It’s fun stirring you guys up. I’ll just enjoy my music.
Peace out
It’s fun stirring you guys up. I’ll just enjoy my music.
Peace out
But isn’t saying that exactly what it does mean? Saying you don’t believe in ghosts means that you believe in not ghosts – that is, you believe there are no ghosts.
I am sure there is an opportunity for sound improvements. I guess the opportunity is in having better integration with the service providers Todal Qobuz. Today I guess there is some transformation happening while Roon streams through the APIs of the providers and the transformation data into Roon’s format. If the streaming from the providers is PCM, then it is not bit-perfect at the source.
So, such improvements might require better architectural integration between provider-Roon Core and Roon endpoints (Streamers). Such change is hard to make and beyond one company’s immediate control.
I think the OP’s question is fair. Part of the value proposition for Roon is good sound quality and the company goes to great lengths to ensure that users have the system set up in ways that maximize sound quality - and many of those recommendations shouldn’t matter if bits are bits. From what I gather, Roon hasn’t made any changes to the sound engine or how it handles data transfer, so that’s probably a better answer than giving him a hard time about the question.
But if it’s supposed to be “bit-perfect”, how can it get any better? Unless it’s NOT bit-perfect today.
It can be bettered by overriding the Dac filters , in software.
This is what HQplayer is bringing to the table.
I believe that Audirvana is suppoting SOX to that end.
Also JRiver has a (minor) SOX implementation.
For more info on how HQplayer works? There is a specific category on this forum
Dirk
Assuming your dac supports direct streaming** and it’s true that this in fact does make a difference then it’s simply a matter of up-sampling to your dacs native rate and roon can already do this.
** from my brief look into this, it only truly happens if you aren’t using your dac as a preamp (have volume control) as if it is then the dac still applies its filters.
Yes, put the available filters in e.g. are of higher/better quality than Roon’s own filters.
That is in factthe one and only purpose of HQPlayer and also SOX.
Most of the time, HQplayer does require much stronger CPU power than Roon ( do not know if that is true for SOX as well).
Ifyou really believe that Roon SQ needs to be bettered, test it out in combination with HQPlayer. You could be surprised.
Dirk
While there are definitely more options in HQPlayer, and you may prefer some of these, the filters that are the same, eg. linear or minimum phase are relatively well understood concepts and I doubt have any real functional difference between implementations.
I think one reason many still use HQplayer is they had purchased it before roon had native DSP, and why not let people use what they have already purchased and are used to.
I use Squeeze Server which is freeware as bundled (and tweaked) with Antipodes gear. It has a more natural and organic sound than Roon, which in my opinion sounds like “hifi” in comparison. I’ve heard that other DLNA/upnp players sound better as well. I don’t use any DSP. Of course, they don’t have anywhere near the GUI and info that Roon serves. It’s all a matter of priorities.
Yeah, what Roon really needs is more demanding specs. Obsoleting all the Nucleii in that same move would make it even better…
Looking at it as a non-user, what I’d like to see from HQPlayer is a clear algorithm quality / tradeoffs matrix rather than ad-hoc answers from Jussi (who’re truly enlightening and fascinating), alongside reference builds for given use cases. There’s definitely a space for two-box solutions, but it’s something that should remain on Signalyst’s plate, not RoonLabs’.
I’d also like measurements at dac output showing how upsampling affects the sound.
As usual with these claims: measurements or, well…
As @sbr rightfully put it,
There are good reasons to why you may be right in some cases (DSP or volume levelling come to mind), but it is extremely likely that none have to do with the brand of your server. If it does, then the people who make it should publish evidence that their approach is hard science, and not salesmanship.
Just an opinion, dude. Thankfully I have no need to convince you or anyone else.
Of course ! And Antipodes (and the others) make great hardware, don’t get me wrong there… it’s mostly that it’s a case where demanding evidence could save you a couple of thousand if you need to update your server down the line… I’m sorry if I didn’t make it extremely clear that in my mind, the onus is on the sellers, and not the buyers.
I’ve been a beta tester for a couple of well known manufacturers, both hardware and software.
The complexity of the software used to play back files and and whether the files are played from RAM memory or cache makes a difference to the sound quality of the playback software. In general the busier the software is doing yet file playback the more compromised the sound quality gets. For ROON to achieve all the magic it does it causes your PC or streamer to work harder. The files are 1s and 0s of course but the busier the computer the more “software jitter” according to Paul McGowan of PS Audio. Software jitter is what affects the sound quality.
My understanding is the addition of a RAM cache or other method of sending the files to RAM for playback makes quite a difference.
ROON’s sound quality is very good. Not the best. It’s very easy to hear the difference if one has an Auralic Streamer. It’s easy to play a file through the ROON Ready method or use Auralic’s Lightning DS. Lightning DS sounds just a bit more real, dynamic, and lively by comparison.
If one wants to try HQPlayer it’s easy to discern the difference as well. HQPlayer offers free trials.
Don’t get me wrong I love ROON. I’m sure all the new changes will be great fun to use. I’m a lifetime user.
I do hope Wizards of ROON have made some strides to better the sound quality too.
I know several users both lifetime and yearly subscribers who don’t use ROON much anymore because other software sounds better.
For those worried that ROON will be around for their Lifetime, I think ROON has to address sound quality.
Audiophiles are a fickle lot and are always looking for away to make their systems sound better. Thus the continuous turnover in equipment that some pursue. I don’t believe software will be immune to this phenomenon.
But of course if you believe you can’t understand the above, discern these differences or they mean nothing to you, then please disregard the above and don’t read it.
Oops, if read this far you did read it. Drat. So sorry to bother and waste your time.
And there you go - the guy whose company is trying to build a competitor is spreading FUD, fancy that…
He likes to pose with measurement gear, and he’s said he uses it, so where are his measurements and his null tests to back up his claims ? Really. I don’t mean this in an assholeish way, I’m honestly, genuinely curious.
HQPlayer actually does something to the bitstream. It isn’t just a transport, there’s the whole filtering deal going on, and I’ll completely buy that that’s audible in a number of cases.
This certainly must be measurable. Why isn’t Auralic providing measurements, and why haven’t Auralic customers demanded that they be provided if Auralic themselves have supported the belief ? Because if you’re right, that means one of two things: there’s a problem with RAAT itself, or there’s a problem with their RAAT implementation.
Another McGowan myth.
See this discussion around memory playback. Does not really work for Roon and the core/endpoint/RAAT architecture means you avoid most of the problems it might solve and all the loss of functionality/usability it would cause.
You know I think it’s a testament to the folks who code Roon that they don’t make any claims of sound improvement. No one could quantify it, but they could end the claims of less than optimum sound. They could then also sit around and laugh at the “expert” description of the improvements.
I’m very familiar with all the posts here. I read most of them and understand why it’s difficult for the Wizards of ROON to do with software versions 1.0 -1.7. I was hoping that with the rewrite of version 1.8 the Wizards might have addressed these problems. As brilliant as the ROON Wizards are (not being sarcastic, I love ROON!) I was hoping they might have rewritten this portion as well.
It’s been my experience and many others that the busier the program is the more the impact there can be. It’s called software jitter.
Problem is this isn’t really true, a statement about an objective thing like:
Thing X does something different than thing Y isn’t a matter of opinion. It’s either true, and provably so or it isn’t true.
Obviously this assumes the different things are playing the same game, eg. bit perfect audio playback.