advertisement


Chord Qutest question ...

Keith Howard published an article in HiFiCritic on the M Scaler which does reveal the Chord approach to audio reconstruction. The article is on the Chord site.

From that source, Chord's approach, which is in line with their marketing and with the online tests I have seen, is very conventional although heroically engineered.

According to Howard, the proprietary "WTA" part is the windowing function which is normally applied to the conventional filter coefficients, to deal with the impact of a non-infinite filter length. This is apparently not one of the conventional windowing functions. It is one that RW has said that he tunes according to his ear. So I think we do have a very conventional DAC with RW's proprietary twist.
Thank you. If it is all down to his proprietary twist then I am grateful as it suits me well and lets me get on with enjoying music without wondering why it doesn’t sound quite right.
 
Exactly. I find Chord DACs very natural too but observation of others' negative opinions of them reveals we are all different in what we perceive. It is only a problem when someone gets didactic and insists that what he/she perceives from listening to a DAC or from listening to a fuse is the only valid perception and that other must perceive that too.
Quite right. My recent hearing woes, thankfully partially recovered from, have taught me just how differently we can hear exactly the same sound and that is before we consider the possibility of expectation bias and other psychological factors.

In practice most hearing changes, other than ear wax perhaps, are slow enough for our brains to adapt. An overnight change reveals just how variable our hearing is. Little wonder we can’t be in accord with every opinion expressed!
 
I

I think you need to go back a read what you put. I’m sharing my experiences but they your are basically dismissing it while offering no substance in your experience and you wonder why. You might as well say I don’t believe you. I’m more than happy to talk about things with folk.
There really is no point in continuing this other to note that you joined this forum on April 1st and ask if that is relevant ;), but seriously, best wishes but over and out.
 
That's a very exact recipe.

I take it you've spent a lot of time tweaking your Qutest to the very best it could possibly be!

What did it sound like naked to you?
In stock form, it is very sensitive to the quality of your local mains. I've heard it in 3 different grids across two US States and, with the OEM SMPS, the sound quality in stock form can vary quite greatly in my experience. It could sound slightly thin and etched. But not always. Late nights (with cleaner power) in the PNW always made it sound smoother and more organic. So I am inclined to believe both sides when they say "Qutest sounded like X", because it probably did.

Upgrading to an iFi Elite or Allo Shanti power supply was the first major step in reducing this variability; I settled on the Shanti mainly because it was cheaper and sounded very nearly as good. I then upgraded to an SBooster with the little add-on ultra dongle thing (an improvement, but not a huge one, tbh. seemed to rob some of the life and punch from the music). I then upgraded to a Plixir Elite BDC after not being able to secure a Sean Jacobs or Paul Hynes PS for purchase, new or used (and getting tired of waiting for one to pop-up). The Plixir was an enormous increase in fidelity and is mandatory, IMHO. Just like Jay at AudioBacon, the degree to which it bolsters and reveals the Qutest's true character is hard to digest from a purely logical perspective. However it's a definite "wow" moment for the Qutest. I found Jay's description of the Plixir Elite BDC, with a Qutest, to be spot-on. Though I suspect his Plixir wasn't fully broken-in by the end of his test.

But yes. From a BS Node 2i via USB.... to the Roon Nucleus via USB....to the Denafrips GAIA DDC via USB/BNC.... to Nucleus + Intona USB Isolator..... to adding the SRC-DX (USB -> dual-BNC)....then adding the in-line BNC filters to reduce unnecessary noise....every step made an improvement.

I actually experimented with the BNC filters quite a bit from Thor Labs and MiniCircuits. The tighter you can get from 1Mhz to 100Mhz, as a passband, the cleaner and more organic the sound you get. In my experience. I spent the better part of $1000 USD just in BNC filters to experiment, as an example. There were some theoretical combinations that should have worked, but made getting the signal-lock really unstable, as an example.

In sum, yes...I have experimented extensively.

And, I'll note that the Hugo TT2, while quite a bit better in stock form than the Qutest, and apparently insensitive to PS & upstream USB upgrades/enhancements.....the Qutest had so much more gas in the tank as compared to stock form and responded very well to these upgrades. It's, in my mind, a credit to Rob Watts at incorporating so many genuine improvements into the TT2. Once you start to add up the cost of all these various things....the TT2 isn't a bad value at all.

Though....controversial take: I found my fully hot-rodded Qutest to be slightly better than a TT2 in my system, upon direct comparison. And there was one mod that put that over the top (for another discussion).
 
About the Qutest, I've read somewhere that the 2V and 1V option is digitally attenuated. So in essence the 3V option would be for the purist!

All interesting stuff.
I’m not so sure.. there is a massive thread on headfi on the qutest and I’m pretty sure that digital resolution remains and the output stage is padded down.
 
