Roon Optimization Guide For Increased Sound Quality
Feb 21, 2024 at 10:32 AM Post #91 of 132
As for why Roon sounds better than Qobuz to you, it might be some of the settings you have set up. I mean, yes, with like-for-like material, Qobuz and Roon do sound marginally different under the right conditions, but they don't sound *that* different. More likely is that you have upscaling on or that Roon is outputting at a higher volume or something similar.

On a broader note, Roon is the single-most significant part of all of my audio setups and gear. I use it for pretty much everything audio-related (if only Roon Arc didn't crash every twenty seconds on my phone. Alas). The DSP tools have a bit of a learning curve if PEQ and whatnot are new to you, but once you get the hang of it, you'll swear by it. And no other service or music player that I've tried has the same ease-of-use as Roon in terms of music playback, browsing, and discovery.

The way to get a discount on a lifetime sub is to get a time machine. It was a lot cheaper early in its existence and the price has gone up ever since. If you're looking to save money on a lifetime sub, your best bet is to get it sooner than later. I pay annually since I'm never confident that any subscription-based service will last that long. With Roon's recent buyout and the growing threat of a crash across all streaming media, I really don't know what to expect in terms of the service's lifespan.
I used all the settings for Roon that the OP outlined. It's not a tremendous difference, but everything is just a bit more clear to me. I'm not really sure why that is and I've heard it each time switching back and forth.

I gave Audirvana a try too. Initially I thought it sounded even better, staging seemed a bit more whimsical and fun. In hindsight it sounds brighter and less warm, everything sounded super processed and artificial almost. So strange I've never dove in this deep to audio, every little thing makes a difference. Just shows how important it is to pick what sounds best to you

Can anyone help explain how sample rate conversion works or at least guide me to something to help understand? I see under the PCA you can convert sample rates, there also a bunch of stuff on PCM to DSD which I don't understand.

I've never gone this deep before as said and it's all so confusing lol
 
Last edited:
Feb 21, 2024 at 10:54 AM Post #92 of 132
Far bigger brains here to explain deets but up sampling to higher rates and in megabytes say 196 32 or 386 32 may improve sq depending on the server and dac. For me it’s antipodes and chord so max pcm rate. Converting to dsd is also dependant on your music server and dac if it’s preferable. For chord it auto changes dsd to pcm so upsampling to dsd makes little sense.
but the great thing is you can try out all options incrementally and see what works for you. But again I can’t explain squat nor direct you to something, my trials are just that, trial.
 
Feb 21, 2024 at 11:02 AM Post #93 of 132
Far bigger brains here to explain deets but up sampling to higher rates and in megabytes say 196 32 or 386 32 may improve sq depending on the server and dac. For me it’s antipodes and chord so max pcm rate. Converting to dsd is also dependant on your music server and dac if it’s preferable. For chord it auto changes dsd to pcm so upsampling to dsd makes little sense.
but the great thing is you can try out all options incrementally and see what works for you. But again I can’t explain squat nor direct you to something, my trials are just that, trial.
I appreciate that at the least. Can you explain "servers" a bit?
 
Feb 21, 2024 at 11:07 AM Post #94 of 132
Music servers like antipodes audio, they stream as well. They are roon core. Course you can use a Mac/pc/ras pi with lesser results but better than iTunes.
 
