about features and SQ in version 1.8.
The damn thing is not even out yet.
Wait and see what it is bringing and based on that, go into a discussion about what is and what is not.

[Xekomi]

@Sgr

Why would do that? Why of course so they won’t be outdone by their competitors.

Exactly. Streaming (at the receiving end) takes up single digit percentages of the available computing power from even a Raspberry Pi from a couple of generations ago. There are good reasons to prefer a new Pi over an older one, but those aren’t related to an interplay between computing power and playback quality.

I agree with @sbr when they say that you’d save yourself a lot of anxiety and money by not being taken by server maker nonsense - I’d rephrase that as always putting the ball back in the court of those making the extraordinary claims, though. It isn’t about loosing the fun and romanticism of thinking that our ears are better than measurement gear, it’s about making certain that we’re paying for more than support, casework, and a lot of flattery.
Asking manufacturers to prove their claims, and then checking with others that the proof is valid, and if it isn’t, asking again until you have an answer that can be independently and objectively verified is not only free, but you’ll learn a lot more along the way than just what cable or magic rock sounds best.

[BigAIMc (BigAlMc)]

I know this is intended helpfully but sadly it’s also symptomatic of the problem with Roon Labs being completely dismissive when paying subscribers raise valid requests that Roon focus on improving the sound quality.

I should buy an optical Ethernet cable because it’s not realistic to expect Roon to take my concerns seriously. Sorry but that’s not good enough and I along with many others will vote with my feet when a more compelling software offering comes along.

I have a lifetime subscription and there are many things I love about Roon. Roon 1.8 sounds very intriguing. But at the end of the day SQ is extremely important to audiophiles and continually fobbing us off is a reckless approach.

I’m fortunate that in my case my server has largely ‘fixed’ Roons SQ because Innuos implemented Roon playing to a virtual Squeezebox endpoint and implemented RAM playback. This sounds WAY better than Roon playing itself.

But on point of principle I shouldn’t have to rely on a hardware manufacturer fixing the software shortcomings that the software manufacturer refuses to take seriously. The difference is that Innuos listened to its customers.

And I know I’m about to get flamed by the ‘bits r bits’ flat earther regulars who would rather get snarky, rather than god forbid have their sound quality improved even if it involved techniques they are doubtful of. God forbid! Sadly one of the reasons I barely post on this forum these days.

Taiko, PS Audio, whoever. Bring it on. Roon needs the competition because this complacency around SQ isn’t acceptable to many of us.

[andybob (Andrew Cox)]

I’m not a Roon employee, just a volunteer user, so anything I say should not be attributed to Roon. Personally I don’t use optical Ethernet, but it is out there if people want better isolation.

Are you using a setup with a separate Core and Output ?

I wouldn’t say that Roon is complacent about SQ, but the Core is always going to perform more intensive computing than software that does not include the same music database features.

[Xekomi]

@David_Zigas

I made the comparison several times over a number of weeks, and was, in fact, biased toward wanting Roon to sound better.

I’d suspect it isn’t impossible that maybe herein lies the problem: because SqueezeServer is less comfortable to use, there’s probably a reason to why people put up with it, and that reason’s got to be that it sounds better typalogic, whether conscious or not… one could of course push the fallacy further and rationalise that as “because it’s less graphically intensive, it sounds better” or something of the sort, eh.

Once again, it’s not impossible that there are differences, but it is extraordinarily unlikely. What’s certain is that there is good money to be made in perpetuating the idea that the unlikelihood is a reality.

[Xekomi]

@BigAIMc

Taiko, PS Audio, whoever. Bring it on.

Yeah baby, show me your measurements before ripping those suckers off !

[Peter_Bruderer]

Sound quality of Roon build 537 [retracted claim]

Heads up for everyone though… please do not pile-on, and try to be as nice as possible. I happen to know what Taiko Audio did to make things “sound better”, but it would be inappropriate for me to spill their internal details. You will have to ask them. I agree with some of their choices (nothing to do with CPU core allocation or anything noted above), and because those changes were made at the same time as the Roon update, the confusion is understandable. Our position remains the same, nothin…

I refer to this thread. @xekomi you have been quite active there. I did not read everything again, but Taiko was claiming to improve Roon. Well, Taiko improved maybe their server, but finally it was again just hot air.

I agree that there can be subtle differences. But always if you start asking why, you get either the answer

• It is not measurable, but you can hear it
• Some marketing mumbo-jumbo, no facts
• You do not hear well enough
• Your system is not transparent enough

Another thing which highly irritates me.

