Funny that you brought this up since I’ve been playing with Diretta a bit just this past week. In “audiophile” terms, I would not describe it as costly, but it’s not free; I paid €100 for my Diretta Target license. Money well spent based on what I think I’m hearing so far.
This question has been endlessly debated on these forums and many others for years, but that’s ok because it was a new one to me at one point too—until a few wise and knowledgeable folks tore me away from the clutches of Head-Fi, explained some basic concepts to me that after a period of confusion suddenly made perfect sense, and saved me a whole ton of time and money (thankfully). So I don’t mind paying it forward, if others on here don’t take offense.
It’s actually very simple stuff, basic computational math—and it will save you so much time, money and anxiety if you just acknowledge that our ears are very tricky things, and abandon all the standard audiophile adjectives they hold aloft on the fancy forums when it comes to the sound quality of software such as Roon. Why?
Because this is digital audio we’re working with, not analog. Roon provides a bit perfect audio stream to your DAC, with a suite of handy (digital) adjustments available to you to apply before it arrives there to suit your tastes, if you so choose. Nothing less, nothing more.
It’s a bunch of ones and zeros, and has no “sound” by definition—until it gets converted to sound waves that move air. The “quality” it provides you is determined solely by the quality of the original master that was sampled into those ones and zeros at a particular frequency and bit depth, and stored in a FLAC file for you to stream. The rest is up to your amplifier and speakers, which are the only two components in a digital audio setup that add any “qualities” to the sound—in ascending order. This assumes that you’re using a transparent DAC, which basically means any delta sigma unit that was produced within the last seven or eight years (probably more than that).
There are a lot of folks who will argue vehemently that there’s more to it than that, many of whom want to sell you some audio gear, and many more still who want you to buy the same gear they did so they can feel justified in how much they spent. But you’re already on the right track right here with Roon—because it offers you access to a near limitless supply of digital, lossless, soundless files in a nifty package at a reasonable price—waiting for you to convert them back into music. You won’t find any better sounding soundless files to stream anywhere else at any price. Enjoy, and peace
So, I had a DAC (PS Audio DirectStream) that had a PCM test track I could play and the DAC would indicate on its display if the track was received in bit perfect form. If one bit were off it would not display the bit perfect icon on the display.
You could play this PCM test track from any player software on any server and using a direct connection or a streaming over a network solution. Of course, you had to set the player up so no upsampling or DSP would be done at any time.
Would you agree that if, over a network, Roon, Audirvāna, HQPlayer, and JRiver, got the music file to the DAC and the DAC display showed the file got there in bit perfect form, that the DAC would output the same analog signal regardless of source?
Because I did this exact test and none of the players sounded different than any other.
This told me that RAAT versus whatever Audirvāna, HQPlayer, and JRiver use does not make one bit (literally) of difference to what PCM data the DAC receives over the network.
Now, when you started doing upsampling and DSP, there was a marked difference in sound quality. But that down to sampling and DSP engines and not what transmission protocol is used.
I wish I had a listening environment where these were the considerations. Outside traffic, parking lot noise, neighboring air conditioners, etc have a way bigger influence on my listening than top end equipment. I don’t need a better DAC or amp, A few hundred dollars of modern and used equipment serves my needs well.
I need a landlord to install double windows in my apartment and quiet AC in my neighbor’s space.
It looks like you can’t actually buy them with that finish. In any event, I’m not currently looking to replace my Sasha ll’s so for me, the price is moot.
Sorry bud; I am in Buenos Aires and having a multitude of issues between my Airbnb and my Chase Bank with bogus “fraud alerts” that were not based on any credible reasons so you caught me in a foul mood. No excuse for my harsh tone but regardless, apologies. Peace
My original comment on your post was also a little sharper than it needed to be so I owe you an apology as well. No excuses here either. Plus, I was wrong so there’s that…
All good Dave; I confess I’m guilty of being a little sensitive when I see false information which began when our politics took a turn for the worse with all of the hate and disinformation permeating our culture. So I confess I am a bit sensitive to "truth and non truth from a pure objective perspective. I don’t pretend to be George Washington, but I’d like to think that I do my best to never tell a lie! Haha
Like you, I think that my Diretta implementation under Audiolinux for Roon server is sounding better than standard RAAT. Now I’d like to hear from other delusionals who have taken the time to experiment with different configuration settings to further tweak performance.
You have been a leader on this forum exploring ways to coax great SQ from the lowly Raspberry Pi, including the use of free VitOS software a few years ago, and more recently posting a helpful guide for Diretta.
Let’s start a new discussion among open-minded hobbyists pursuing this ever-evolving topic. Time for another fresh start!
If you think you hear a difference, you should definitely make an effort to quantify what that difference is. By looking at the frequency spectrum between two configurations/chains/sources, which you can do using a relatively cheap, calibrated microphone and a good ADC (or direct to ADC for DACs), you can get a handle on where in the spectrum changes are taking place. This leads directly to subjective impressions like “flabby bass,” “good instrument separation,” and so on. With that data in hand, you can better improve your listening experience!
Hey Mark! I did buy a miniDSP UMIK-1 calibrated mic to record frequency sweeps around my listening position for room correction, and each of my Meridian speakers came with a high quality outboard ADC, so I could generate that data. Hmmm…
Great, if you are doing room correction you have a flavor of the kinds of measurements that can lead to insights about SQ and the impact of changes like digital transports (likely none) versus room treatments (likely lots). I use DIRAC in one home theater, REW in a listening space, preconfigured OPRA profiles for headphones, and Audyssey in another AVR setup.
A critical component of open-minded SQ discussions is the baseline truth of measurements!
[In my opinion] there is a difference in digital sound between players (Roon, audirvana, Qobuz connect), differences between hardware that Roon runs on, dacs, usb vs i2s vs Ethernet.
Audirvana on a Mac sound d better than Roon on the same mac. Roon on Linux using Mac hardware sounds better than Roon on a Mac running osx. Bypassing Roon/audirvana and using Qobuz connect directly to my streamer sounds the best. Qobuz connect has a terrible interface so that’s why I still use Roon plus I have thousands of cds ripped to the server.
There is no difference in digital sound between players that just make their function while you don’t use DSP. Once you use DSP, things change… for better or for worse. Until then: No difference.