Feb 21, 2024 at 11:57 AM Post #95 of 132
This guide assumes you have already downloaded and installed Roon Core/Server 2.0 on your source and Roon Endpoint version 2.0 on a Mac or PC.
  • Access the settings menu by clicking on the hamburger menu (three horizontal lines) on the top left of the Roon app. Configure the following:
    • Disable the default music folder (not the folder where your music is stored. Roon sometimes automatically adds the default system music folder and this is the one we want to disable)
      • Settings > Storage > kebab menu (three vertical dots) > Disable
    • Set Background audio analysis to Off
      • Settings > Library > Background audio analysis > drop down menu > OFF
      • Note: changing this setting is vitally important for SQ, but also a PITA. You'll need to turn this back on after adding new media to your library then turn it off again.
    • Set On-Demand audio analysis to Off
      • Settings > Library > On-Demand audio analysis > drop down menu > OFF
      • Note: changing this setting is vitally important for SQ, but also a PITA. You'll need to turn this back on after adding new media to your library then turn it off again.
    • Disable all audio zones and devices except for the upstream device (DAC/Network Player) you will use to connect to your Roon Core.
      • Settings > Audio > Connected to Core > DAC/Network Player > Enable
      • Note: ASIO drivers for Windows and ASLA for Linux tend to yield the best audio performance. Though I have heard instances of WASAPI performing just as well in Windows with the right software.
    • Audio Device Settings
      • Settings > Audio > Connected to Core > DAC/Network Player > cogwheel menu
        • DSD playback strategy > Native or DSD over PCM (DoP)
        • MQA capabilities > off
        • Volume Control > Fixed
        • Resync Delay > 0ms
          • Note: if you click play and the first second or so of a track is cut off, increase in 50ms increments until it plays. Some DACs need a few milliseconds to sync with the source before initializing playback.
      • Show Advanced > Drop down menu >
        • Set your DAC's max sample rate for PCM
        • Set your DAC's native bits per sample for PCM (bit depth)
        • Set your DAC's max sample rate for DSD
        • Enable MQA decoder > No
        • Use maximum Buffer size > No
        • Power of 2 Hardware Buffer Size > No or 0
    • Disable DSP and remove filters (don't worry, you can always add these back).
      • Home (main music playback bar at the bottom) > Speaker Icon (lower right) > Click the Sine Wave/hairpin looking thingy (this should open the DSP menu)
        • From the DSP menu:
          • Headroom Management > Disabled
          • Sample rate Conversion > Disabled
          • Parametric EQ > Remove Filter (top right)
          • Convolution > Remove Filter
          • Crossfeed > Remove Filter
          • Audeze Presets > Remove Filter
          • Note: it is important to remove these filters, not just disable them.
The rest of the optimizations will be done at the hardware, BIOS and operating system level. That will be covered in the "Building a Custom Digital Source" thread.

You can also use these optimizations as a litmus test of sorts. If you change these settings and can't hear a difference in your rig. Leave this thread and don't give "optimizing" Roon another thought.

Hope this helps!

-lj
Late to the party, but thank you very much for this. SQ definitely enhanced. Particularly, removing already disabled filters. I did this last in the middle of listening to a DSD recording of Vivaldi’s 4 Seasons. Immediately obvious improvement.
 
Feb 21, 2024 at 12:14 PM Post #96 of 132
I used all the settings for Roon that the OP outlined. It's not a tremendous difference, but everything is just a bit more clear to me. I'm not really sure why that is and I've heard it each time switching back and forth.
Cool. Just go with it then. You might discover an explanation for the difference later down the road as you get more familiar with Roon and your gear.

I gave Audirvana a try too. Initially I thought it sounded even better, staging seemed a bit more whimsical and fun. In hindsight it sounds brighter and less warm, everything sounded super processed and artificial almost. So strange I've never dove in this deep to audio, every little thing makes a difference. Just shows how important it is to pick what sounds best to you
Yup. While the general advice is to focus on the big things first (headphones, amp, dac), then start the endless process of tweaking "every little thing," I can't fight my impulse to keep tweaking, adjusting, and modding the details. It's fun.

Can anyone help explain how sample rate conversion works or at least guide me to something to help understand? I see under the PCA you can convert sample rates, there also a bunch of stuff on PCM to DSD which I don't understand.
I can't tell you how it works, but I will say that there are two big schools of thought with respect to upsampling. One group doesn't upsample and just makes sure Roon puts out the same sample rate that it receives from the music files. The other group upsamples to their DAC's ideal/highest sampling rate. The nice thing is that neither choice costs you anything but time and patience. Just listen and decide for yourself. For me, I found upsampling to produce a glassier, more liquid sound for some albums, but it ultimately sounded too artificial and "crystalline" across the majority of my music. So I just do native sample rates now.

I've never gone this deep before as said and it's all so confusing lol
Totally. The need to learn, discover, and experience more in this hobby is what keeps me going.
 
Feb 21, 2024 at 12:46 PM Post #97 of 132
For anyone with a squeezlite capable streamer, try squeezlite through roon over raat. It's a fairly stout difference for such a small change so its worth giving a shot
 
Feb 21, 2024 at 1:01 PM Post #98 of 132
Cool. Just go with it then. You might discover an explanation for the difference later down the road as you get more familiar with Roon and your gear.


Yup. While the general advice is to focus on the big things first (headphones, amp, dac), then start the endless process of tweaking "every little thing," I can't fight my impulse to keep tweaking, adjusting, and modding the details. It's fun.


