What do most people use HQ player for?

I think most people who use HQPlayer with Roon use it because they think they can achieve better sound using HQP as Roon’s playback engine than they can using Roon alone. If you want to check this out, you can download a trial of HQPlayer and see for yourself whether you think it’s worthwhile.

A few things you might want to know:

HQP’s user interface is somewhat impenetrable at first glance. Most likely, you will need to read the manual to get it set up. To minimize problems, I would get HQP up and running to your satisfaction by itself first and then proceed to set up the integration with Roon, which is pretty easy if you follow these instructions.

If you really don’t want to fiddle around with HQP’s settings and try its different filters, I would suggest that HQP isn’t for you. It’s worth noting, though, that once you have HQP set up to your satisfaction and integrated with Roon, you don’t really have to do much with HQP other than launch it (although for those so inclined, you can try different filter settings more or less endlessly).

Many people (including me) think that the sound quality achievable with HQP is great. Others, however, are not so impressed. My personal opinion is that this is system dependent to a large degree, especially with regard to one’s DAC. Again, you would have to fool with the HQP trial to see which group you fall into.


Thanks for the replies folks! And the endorsement dbtom2!

I’m going to reply specific to my needs, but I would encourage continued contribution to the conversation, for - as Tom noted - this is a question worth asking. And a good set of replies is likely helpful to the community overall. Thanks!

Indeed. It sounds fairly specific a particular need.

[quote=“RBM, post:4, topic:7393”]
HQP has performed miracles by upsampling anything to DSD256 and giving a smooth, smooth (did I say smooth?) sound that is rich in texture and incredibly detailed.
[/quote]THAT sounds interesting. However, my DAC already upsamples everything to DSD, so I’m not sure I’d have any chance of hearing a difference.

Though that would be helpful to my old ears too! :wink:

I like “set it and forget it”, but not sure if I have the patience to get there. Thanks for the insight.

1 Like

I use Roon into a Squeezebox Touch.

If I got HQ player would Roon send the HQ player upsampled music to the Squeezebox or does it not work like this?



Hi Rene,

I have got the identical headphone rig to yourself.

What HQPlayer settings do you use for the Nano if you don’t mind me asking?



There’s a very useful Kick-Start Guide by Geoffrey Armstrong which reduces the learning curve associated with the wide variety of HQP options and it’s functional UI (intuitive it’s not).

There is also this helpful précis by Jussi/Miska (the developer) of the theory and a brief description of the function of the various filters.

I think the trick to HQP is to find what input to your DAC sounds best. In my case I am upsampling everything to DSD 128 and sending it to an Auralic Vega. I would try DSD 256 but my Vega doesn’t have the firmware upgrade.

In your case Steve I’d suggest trying upsampling to 192 kHz (which will bypass your DAC’s internal PCM upsampling). The Orbiter having an NAA is a great invitation to experiment.

Nearly all DACs will also convert upsampled PCM into a Sigma Delta Modulated bit stream (SDM to engineers, DSD to us) and feeding upsampled DSD to the DAC can bypass that step also.

What I find (varying with filters) is enhanced detail and micro-dynamics, larger and more focused soundstage and improved transient response (especially decays with mp). The modulators can have a darker/brighter influence. I generally leave HQP on poly-sinc-short-mp/ASDM5 but it can be fun to try out different combinations.

1 Like

I think most people use HQPlayer with Roon for its up-sampling and sigma-delta modulation (“PCM->DSD Conversion”) capabilities.

There’s a 99% chance that your DAC is performing some sort of over-sampling internally–this is pretty much the case unless you went out of your way to buy a NOS (No Over Sampling) DAC.

There is also a very high likelihood that at some point during the source material->analog process, your signal is being digitally transformed from PCM (pulse code modulation at a low sampling rate and high (16+) bit depth) into an SDM (sigma delta modulation, high sampling rate, low (1-5) bit depth) signal inside the DAC chip before being converted into an analog signal.

There are good reasons why it is done this way. The performance/$ is higher, and SDM architecture escapes some pesky analog domain noise-floor issues inherent in building DACs that handle many (18+) bits directly. Virtually all of the popular dac-on-a-chip solutions today (ESS Sabre, Burr Brown, …) do something like this.

There are some issues to consider with this approach:

  • The MIPS budget of the DSP in most DAC chips isn’t very high, so quality compromises are likely during the upsampling/sdm process
  • Putting a hard-working DSP right next to the delicate analog bits within the same chip isn’t the best for analog-domain performance
  • The DSP within your DAC does not change or evolve much after you’ve purchased the product.