I’m not so sure.. there is a massive thread on headfi on the qutest and I’m pretty sure that digital resolution remains and the output stage is padded down.

Correct, there's no decimation occurring in the Qutest. I.e. you're not losing bit depth going from 3v to 2v to 1v. RW has explained that he does the attenuation in the FPGA and that bit depth resolution at output is exactly the same for all 3 settings.

I have used the 1v setting for years and, all the times I have A/B'd it with 2v/3v, have not bee able to discern a difference I could attribute to the Qutest.
 
Anyone any idea where 3V as an acceptable line level came from?

I don’t understand why anyone would want this high a level. It is far higher than most power amps could handle (typically 1-2V) so no advantage when using a passive preamp even into amps at the upper end of the range. The now standard 2V for digital was viewed as too much for many preamps when it was introduced and arguably a key reason people didn’t subjectively like CD as it often subtly clipped their preamp stage leading to a perception of harshness and thinness. I don’t see why the figure seems to be creeping ever upwards. It basically means I can’t use the fixed outs on the DSX.
This is speculation. I suspect the 2 V level for CD players and DACs came about because of the capability of digital audio to represent very low level audio signals.

A higher full scale voltage than previously standard was needed to allow those very low-level signals to stay sufficiently far above the thermal noise floor in an analogue preamp to be clearly audible. A 2 V level allows a good margin for 16 bit audio into a normal preamp input.

And then as the resolution of DACs got better some makers succumbed to the temptation to raise that 2 V to get the highest possible dynamic range. Even well above what is really needed for music but looking good on the spec sheet.
 
@John Phillips

I concur, exactly that.

Many of the favoured, low price but apparently-excellent DAC & headphone amps that are praised over on ASR have as much as 4-5vrms outputs, yet the hallowed (there) SINAD measurement does not correct for such, to say a constant 2Vrms per redbook, to allow true comparisons to be made over time. That's up to a 6-8dB 'advantage' sneaked into plain sight - or rather, plain-overlooked in the regime.

So - let us play that game. 'My' headphone amp was tested by stereophile c 1995 at -108dB SNR relative to 1V output.
Hey, that also means today it's right at -121dB SNR at 5v output... take that, suckers (PS: collect remaining bits of your own eardrums on the way out, please..)


Lies, damn lies, and statistics... 'in matters of music, for judgement, ultimately everything comes to the ear.' - said Lord Kelvin, at least 140yrs ago. I cannot argue with a word of that.
 
This is speculation. I suspect the 2 V level for CD players and DACs came about because of the capability of digital audio to represent very low level audio signals.

A higher full scale voltage than previously standard was needed to allow those very low-level signals to stay sufficiently far above the thermal noise floor in an analogue preamp to be clearly audible. A 2 V level allows a good margin for 16 bit audio into a normal preamp input.

And then as the resolution of DACs got better some makers succumbed to the temptation to raise that 2 V to get the highest possible dynamic range. Even well above what is really needed for music but looking good on the spec sheet.

From a technical perspective I’m sure this is correct. The problem is no one even remotely predicted such vast dynamic range potential would usually be wasted pegging everything right at the 0db red line! Really well mastered digital, e.g. ECM, have a very low median level, if anything gets even remotely close to that 0db max it will be maybe one snare or cymbal hit in the course of the whole album. Beyonce, Kendrick Lamar, Flaming Lips etc will seldom drop below the max output voltage.
 
I apologise for the blunt attitude. I thought you were another one of these folk that will argue black is white. I'm sure youve seen them as well. I came out all guns blazing to nip it in the bud and shouldnt have.
Thanks, good of you to post that and pleased to accept your apology.

As it happens I’m fairly centrist on the objective/subjective divide. I’m interested in measurements, how things work and why they sound as they do but also feel that in the end it is whatever brings my favourite music to life as though it is being played in my living room which I choose to use. My signature on the now defunct Wam was something like “it’s what you think you here that really matters”, and that’s a product of what I actually hear and can be measured and verified and how my mind interprets it to recreate the illusion of music. It’s complex, particularly when you take the whole chain into consideration; original performance (assuming it exists) - recording - hifi equipment - listening room - listener. I’m lucky, in that after years of faffing about I have a couple of systems that now work very well for me.

edit: oops, missed out the all important listening room.
 
From a technical perspective I’m sure this is correct. The problem is no one even remotely predicted such vast dynamic range potential would usually be wasted pegging everything right at the 0db red line! Really well mastered digital, e.g. ECM, have a very low median level, if anything gets even remotely close to that 0db max it will be maybe one snare or cymbal hit in the course of the whole album. Beyonce, Kendrick Lamar, Flaming Lips etc will seldom drop below the max output voltage.
Yep, I’m thankful that most of my listening is jazz and classical where dynamic range, which is an essential quality of the music, is preserved (other than being reduced a bit for a smaller listening environment?). The only album I’ve bought that was compressed to whatevery was by Bellowhead; great fun music made unlistenable.
 