I can't tell you how it works, but I will say that there are two big schools of thought with respect to upsampling. One group doesn't upsample and just makes sure Roon puts out the same sample rate that it receives from the music files. The other group upsamples to their DAC's ideal/highest sampling rate. The nice thing is that neither choice costs you anything but time and patience. Just listen and decide for yourself. For me, I found upsampling to produce a glassier, more liquid sound for some albums, but it ultimately sounded too artificial and "crystalline" across the majority of my music. So I just do native sample rates now.


Totally. The need to learn, discover, and experience more in this hobby is what keeps me going.
Good info appreciate the response. I suppose you're right it's time to just start changing things up and see how I like it! Got my chain out together, waiting on my AudioQuest cable to come in today and Friday I should have my Aryas. That marks my set up as "complete". So naturally I'm going to spend time messing with the source 😁

What about pushing out higher sample rates with NOS mode? Or would filters work better in that regard? Not that I know anything about either.
 
Feb 21, 2024 at 1:11 PM Post #99 of 132
Good info appreciate the response. I suppose you're right it's time to just start changing things up and see how I like it! Got my chain out together, waiting on my AudioQuest cable to come in today and Friday I should have my Aryas. That marks my set up as "complete". So naturally I'm going to spend time messing with the source 😁
At least you have the self-awareness to put the word "complete" in scare quotes. :beyersmile:

What about pushing out higher sample rates with NOS mode? Or would filters work better in that regard? Not that I know anything about either.
Depends on... well, lots of things. Easy enough to try it for yourself and see. There's no general advice on this front. In my experience, DAC filters make only a marginal difference to overall sound quality, but as you've already illustrated, everyone's sense for what is "marginal" and what is "major" is rather different.
 
Feb 21, 2024 at 11:40 PM Post #100 of 132
Ohh yea Roon is incredible. I love how I can just set my laptop down as the server and use my tablet or phone to control everything. Sounds great, the UI is super intuitive and easy to navigate, I'm really loving it. I see what you mean by just play around and see how it shapes the sound
 
Last edited:
Mar 2, 2024 at 2:03 AM Post #101 of 132
@littlej0e thanks for this great thread, it made a subtle beneficial difference in my setup, noting I don't have a local library and output via HQP server which limits the Roon audio settings in play.

<Edited for brevity>

My question for you and others on this thread who use Roon Remote is have you experimented with optimisations (either software settings and physical) of your Roon Remote device? As I couldn't see much on this thread or elsewhere on HF re this point.

I ask as I found repeatable 'spooky action at a distance' improvements with these tweaks to the Roon Remote device:
- iPad Air superior to Samsung phone
- turning on airplane mode (LTE + BT OFF), with only wifi left on
- physically damping the back of the iPad /phone with piezo-electric materials (an fo.Q SH-22K A5 polymer sheet and/or a zip bag with say 5-10 grams of Rochelle Salt powder) and/or 3K carbon fibre fabric folded over a few times, in each case placed underneath the device.

I found the more damping types in combination the better. Have found this previously with Bluetooth to my iFi xDSD Gryphon.

These tweaks brought a more settled sound, the cliched but very real quieter backround allowing greater separation and focus, particularly a deeper soundstage with more decay and sense of acoustic space, and a little greater dynamic scale. Removing the damping and/or turning off Airplane mode the reverse happens, it all gets a little congested, flatter.
 
Last edited:
Mar 2, 2024 at 2:11 PM Post #102 of 132
@littlej0e thanks for this great thread, it made a subtle beneficial difference in my setup, noting I don't have a local library and output via HQP server which limits the Roon audio settings in play.

Roon Remote & spooky action at a distance
My question for you and others on this thread who use Roon Remote is have you experimented with optimisations (either software settings and physical) of your Roon Remote device? As I couldn't see much on this thread or elsewhere on HF re this point.

I ask as I found last year repeatable spooky action at a distance improvements with the sound quality of Bluetooth streaming with multiple send (stream) and receive (playback) devices (was originally in a portable context with my Gryphon, but broadened the test to a range of components). But till yesterday I'd not bothered to try same for Roon Remote devices for a number of reasons - I mean it's not a continuous audio stream from the client device like BT or Airplay after all, just an instruction to Roon Core to play/pause etc, with maybe a packet or two to keep it in sync, why/how would it make any appreciable sifference, yadda, yadda. Anyway I finally did some careful AB'ing with the two devices I have Roon Remote on, a Samsung phone and an iPad Air M1 LTE, incl the range of tweaks as below.

