[Scarydog]

I have played around with HQPlayer a few times and finally bit the bullet and bought a copy 6 months ago. I have repurposed an older Mac Mini now running linux and HQplayer Desktop.

It works fine, I can upsample PCM ok, not really enough grunt to upsample to DSD but that’s fine.

But I have a coupe of issues with HQPlayer.
The interface is terrible, really poor feedback and some really quirky bugs. I saw that new version has been released with a new interface. I downloaded it and tried it, it was better so I thought I would buy it. Unfortunatley the price to upgrade is nearly as high as the original purchase price, which I already thought was not great value.

The only alternative I can think of is Roon. It doesnt sound as good, but is certainly much easier to use (and much less expensive)

I have had a bit of a search for hardware solutions but can’t find quite what I am after.

I’d be grateful for any suggestions for software or hardware upsamplers.

 

[m-i-c-k-e-y]

How about Roon + HQplayer? Roon has an option to use it as an interface and use HQplayer as your audio engine.

What I have right now is a hardware solution: Audiobyte Hydra.Hub a DDC, upsampler/format converter/reclocker. Takes away the brunt on my PC (so now i can use even any low power streamers). But it works best when used together as a stack w/ its accompanying DAC and PS. They have announced however a new product that is more independent. The SuperHub.

There is another harware solution. The Mutec MC3+ USB. Have many connections but somewhat limited in conversion.

 

[Scarydog]

Thanks for that I will have a look at the Audiobyte stuff that sounds like what I am after!

 

[GoldenSound]

There are plenty of other upsampling tools, including as mentioned above the Hydra Hub for example, but in terms of stuff that will compete with the level of quality HQPlayer's most advanced filters can offer, you kinda have two options:

The Chord MScaler, or PGGB.

MScaler is a great tool though limited to 192khz output with non-chord DACs, and is quite pricey.
PGGB is currently the highest performance PCM upsampling tool available period, both in terms of the filter itself and its noise shapers (accurate to beyond -700dB!). But because of how processing intensive it is, it's not intended to run realtime like HQPlayer. It upsamples to a file which you then play back as normal in roon or your player of choice

 

[spacecase25]

I would tend to agree that HQPlayer + Roon is your best option, especially since you've already bought the license for HQPlayer and you've bought a year (assuming you're not just on the trial) of Roon. The two are a great combination. HQPlayer also works well with LMS, assuming you have the embedded version of HQP, as the desktop version doesn't support UPNP. LMS + HQPlayer is what I use when I want to use HQPlayer since I ditched Roon.

I also don't like the interface for HQPlayer. Going the some other player + HQP route means you'll only have to deal with the HQP interface to change filters / settings and you'll do all your browsing in the media player (either Roon or LMS).

I pretty much always listen with NOS tho, I prefer it to oversampling.

 

[gregorio]

I’d be grateful for any suggestions for software or hardware upsamplers.

Why don’t you just let your DAC up/oversample the PCM? After nearly 40 years development, they’re very good at it, even the cheap ones.

G

 

[GoldenSound]

Why don’t you just let your DAC up/oversample the PCM? After nearly 40 years development, they’re very good at it, even the cheap ones.

G

Most DACs don't have filters that actually adhere to Nyquist theorem.
And even if you do have one that adheres to Nyquist and doesn't attenuate audible band content, there are still benefits to be had by using a higher performance reconstruction to preserve more timing information

 

[gregorio]

Most DACs don't have filters that actually adhere to Nyquist theorem.

Technically no filter (in DACs or in software) adheres to Nyquist/Shannon Theorem. However, for 30 years or so even cheap DACs have adhered to it beyond the threshold of audibility.

And even if you do have one that adheres to Nyquist and doesn't attenuate audible band content,

Again, you’re describing what a cheap 30 year old DAC could do.

there are still benefits to be had by using a higher performance reconstruction to preserve more timing information

What timing information isn’t preserved by oversampling in a DAC?

G

 

[GoldenSound]

Technically no filter (in DACs or in software) adheres to Nyquist/Shannon Theorem.

