Where are improvements to sound quality?

Err, I don’t like your timbre :wink:

It sounds much clearer to me with the Convolution DSP to do room correction.

For more amusement, why don’t you tell us your theory on why these CD transports connected directly to the DAC sound better than streaming?

Let me discount some things first:

A laser reading from a CD is NOT better than a SSD or HDD being read.

A CD transports generates more noise than a quality streamer.

Using a network provides better isolation, not worse.

I can’t wait to read what you have to say…

1 Like

One theory is perhaps that the CD/Transport builders used the best analog for their inspiration. That’s what they were trying to out do. Of course they never beat the best analog rigs. But they worked hard at it.

I’ve found that too but with headphones. Using the convolution filters (from Github) for Sennheiser HD650 is better than anything I managed to achieve with parametric eq.

bye deleted

Yes, wrote about that in my initial post, the PEQ in Roon is not as good as it could be, and using convolution is better. But since I posted this, I discovered that convolution in HQPlayer sounds slightly more clear than in Roon so convolution in Roon isn’t perfect either.

What does “best sound quality attainable” mean, though? Most transparent? Most like the recording? Most like the original performance? Most pleasing?

1 Like

Best sound quality obtainable.

Easily heard to be as good as in most audiophile categories and clearly better in others. I can listen for certain characteristics but I can usually hear two versions of something and know which is musically the best. I try to compare to one version to another and then think about which was sounds most like the real thing. As compared to what instruments and voices sound like live.

@Speed_Racer Why such hate of CD Players and transports? It’s not the technology but the way it is implemented that matters. My old Accuphase CDP still outperforms my Roon/PSA DSD streaming setup but I am just too lazy to use it most of the time.

Plus, pretty well every CD transport uses a (usually) well implemented SP/DIF interface, while the SP/DIF interface on many streamers is an afterthought, so depending on the DAC the CD transport may give it an easier job to do and thus sound better.

I don’t hate them. If people want you use CD players and CD transports, more power to them. I am saying that quality transports aren’t better than a quality streaming setup. There is no technical reason why a transport would be inherently better than a streaming setup.

1 Like

Why would I want to limit myself to to an S/PDIF interface with my streaming setup?

Regardless, yes, I am sure some transports do a better job with S/PDIF than some streamers that use that technology. But that does not make transports inherently better than streaming setups.


In my Meridian set up, I never think ‘oooo this will sound better on my 596’, never.
Roon streaming to the 218 DSP SE set up is very satisfying to my ears I have no upgrade hankering at all. There is nothing wrong with Roon’s audio quality here, it think it’s the endpoints that count.


It’s all about the speakers…wish I had some speakers.

1 Like

In audio tests, the problem is that the human ear is very poor at recognizing small differences suring short samples. Sample length must be at least 20-30 minutes each for the human auditory system to lock onto and recognize the differences. Than makes AB testing impractical in this case. When these tests actually support a difference, it is only because the difference are large enough to be recognized in a minute or two.

Well, that claim is not an established fact.

But let’s just accept it for purposes of discussion. You are saying that listening tests extending over many minutes, hours or days are capable of detecting subtle but perceptually important differences in sound quality. The problem with such testing (at a store, in your home) is that you are influenced by so many subjective phenomena in your normal daily experience while evaluating “B”, and your memory of the prior “A” experience is also modified. For example your interactions with your friends or thoughts about your next Roon posting as you evaluate the new $50k chocolate-covered nuclear-powered amp you are considering buying. (<== parody to help make the point.) That means you do not know how you ultimately came to your conclusion (i.e. you do not know what you are actually testing). If you want an objective measure, it has to be a controlled experiment, in this case double-blind testing. If achieving that is so impractical as to be effectively impossible, then we can comfortably agree that discussions of sound quality are discussions of opinions. If so, and that is what you want, go for it! For myself, I would refer you to considering the early history and ultimate outcome of mentalism in psychology (late 1800s to early 1900s), but otherwise have no interest in such conversations (except perhaps as a sort of bystander).


This is what I miss about being able to integrate vst plugins in other media players.
Fabfilter Pro-Q sounded better to me when I demo’d it vs many stock EQs

I think the problem is that playback of a file requires the participation of a computer. It’s unavoidable. Computers are noisy critters. A CD transport is a much simpler collection of electronics. Look at what is being done in the ultra-high-end of servers and endpoints. For example, Pink Faun is selling tightly engineered oven-controlled-crystal-oscillators to achieve lower levels of noise in the motherboard and output cards. Software like Euphony is controlling CPU cores to slow them down to the right speed to process audio as electrically quietly as possible.

The number of things that can go wrong in a file-based playback device (or chain of devices) is huge compared to a disc transport.

That Auralic streamer of yours is probably hard to beat below the $10k mark (barring DIY options). Maybe the dCS Network Streamer? What I recently did might be a step backwards for you, but when I switched from ROCK to Audiolinux (running Roon as an endpoint only) on my NUC, the sound quality improvement was significant. I have no clue how well it would do against your transport, but I’ve been content ever since.

There’s no doubt DSP can materially affect what you hear. I just don’t think it should be Roon’s primary focus. User who complain about the bit-perfect path from Roon versus the bit perfect path from JRiver are a whole different kettle of fish. So that’s the first thing that should be de-confounded. If such differences do exist it’s almost certainly a function of how the software interacts with the users equipment, which is not reasonably predictable under all circumstances.

Dirac integration is one thing (and one user actually got the math geek joke I made about it). It should be easy for users to incorporate the software they choose to do room EQ if it’s not an undue burden… But Roon should focus on perfecting the user experience, not obsessing over why one DSP over another sounds “veiled”.