The outcome was, let's say, inconvenient - not what I was hoping for. In short these measures all improve the improve the sound, and not subtly:
- iPad superior to Samsung phone
- turning on airplane mode (LTE + BT OFF), with only wifi left on
- physically damping the back of the iPad with piezo-electric materials (an fo.Q SH-22K A5 polymer sheet and/or a zip bag with say 5-10 grams of Rochelle Salt powder) and/or 3K carbon fibre fabric folded over a few times.

I found more damping types in combination the better whilst CF fabric alone could overdamp.

The sonic delta: a more settled sound, the cliched but very real quieter backround allowing greater separation and focus, particularly a deeper soundstage with more decay and sense of acoustic space, and greater dynamic scale. Removing the damping and/or turning off Airplane mode the reverse happens, it all gets a little congested, flatter.

The piezo-electric damping idea dates back to my BT findings last year (and prior to that a lot of experientation of piezo-electric damping materials on/around ground boxes, cables and components) incredibly not merely the same character of effect with Roon but also similar scale, notwithstanding the fact Roon Remote likely only sends occassional instruction or sync packets to Roon Core vs the BT continuous audio stream. Sigh. It really seems EVERYTHING matters.

I do realise the fo.Q observation is not super helpful to folk in this thread as they closed up shop last year so it's now unobtainium. However the other materials I refer are also very effective and easily accessible - hopefully some of you may even have them to hand to try. For those who use Roon Remote (maybe a minority given this is Headfi) give it a go and let me know what you find.

And more generally apologies if all this seems a bit out there/woo-woo or just off-topic for a software settings thread, particularly for folk who've yet to venture down the damping/grounding rabbit hole. But if you can suspend disbelief long enough to try some of this, and provided you have a reasonably resolving system, I will wager you will end up as bemused as I am. It's damn inconvenient is what it is! (The fo.Q sheet was already deployed elsewhere in my system, I'm going to have to rob Peter to pay Paul dammit 😅)

Cheers
Jake
I mean, it makes no sense why the remote device should matter at all to the quality of the stream (since your remotes just tell your server what to stream), but stranger things have been known to happen in audio. This could be down to the Roon settings you have on your individual remotes.
 
Mar 2, 2024 at 2:24 PM Post #104 of 132
Similar things have been discussed in the Taiko thread over on WBF. IIRC they prefer the ipad too, possibly narrowed down to a specific generation.
But then what happens if you pause the stream on another remote, skip to the next track on a third remote, and add a new song to the queue on a fourth device? Which of these remote devices would determine the streaming quality? Believe it or not, this is not a far-fetched scenario. I regularly use about four or five different remote devices in my house.

My hunch is something else is happening. Possible it's something to do with your home network environment and how your various remotes and server are all connected to one another, to your network, and to your ISP. If someone can isolate what's happening to improve the streaming quality, it's possible it can be reproduced intentionally and spread to all of your remotes.
 
Last edited:
Mar 2, 2024 at 2:38 PM Post #105 of 132
Similar things have been discussed in the Taiko thread over on WBF. IIRC they prefer the ipad too, possibly narrowed down to a specific generation.
Thanks for the link, good to know it's not just me, and whaddya know of all the six devices tried the OP preferred the iPad Air 5 which is my model. Not the worst company either as IIRC he is a high end dealer in HK with experience with very good gear.

Edit - here's the link to the thread I believe you may be referring, and my bad the OP CKKeung isn't a dealer but does have an exotic system (Magico/MSB/Taiko) and connections to a bunch of high-end distributors.

https://www.whatsbestforum.com/thre...e-for-roon-ipads-vs-iphones-vs-tablets.36366/

But then what happens if you pause the stream on another remote, skip to the next track on a third remote, and add a new song to the queue on a fourth device? Which of these remote devices would determine the streaming quality? Believe it or not, this is not a far-fetched scenario. I regularly use about four or five different remote devices in my house. My hunch is something else is happening.
I shared your scepticism that any improvement would be possible for these reasons, so didn't bother testing the equivalent of what I'd observed with Bluetooth last year till now. I can only offer a generalised theory or two to maybe explain what I'm hearing: that there is perhaps a higher frequency packet flow between the Remote and Core than we might assume and Core is somehow rather sensitive indeed to some characteristics of this flow. Not to mention that the level of noise/vibration in the Remote device somehow affects this flow.
 
Last edited:

Users who are viewing this thread

Back
Top