It sort of depends on exactly how you define it. A perfect Nyquist reconstruction would require instant and infinite attenuation at the Nyquist frequency which is impossible.
But generally people seem to describe them as adhering to Nyquist when they achieve full attenuation before the Nyquist frequency, thereby eliminating any unintended signal content. But they will of course have some amount of rolloff under that frequency.

Traditionally, filters have been considered 'transparent' whilst adhering to Nyquist so long as they both attenuate fully by Nyquist, and have almost no attenuation under 20khz. Though hardly any DACs achieve this as this still requires some computing power. Most DACs either just do not attenuate by Nyquist meaning the signal straight up is not being reconstructed properly, or they rolloff under 20khz which is a directly audible effect.

However, for 30 years or so even cheap DACs have adhered to it beyond the threshold of audibility.

This isn't the case at all. As mentioned, most DACs have filters which either roll off the audible band (directly audible), or do not rolloff by nyquist and cause imaging and inaccurate timing of transients.

Frequency domain content is not the only issue as our hearing does not adhere to the fourier uncertainty principle, meaning whilst we can only hear frequency domain content upto about 20khz, several studies have found that we can discern time domain differences of around 10uS, which would require a bandwidth of about 80khz to describe. Some studies have even showed audibility of time domain differences down to about 6uS which would require an almost 150khz bandwidth to describe.

This alone logically concludes that therefore accurate preservation of additional high frequency/timing information is audible, but there are also studies on the matter of high resolution recordings vs normal 44.1khz recordings with 'normal' reconstruction filters and a meta-analysis of studies conducted on the topic can be read here and concludes:

"Results showed a small but statistically significant ability of test subjects to discriminate high resolution content, and this effect increased dramatically when test subjects received extensive training. This result was verified by a sensitivity analysis exploring different choices for the chosen studies and different analysis approaches. Potential biases in studies, effect of test methodology, experimental design, and choice of stimuli were also investigated. The overall conclusion is that the perceived fidelity of an audio recording and playback chain can be affected by operating beyond conventional resolution."

What timing information isn’t preserved by oversampling in a DAC?

The time domain and frequency domain are inherently linked. To describe something with more precision in the time domain you need a wider frequency domain bandwidth.
Attenuation of higher frequencies means you lose precision in that regard, so earlier attenuation than Nyquist effectively 'throws away' time domain precision, and insufficient/leaky filters mean imaged/aliased content that should not exist remains there, producing not only time domain (and frequency domain) inaccuracies, but also signal correlated inaccuracies which can additionally be argued to be worse than random inaccuracy.

 

[gregorio]

Traditionally, filters have been considered 'transparent' whilst adhering to Nyquist so long as they both attenuate fully by Nyquist, and have almost no attenuation under 20khz.

They don’t have to “attenuate fully” to be considered “transparent”, only to the threshold of human hearing, so -90dB to -100dB is fine, even -80dB is perfectly transparent except in particularly exceptional circumstances. But I agree they do have to have little or no attenuation under 20kHz though.

Though hardly any DACs achieve this as this still requires some computing power.

Again, even cheap 30 year old DACs were capable of this and they did have “some [enough] computing power”.

… meaning whilst we can only hear frequency domain content upto about 20khz, several studies have found that we can discern time domain differences of around 10uS, which would require a bandwidth of about 80khz to describe.

Ah, you seem to be confused. You’re confusing audio wave period with time differences. For example, you can take a 900Hz sine wave then take another 900Hz sine wave and place it 10uS later. There you go, a time domain difference of 10uS requiring a bandwidth of just 1Hz not 80kHz, we cannot hear 80kHz regardless of it’s time domain difference.

Some studies have even showed audibility of time domain differences down to about 6uS which would require an almost 150khz bandwidth to describe.

Again, it doesn’t require a bandwidth of more than 1Hz. You seem to be confusing a time domain difference of say 6uS with the different frequency a sound wave would have if it’s “wave period” were 6uS. So this argument is a red herring as even 44/16 can resolve time domain differences around 50,000 times smaller than 6uS!