HQPlayer addresses all three issues:

  • It lets you use your computer’s CPU to perform the upsampling/sdm, so your MIPS budget is only limited by the amount of hardware you want to pay for.
  • By performing upsampling/sdm outside of the DAC chip to the maximum input rate that the chip supports, you’re taking a load off of the internal DSP, which reduces its influence on the delicate analog bits sitting a few mm away.
  • You can continue to benefit from upgrades to the DSP processes as Jussi makes HQPlayer better, and you upgrade your computer.

If you look at the list of recommended hardware for HQPlayer, you’ll see a pattern: it’s mostly focused on less expensive DACs that don’t have their own discrete DSPs. This isn’t an accident–HQPlayer is ~$100 way to make the digital portion of your $500 DAC behave more like a $5000 DAC.

If you already have an expensive DAC with a discrete DSP implemented on FPGA or similar (dCS, Linn, PS Audio, Meridian, Ayre, MSB, and many others), then they’ve already put some space between the DSP and the analog bits, and the device probably has an ample enough MIPS budget to do an excellent job at up-sampling the signal internally. HQPlayer will make much less of a difference with these.

The biggest drawback of HQPlayer is the overall UX and the technical nature of setting it up. I’d love to see a tighter integration with Roon that gets us past the “player stacked on top of player” stuff and lets us treat HQPlayer as more of a DSP plugin. At the moment, it isn’t built to be used that way, so we went with a more direct integration.


Thanks Andybob. But that’s what my DAC’s doing now. Why would I want a PC to do that for me? I assume the DAC was engineered to do an adequate job of that on its own.

Indeed. Which is really why I asked the question in the first place.

This is a bit OT, but you have just perfectly described the effects of a fairly recent change to my system. I took the inadequate, passive, stock crossovers out of my speakers, and replaced them with new XOs, but ones I built using much (much) higher quality components. I did not change the electrical schematic one iota. So the signal roll offs to the three drivers happens exactly as before. Instead, the change in sound is exactly what you described above. I shudder to think what it would sound like if I used HQ Player to do it even more! :astonished:

1 Like

Brain - Thanks! You ninja posted on me. Am going to have to try to digest your response (assuming I can understand it).

But looks like you may have just answered my question to Andybob. Thanks. Reading it now…

1 Like

I’m one of the few (I’m guessing) that has tried a couple of times and not found any obvious SQ benefit.

It may be that I never found my DACs sweet spot, but I didn’t really have the patience to go more than two sessions with it to find out. The interface and the 30 mins limit on OSX demo didn’t help to be fair.

My Devialet (200) is currently limited to DSD64, or 24/192 and both are up sampled internally from here to its internal native 40/384. Perhaps this is a factor? Who knows.

I found it a real trial to experiment. The ability to change the settings, with play continuing from where it left off - even if it had to pause to sort itself out - would really have helped. One idea would be for HQPlayer to do a ‘let me show you all the settings’ type thing, and just adjust as it went along with all sensible combos. If you read the manual, bizarrely some included filters aren’t really designed to be used, they’re just ‘there’.

If I’d found a hint of something magical I would have been all over it, but it just didn’t happen. I hope I’m not missing out……

1 Like

OK @brian, you DID answer my question to Andybob, in quite a delightful bit of detail. Thank you. I almost understand it. :wink: If so, what I’m taking away from it is…

My DAC uses the Burr-Brown PCM1795 chip: 32-bit/192kHz multi-bit Sigma-Delta variety, which is specifically made to allow the use of external DSP. And I believe that is done in my DAC with a TI SRC439 sampling rate converter. If so, and if I’m understanding, it sounds like I might not be one of those people that see (hear) the benefit over what their DAC already provides.

However, my DAC will NOT improve over time, and HQPlayer will. So it might be worth the investment in time/$ just to “future proof” my DAC to some extent.

Hope I got this right. Thanks.

1 Like

You ninja posted on me. But thanks for the insight. It’s clear that there is not always an audible benefit. And it’s sounding more and more like the higher up the digital food chain your gear is, the less likely you’ll hear a big difference.

There’s a strong following over at ComputerAudiophile, and I’m pretty sure there’s some seriously high-end DACs being used with it - well beyond the cost of mine anyway. It’s so hard to say exactly what people hear though, without being able to listen to everyone else’s systems before and and after. Which would be kind of nice actually, I’ve not really heard too much other hifi….

