Harley's at it again!!! - FOQUS (bastard son of MQA)

[fatoldgit]

"I’ve had the opportunity to listen to FOQUS and can say unequivocally that it represents a significant improvement in sound quality that will benefit all music lovers, no matter what their playback hardware. —Robert Harley"

see https://www.theabsolutesound.com/ar...f-the-year-foqus-analog-to-digital-converter/

see also https://mqalabs.com/foqus/


In 2015, in his "review" of MQA, Harley stated "MQA accomplishes the seemingly impossible—delivering extraordinary sound quality in a form that easily integrates into the existing music distribution infrastructure, is backward compatible with existing hardware, and serves all applications, from smartphone streaming to high-end home downloads, with a single file. The combination of streaming—which is poised to become the dominant paradigm"

The song remains the same.... Does Bob [Stuart] have some Harley "pee" tapes?

Peter

 

[Nutul]

Oh noe...

 

[fpitas]

Lol!

 

[IngMishael]

he has always been shameless

 

[Curvature]

Harley wrote possibly the worst book on audio in existence so this is right on par.

 

[beeface]

The song remains the same.... Does Bob [Stuart] have some Harley "pee" tapes?

This is funny, but the true explanation is surely as simple a cash payment, not to sound conspiratorial. TAS in 2026 is a rinky-dink outlet. Too small for anyone to care or do anything about them not declaring an advertorial, not enough revenue to be principled about what they write.

 

[RandomEar]

Hmmm... took a look at the whitepaper. This is the supposed impulse response for their ADC:

That's just... a Gauss filter or maybe an even more trivial moving average? That will start to droop beginning at 0 Hz and will be down multiple dB by 20 kHz, right? This is garbage. What am I missing here?

 

[pjug]

Hmmm... took a look at the whitepaper. This is the supposed impulse response for their ADC:

View attachment 557243
That's just... a Gauss filter or maybe an even more trivial moving average? That will start to droop beginning at 0 Hz and will be down multiple dB by 20 kHz, right? This is garbage. What am I missing here?

I don't know what they do at 192KHz but at 1X rate I think they use a flattening filter along with their slow filter. The usual Rube Goldberg M.O. by the MQA people.

From: https://www.audiosciencereview.com/...ter_for_high-quality_audio__doc-3-pdf.456132/

 

[Curvature]

Hmmm... took a look at the whitepaper. This is the supposed impulse response for their ADC:

View attachment 557243
That's just... a Gauss filter or maybe an even more trivial moving average? That will start to droop beginning at 0 Hz and will be down multiple dB by 20 kHz, right? This is garbage. What am I missing here?

I think they are prioritizing a minimum phase response, which will have no pre-ringing. The diagram seems to suggest that there us no post-ringing either, which as far as I know is impossible. The engineering is sure to be solid, but it will likely lack practical merit, and the marketing will be deceptive, as it was with MQA.

There is a gigantic increase in new-generation rich folks from successful businesses. Maybe this is why the industry is still pursuing these schemes. There has to be a business case for investors, and a target demographic, otherwise it would be a waste of time. @kemmler3D Thoughts?

 

[fpitas]

There is a gigantic increase in new-generation rich folks from successful businesses.

That may be the demographic. But a lot of older audiophiles seem to have money burning a hole in their pocket.

 

[gzost]

With everything regarding Robert Harley it helps to remember that he thought the adjustment of the rise times/pulse width that Wadax offered for digital transfer between two of their ridiculous overpriced boxes allowed fine-tuning the sound of the DAC!

 

[NTK]

I think they are prioritizing a minimum phase response, which will have no pre-ringing. The diagram seems to suggest that there us no post-ringing either, which as far as I know is impossible. The engineering is sure to be solid, but it will likely lack practical merit, and the marketing will be deceptive, as it was with MQA.

There is a gigantic increase in new-generation rich folks from successful businesses. Maybe this is why the industry is still pursuing these schemes. There has to be a business case for investors, and a target demographic, otherwise it would be a waste of time. @kemmler3D Thoughts?

The Gaussian filter as shown is linear phase. You can tell that from the symmetry of the impulse response. The typical response of a Gaussian filter to a square pulse is as follows:

 

[Curvature]

The Gaussian filter as shown is linear phase. You can tell that from the symmetry of the impulse response. The typical response of a Gaussian filter to a square pulse is as follows:

View attachment 557246

I believe they are showing the measured rather than ideal impulse response.

