I would argue that most DACs don’t do it particularly well. Only some do. Even then, I think you can do it better outside of the DAC where you can more utilize computing power available.
John Siau is selling his products. Of course no one else does it as good as he thinks he does. I can’t stand his amp…and his DACs are just ok.
Thank goodness for that. If I can buy an old Kenwood integrated with HQPlayer upsampling, and get superior sound, I’ll save a bucket load of money. You credibility has been established.
All kidding aside, best of luck with upsampling. For me, it causes fatigue faster than poor acoustics.
Strange. It should reduce fatigue as upsampling, done right, moves the noise farther away from the frequencies we can hear.
I don’t have any problem with your posting and if you think I am the one constantly nagging here.
As long as you are ok being called out for being wrong.
Likewise if I constantly badger a point and someone points out I have no understanding of the topic , then I deserve to be called out too.
All fair.
Someone mentioned something about the thread getting heated. I don’t think that’s the case at all.
There’s just been too many posts that clearly demonstrated the actually opening post has not been understood at all. Posts with arrogant tone too, funnily.
It doesn’t make sense to me for people to chime into a feature request thread and say, sorry OP but this request doesn’t make sense.
Especially if you firstly clearly didn’t understand it and even if you then finally did come around to understanding it, it still doesn’t make sense to me to behave that way in a feature request thread.
In all honesty I have been watching this in dismay because it is as you rightly state…
A feature request thread, not an all out debate.
May I suggest that anyone who has issues with the actual premise of the suggestions contained herein start a new thread to debate it.
Literally I could and maybe should delete half of the posts here.
Although that’s a little late now.
In hindsight it probably should have been moderated a little better at a much earlier stage.
Ah well.
And that’s not aimed at anyone in particular but just a comment in general.
@AceRimmer , correct, of course. My point was very simple: there are numerous other enhancements/features far more useful than an enhancement 99.9% of Roon subscribers can’t (and won’t) ever use. Thanks for your help in these groups. I know it’s very difficult to moderate this stuff.
Great post. Your two Holo Spring 3 PCM show why there’s argument that such high up/over sampling isn’t adding any sonic benefit to a human. Folks see those harmonics in the first of those and think, “gee that looks messy”. Those harmonics are so far below the level of audibility as to be effectively silent, so far below the level of performance of your amplification stage, and so far below the distortion floor of your headphones and loudspeakers as to have no impact. Keep in mind that a quiet bedroom is 30 dBA. Yes, graph 2 is cleaner but it’s a distinction without a difference.
Would be good if you could add those practical details to your post so the audience can put the math into the appropriate context of a human involved system.
Cleaned up the Feature Request thread and created this one specifically for you guys to carry on the debate over the merits of the idea.
It’s not likely that this would take any real resource to implement.
Adding support for 1.536mhz/DSD1024 output should be pretty trivial (may even literally just be altering a limit somewhere).
And even for Roon’s own upsampling, which uses SoX, SoX can do 1.536mhz anyway so that’d just be a case of adding the option to the dropdown.
You subjectively prefer PCM1.5MHz with your Holo hey?
But objectively, no difference between PCM1.5MHz and DSD256 (or higher)?
Or one does measure better?
I personally prefer PCM on the May.
The May effectively has two DACs internally, a discrete R2R PCM converter and a discrete 1-bit DSD converter, which do objectively behave and subjectively sound different.
I swap between the two but overall prefer the R2R/PCM.
On most delta-sigma DACs though I tend to prefer DSD if they have a true DSD conversion capability.
If it’s one that converts back to PCM internally though like ESS then I would use PCM instead.
Can you share some more info on this objective difference? With Holo May specifically.
I used to own a Spring 1, Level 1. My preferences aligned with what yours are for the May, namely that PCM sounded best to my ears. PCM had more audible attack whereas DSD sounded smoother, but lacked sparkle. This applied both when fed native rates and when fed PCM705 and DSD512.
If I’ve used terms there that don’t satisfy an objectivist’s vocabulary then tough, At the end of the day, it’s me who has to listen to it and I don’t listen to graphs. Just so happens that, like the May, the Spring measured OK for an R2R DAC.
Interesting, to me all the pcm filters sound bad to worse. DSD upsamplng at first was thin and brittle, but my bad, After vibration mitigation, true galvanic isolation and appropriate cabling it sounds truly eargasmic. I want more, DSD 256 is wonderful, what does it sound like at DSD 22.4 mhz? I don’t know cuz Roon isn’t cutting the edge right now. Bummer. What’s that HQ Player? Be right there……
What is vibration mitigation?
Depends on component. Amp, pre, source different. Anything with a clock I like roller bearings. But generally speaking, I like to couple the component to a base that is decoupled. The idea is to disallow internal and external vibration from destroying specificity of soundstage.