And I’m not ruling out the fact that maybe there is an improvement but I can’t hear the change for other reasons, like my hearing (not that I think its bad in any way, but you know what I mean), speakers, room setup, etc, being in the way….

I was going to try it for the NAA aspect, which I think would be more beneficial as my source is a fairly standard mac which probably isn’t balanced with the rest of the system. But I just cant be bothered investing the time setting up a new system for Room Correction, and its the one thing I dont want to take out of the chain now Ive finally found it. I stupidly bought a linear PSU for my RPi2 as I really wanted to be ready for RAAT, but now I think the RC thing will keep me tied to my mac for the time being……

In any event, I’d definitely give it a try if you haven’t already. At most you’ll lose an evening or two tinkering.

it sounds like I might not be one of those people that see (hear) the benefit over what their DAC already provides.

I don’t think I would put a PCM1795 + SRC439 configuration in that category. It’s not comparable to running customized DSP like you typically see in the hifi brands that I mention, or in the most popular HQPlayer configurations.

In those cases you’d be using synchronous sample rate conversion built on top of a small family of techniques (most commonly, zero-stuff + apply FIR interpolation filter, or poly-phase resampling). These are almost the same thing, just one is built for integral conversions and one is built for rational conversions.

The important thing is: whether your vendor has built upsampling in a discrete DSP or whether you’re using HQPlayer, the math and techniques are likely to be basically the same, and now you’re down to whether you prefer Jussi’s filter design or not, and whether or not your preferred filter fits into the MIPS budget on the DSP vs your computer.

Asynchronous sample rate conversion is fundamentally different and has a worse quality/$ characteristic.

The most critical part of upsampling happens near the source rate. The trip from 44.1k->88.2k->176.4k. What comes after is a lot less crucial. That’s why products like HQPlayer can confer benefit even if you can’t feed your DAC quite at it’s completely native rate.

All that said, you should listen for yourself and decide. You may hear nothing different. There may be other quality bottlenecks in your system (like the room!) that make this small issue less important. Or it may make a noticeable difference.


I’ve found both of my commonly used DACs (LH Labs Pulse X Infinity and iFi Micro iDSD) love to be fed SDM (DSD128 for the Pulse, DSD256 for the Micro). Since 95% of my music library is PCM :wink: Jussi’s poly-sinc-short, ASDM7 conversion to DSD makes all my music sound just right :smile:

Seriously, the ability to take the upsampling (which, as Brian points out, just about all DACs do) and the conversion from PCM > SDM out of the DAC and into HQPlayer is what “makes” it worthwhile to me.


Thank you Brian and all for the excellent analysis. HQP seems like a worthwhile investment especially given my low end DAC.

I wouldn’t have a DAC in the first place if I hadn’t visited the CA forum a year ago and read about how inferior the PC’s built-in DACS were compared to external DACs. Which led me to question why I would want to take the PCM processing that is now done in the DAC and return it to the PC given the DAC was supposed to be better at the conversion.

Though still flimsy, my grasp is getting better on how the filtering is done by DACs. At least the theoretical versus actual part of it.



Many users find a upsample/filter process etc. setup in HQP that sounds better than what their DAC does by itself. Brian explained it very well, but the 2 basic reasons for this are: a) The filters Jussi writes for HQP are often higher quality than those built into your DAC; b) and part of the reason Jussi can do this is that a PC will have much more computing power than almost any DAC - so by using HQP to do the processing you get better filters that are possible only on a more powerful processing platform than your DAC, and you consequently also “lighten the load” on the DAC.

Both of those above reasons may result in better sound.


Aaaaah (Foley: several pennies dropping). That explains why there isn’t a “cumulative filters” problem.

Well I can’t thank everyone enough for the outstanding, insightful, and information laden replies. Thank you.

After reading this, IMO I’d be a fool not to try it.

Well, I am going to be totally different here than others.
I use HQP as an endpoint on my 2012R2 Server to connect/talk to my DAC (USB), while I have my Roon Core on a separate Win10 Server running RoonServer. Both servers using AO. remote with iPad. This way my Roon Core and HQP are running on separate machines. Network is GB Ethernet/Fiber.

I do no upsampling with HQP as my settings are as follows: None, None, Auto, Auto. I have played around with many HQP settings and while each has its merits there is always something to give up sound quality wise IMO.

Watching units like the SonicOrbiter and the upcoming microRendu (waiting for cost).

@Wayne_Bull1: poly-sinc / ASDM5 mostly. Currently exploring poly-sinc-short. And at times I quite like the freshness of closed-form / DSD5v2 256+fs. All DSD256.

And yourself?