• To improve SQ, you always have to spend a lot of money.
• Expensive things make it never worse.
• It is a problem to process digital data in your DAC. But if you add a different switch, you do not have this problem anymore? Why does the manufacturer not apply the solution directly in the DAC?
• Why do these problems only exist in audiophile music reproduction?

I did work as an IT guy for IBM Research Lab and the CERN. Those guys count atoms. They can measure everything. If you can hear it, you can measure it.

Instead of buying a server for 24k or an Ethernet cable for 2k, you should buy an oscilloscope and learn how to use it. You would be surprised how much you can see. Even without being an expert.

[David_Snyder]

@Frudogg

I would like to see an option to reverse absolute phase. Since we already have DSP, maybe this would not be difficult to include.

So easy, in fact, that it’s already there. I save it as a DSP preset to make it convenient to switch it on/off from my phone.

[Xekomi]

@Peter_Bruderer

I did work as an IT guy for IBM Research Lab and the CERN. Those guys count atoms. They can measure everything. If you can hear it, you can measure it.

I’ll add that Team Roon was behind Bloomberg’s trading platform… not the type of environment where you mess around with information loss, either.

[Nyquist]

Why is it that the same people who are so worried about their CPU doing one cycle too many because it impacts sound quality still use upsampling to DSD , dsp and room correction? Isn’t that the very first thing to abandon then?

[Johnny_Ooooops (Listener in Chief)]

@BigAIMc

And I know I’m about to get flamed by the ‘bits r bits’ flat earther regulars who would rather get snarky, rather than god forbid have their sound quality improved even if it involved techniques they are doubtful of.

I think I just got called a flat earther. Which is actually kind of amazing. I feel a bit like if I were a follower of copernicus who was jailed for heresy.

[Billy_k]

I’m really looking forward to roons patented double bit perfect technology that guarantees a 100% sound quality improvement.

[David_Snyder]

@CrystalGipsy

Those who deny there is a difference without listening try it out its free to do so and doesn’t take long to set up and is easy to switch playback between them on the same device.

I like to check myself from time to time as well. The test I’m running at this very moment is (essentially) RAM playback of WAV files local to the Raspberry Pi with the Ethernet cable disconnected and both Wi-Fi and Bluetooth disabled.

I followed these steps to set this up:

1. Convert a familiar album from FLAC to WAV
2. Use scp to copy the converted files to /tmp on the Raspberry Pi 4
3. Install SoX and screen (if not already installed)
4. Kick-off playback in a detached session with: screen -dm play --replay-gain off /tmp/*.wav
5. Logout of the RPi
6. Disconnect the Ethernet cable
7. Enjoy the album

If I find myself more engaged in listening with this style of playback vs. using Roon with volume leveling, DSP, and MQA Core Decoding all disabled, there’s work to be done somewhere.

This test eliminates a few different concerns that could potentially affect sound quality:

• Uncompressing FLAC – Roon does this before streaming music to the endpoint, so for parity, this must be done before copying the files to the RPi4
• Disk I/O – Most Linux distros mount /tmp as tmpfs, a file system which keeps all of its files in virtual memory. Playing files from a tmpfs filesystem minimizes disk I/O to the microSD card
• Network “noise” – People are falling over themselves in other threads to buy $1k audiophile Ethernet switches, optical media converters, AudioQuest Diamond patch cables, and what not. If network noise is a $1k problem, let’s remove it entirely…just to be safe

If you’re using an RPi4 (or similar) endpoint solution, this is the sound quality benchmark for Roon to aspire to. Spoiler alert…

YMMV, but as far as I can tell, they are already there.

Edit: System context in case you thought I had plugged headphones into the 3.5mm jack on the RPi4

• Raspberry Pi 4 in a Flirc passively cooled case with official USB-C power supply
• microSD card flashed with VitOS (RoPieee would work too)
• iFi Audio Mercury USB cable
• iFi Audio micro iUSB3.0 power/signal regenerator with iPower 9V power supply
• iFi Audio Gemini3.0 split USB Cable
• iFi Audio iPurifier3 (Type-A)
• iFi Audio micro iDSD Signature DAC/Amp
• Drop E-MU Wood Series Headphones (Teak cups)

So, not the best system in the world, but not the worst for this kind of test.

[Martin_Kelly (Roon Misfit)]

Not according to Audioquest:

Cinnamon · USB-A to USB-B · Digital Cables · Cables · AudioQuest