This alone logically concludes that therefore accurate preservation of additional high frequency/timing information is audible …

It doesn’t logically conclude that, only your confusion between a time domain difference and a wave period has concluded that.

a meta-analysis of studies conducted on the topic can be read here and concludes

Ah the famous “paid for” Reiss study that cherry picked a bunch of studies to show a minuscule statistical difference that unsurprisingly supported his wishes and the wishes of the company that paid for it.

To describe something with more precision in the time domain you need a wider frequency domain bandwidth.

Same confusion again, where did you get this from?

G

 

[GoldenSound]

Again, even cheap 30 year old DACs were capable of this and they did have “some [enough] computing power”.

I'd be very interested if you can share any examples.

Ah, you seem to be confused. You’re confusing audio wave period with time differences. For example, you can take a 900Hz sine wave then take another 900Hz sine wave and place it 10uS later. There you go, a time domain difference of 10uS requiring a bandwidth of just 1Hz not 80kHz, we cannot hear 80kHz regardless of it’s time domain difference

Again, it doesn’t require a bandwidth of more than 1Hz. You seem to be confusing a time domain difference of say 6uS with the different frequency a sound wave would have if it’s “wave period” were 6uS. So this argument is a red herring as even 44/16 can resolve time domain differences around 50,000 times smaller than 6uS!

I'm not sure how you're coming to that conclusion

Additionally, even if we were to ignore the discussion about making a conclusion from the study of audibility of time differences, we can just look at the study showing our ability to discern high sample rate content from redbook content. That quite definitively says 'this makes a difference at least for some people'

 

[gregorio]

I'd be very interested if you can share any examples.

I measured quite a few DACs in the 1990’s most of them fit the bill but I don’t still have those measurements. However, Here’s the measurements of a 7 year old DAC that cost $9. Note that at 19,250Hz the roll-off is -0.15dB and by 21,700Hz it’s -95dB.

I'm not sure how you're coming to that conclusion …

Easy, why don’t you try it yourself, I presume you have some sound editor software or a DAW? Create a say 1kHz sine wave, duplicate it to the other channel, move it by say 6uS and record the stereo output down. If your software doesn’t allow movement by a fraction of a sample then move it one sample using a sample rate of say 176.4k or higher and down sample it to 44.1k or even 32k if you want, no difference.

Additionally, even if we were to ignore the discussion about making a conclusion from the study of audibility of time differences …

We don’t need to ignore the audibility of time differences. Those time difference audibility amounts are well supported by several/various studies, independently verified and generally accepted as between 5-10uS.

That quite definitively says 'this makes a difference at least for some people'

Yes it does, based either on no evidence or cherry picked studies. 52% positive result rate is very easy to demonstrate if you just exclude one or two studies but even then, 52% is hardly “quite definite” compared to 50% being completely random guessing.

G

 

[GoldenSound]

Easy, why don’t you try it yourself, I presume you have some sound editor software or a DAW? Create a say 1kHz sine wave, duplicate it to the other channel, move it by say 6uS and record the stereo output down. If your software doesn’t allow movement by a fraction of a sample then move it one sample using a sample rate of say 176.4k or higher and down sample it to 44.1k or even 32k if you want, no difference.

This is referring specifically to entirely periodic content, this is not the same for transient content

However, Here’s the measurements of a 7 year old DAC that cost $9. Note that at 19,250Hz the roll-off is -0.15dB and by 21,700Hz it’s -95dB

Can't see the graph there accurately enough to verify. I've ordered one to test though as I appear to have lost mine.
Additionally it's still attenuating by a notable amount before 20khz iirc, and additionally was not phase linear.

 

[GoldenSound]

Yes it does, based either on no evidence or cherry picked studies

Do you have examples of contrary studies you feel should have been included?

 

[Scarydog]

Why don’t you just let your DAC up/oversample the PCM? After nearly 40 years development, they’re very good at it, even the cheap ones.

G

My dac is a holo spring 3 and doesn’t offer over sampling. I really like the sound of NOS but also like the upsampled sound from HQPlayer.

I just find HQPlayer itself clumsy and awkward. I also don’t really like having a computer in my audio chain as it nice to get away from them sometimes . (Even running a headless Linux box I find myself having to futz with it semi regularly)

Interesting discussion, I need to do some reading as I couldn’t completely follow it