I used to be a Roon subscriber and yesterday just reactivated my account and now on trial period. I did a A-B test on playing a song using different software; one with Roon and the other with Tidal. All EQ were disabled during Roon playback. I found that song playback by Roon was less dynamic, clarity and separation as compared to the Tidal playback. In short, Tidal is better in sound quality. May I know what I have not done right that caused this issue ? Or this is the characteristic of Roon ? I really like the Roon UI and hope this will not be the reason for me to give up Roon once again. Appreciate your help. Thank you.
What you hear is that the output of the Tidal app is a bit louder. Push up the volume in Roon a bit more and it will sound the same.
In an A/B-Test the first thing to do is matching the volumes with a measuring device.
My first thought was that maybe the playback through Tidal was “enhanced” in some way. Increased volume is one obvious means.
Whose to say the more accurate sound is not through Roon? Sometimes I like to enhance the sound with DSP. That doesn’t mean straight through Roon is deficient.
I see the question of roon sound quality appearing more frequently. It is my experience it sounds different than Squeeze, Audirvana and HQPlayer feed thru. The roon user guru’s will say bits are bits and volume leveling and I can’t argue that but something makes roon sound more constricted. Fortunately it’s operating system is so much more advanced. When doing an in depth listening session I usually switch to Lyrion (squeeze). A couple years ago there was a post from roon in an update that said they had optimized the audio path so I am hopeful they can address it now. Software gets bloated with advancements and optimization may be needed. For now when using roon I send it’s steam to hqplayer and then to an NAA. That helps a bit.
Such comparisons need clarity about data paths from source(s) to DAC. In the case of the dCS Rossini and other Roon Ready DACs that also accept other network sources, we don’t know anything about possible differences in the processing of different sources. One potential difference is in how the protocol stream delivery for different stream types interacts with internal realtime processing for feeding the D2A circuitry. Having worked with realtime code for streaming audio in a distant past, I can say that it’s easy to have subtle, non-fatal bugs or inefficiencies that could affect the realtime properties of the system. And I also have multiple (proprietary, not sharable) “war” stories over decades of subtle synchronization defects in hardware and software leading to inconsistent performance. Processing correctness in this space is really, really hard, and the very few engineers who are excellent at it are in very high demand from tech, defense, and instrumentation giants with much deeper pockets and career growth opportunities than any specialty audio company.
Have you reviewed and implemented the basics? Separate Server and Output devices. I prefer HQPlayer to be my ‘processor’ and output so I have 3 devices. Roonserver does very little ‘work’ in my setup.
RoonServer on Windows 11 (non-dedicated but works fine at very low power, high power savings settings) > dedicated high power HQPlayer machine running HQPOS for upsampling and convolution > UPBoard gateway mini device running HQP NAA. RoonServer becomes only an orchestrator handing off the hard work to HQP but this can also be done with Bridge without additional expense to listen to the quality of separate devices handling the audio flux.