 

[respice finem]

That may be the demographic. But a lot of older audiophiles seem to have money burning a hole in their pocket.

Maybe that's why the name of this "newcomer" sounds like "fuck us"?

 

[RandomEar]

I think they are prioritizing a minimum phase response, which will have no pre-ringing. The diagram seems to suggest that there us no post-ringing either, which as far as I know is impossible. The engineering is sure to be solid, but it will likely lack practical merit, and the marketing will be deceptive, as it was with MQA.

Oh, I get that they are trying to do this because their market is allergic to "ringing" IRs. My main problem is that the FR would be very, very non-flat in the audible band and not suppress much of the ultrasonic region. So unless I'm mistaken, this thing's not doing either of the things we would want a filter to do. Their 48 kHz filter also looks like a very slow roll-off minimum phase design, so that checks out.

Depending on the exact filter design, both, 48 and 192 kHz, should either sound "dull" due to a reduction in treble amplitude and still have ultrasonics bleed into the recording, or they could be almost flat in the region up to 20 kHz and essentially filter nothing far above that (40++ kHz). That does not look like solid engineering, unless I'm missing something.

 

[Curvature]

Oh, I get that they are trying to do this because their market is allergic to "ringing" IRs. My main problem is that the FR would be very, very non-flat in the audible band and not suppress much of the ultrasonic region. So unless I'm mistaken, this thing's not doing either of the things we would want a filter to do. Their 48 kHz filter also looks like a very slow roll-off minimum phase design, so that checks out.

Depending on the exact filter design, both, 48 and 192 kHz, should either sound "dull" due to a reduction in treble amplitude and still have ultrasonics bleed into the recording, or they could be almost flat in the region up to 20 kHz and essentially filter nothing far above that (40++ kHz). That does not look like solid engineering, unless I'm missing something.

I believe you are describing NOS filter characteristic impacts to FR. I am curious if that's all this is.

 

[kemmler3D]

I think they are prioritizing a minimum phase response, which will have no pre-ringing. The diagram seems to suggest that there us no post-ringing either, which as far as I know is impossible. The engineering is sure to be solid, but it will likely lack practical merit, and the marketing will be deceptive, as it was with MQA.

There is a gigantic increase in new-generation rich folks from successful businesses. Maybe this is why the industry is still pursuing these schemes. There has to be a business case for investors, and a target demographic, otherwise it would be a waste of time. @kemmler3D Thoughts?

To me it looks like a gamble from Lenbrook, I didn't spend very long looking but it doesn't seem like any products actually incorporate this yet. You can see the prominent "become a partner" button on the Foqus page but no mention of actual partners anywhere. My take is they are throwing good money after bad because some exec there thinks the margins on licensing snake oil are too good to pass up. It's not immediately clear how they've improved on the original failed MQA business model, though.

 

[Soundproof]

MQA Stand:

KE + DN High-performance compact demagnetizer - Maurer Magnetic

maurermagnetic.com

 

[RandomEar]

I believe you are describing NOS filter characteristic impacts to FR. I am curious if that's all this is.

No. The minimum phase filter they show for 48 kHz is broadly similar to common minimum phase slow roll-off filters available in many DACs. And they start rolling off at about 14 kHz with -0.5 dB @ 16 kHz and -4 dB @ 20 kHz being common numbers. The FOQUS filter just looks even "slower" than that, so I would like to see the actual FR instead of the impulse response.

 

[NTK]

I believe they are showing the measured rather than ideal impulse response.

View attachment 557247

The figure they showed is, IMO, laughable. Pre-ring, post-ring, and undershoot are seldom used as metrics for filters, unless for specialized situations where your system has a true need to control them, such as your vehicle suspension. (Yes, your vehicle suspension is a low pass filter. Over/under-shoot and ringing are controlled through damping.) Filters "pre-ring", "post-ring", and "undershoot" based on, and are part of their, the filter designs. The important filter design specs for electrical signals are pass-band flatness, stop-band attenuation, width of transition from pass band to stop band, etc.

The figure showed the ADC response, not DAC. So not very applicable to music consumers. Here are my estimated response based on the impulse response shape.
Sampling frequency was 192 kHz, which gave a Nyquist frequency of 96 kHz. Filter (anti-aliasing) response was about -12 dB at Nyquist, with early roll off and excessively wide transition. Broken by design