That's interesting as I found it forward, bright and synthetic sounding, literally couldn't stand listening to it after I'd got over the initial impressive level of detail and speed. I don't remember seeing any reviews that called it relaxed sounding - I think Chord themselves called it "incisive".

I wonder if there's some strange sample variation going on? I tried it on all output voltages and filters etc.
That’s exactly how I often heard the TT2s (long story short, have owned 3 or 4 over the past few years), and wrote here before about how it could sound ‘processed’. It wasn’t until I used Toslink for both inputs (with no other digital input connected) that I realised I was relaxing into the music and not waiting for the next wince-inducing moment of incisive brightness! Oddly it wasn’t one of those day/night changes, but one you suddenly realise has happened because you’re spinning a greater range of music into the early hours again. So it seems to me that Rob’s dacs are just incredibly sensitive to rf noise. Part of me thinks that for DACs at this price point, that shouldn’t be the case and it should have been designed to cope. The other side of me thinks it’s just a highly resolving DAC, and to get the best out of it without spending silly money on cables one needs to use Toslink. I found the same thing (though the effects seemed magnified) with Hugo 2. Could sound incredibly natural/analogue sounding, could also strip wallpaper!
 
That’s exactly how I often heard the TT2s (long story short, have owned 3 or 4 over the past few years), and wrote here before about how it could sound ‘processed’. It wasn’t until I used Toslink for both inputs (with no other digital input connected) that I realised I was relaxing into the music and not waiting for the next wince-inducing moment of incisive brightness! Oddly it wasn’t one of those day/night changes, but one you suddenly realise has happened because you’re spinning a greater range of music into the early hours again. So it seems to me that Rob’s dacs are just incredibly sensitive to rf noise. Part of me thinks that for DACs at this price point, that shouldn’t be the case and it should have been designed to cope. The other side of me thinks it’s just a highly resolving DAC, and to get the best out of it without spending silly money on cables one needs to use Toslink. I found the same thing (though the effects seemed magnified) with Hugo 2. Could sound incredibly natural/analogue sounding, could also strip wallpaper!
I think it is more a case of Chord DACs not pulling any punches. They are highly resolving, in a good way, but with some partnering equipment and speakers I’ve had they can be a bit too much. With Martin Logan speakers, which I’ve found can be a bit upfront and in your face, the combined effect of DAC and speaker can be a tad too much. On the other hand Chord DACs are, for my ears, a match made in heaven with the more polite Quad stats and German Physiks speakers where the aural image is eight feet behind the speakers.

I’m in two minds about RF noise. Obviously bad if intrusive but the trick seems to be getting rid of that slight RF noise without taking the “energy” out of the sound and more importantly the music. A tricky balancing act.
 
And then as the resolution of DACs got better some makers succumbed to the temptation to raise that 2 V to get the highest possible dynamic range. Even well above what is really needed for music but looking good on the spec sheet.
Hmm I would add two other reasons for having a high output.
The first is that it will “sound better”- something which will sell more units than squeezing out some higher specs- a trick which of course long predates ASR.

The second is that all else being equal and ignoring standards it arguably make more sense in general to have a higher source voltage and less gain in the power amp.
 
Last edited:
@John Phillips

I concur, exactly that.

Many of the favoured, low price but apparently-excellent DAC & headphone amps that are praised over on ASR have as much as 4-5vrms outputs, yet the hallowed (there) SINAD measurement does not correct for such, to say a constant 2Vrms per redbook, to allow true comparisons to be made over time. That's up to a 6-8dB 'advantage' sneaked into plain sight - or rather, plain-overlooked in the regime.
…,
Lies, damn lies, and statistics... 'in matters of music, for judgement, ultimately everything comes to the ear.' - said Lord Kelvin, at least 140yrs ago. I cannot argue with a word of that.
Errm Martin they do try to standardise output.
You know that, right?
 
Part of me thinks that for DACs at this price point, that shouldn’t be the case and it should have been designed to cope.
It absolutely shouldn't be the case. The design failure with the M Scaler is even more preposterous. Watts, with all his training, background, fancy test equipment, experience and simulations down to -300db, couldn’t keep RF from getting from the MScaler to the DAVE and has to resort to cables made by an amateur with no electronics experience on his kitchen table to fix the problem.
 
So it seems to me that Rob’s dacs are just incredibly sensitive to rf noise. Part of me thinks that for DACs at this price point, that shouldn’t be the case
Every part of me thinks that DACs at this price point shouldn’t be sensitive to RF noise in the slightest, especially ones that claim to address issues 300db down, have supposedly been subject to thousands of hours of listening tests, and claim to have undergone the most rigorous measurements. (And I’ve owned 3).
 
…,

Errm Martin they do try to standardise output.
You know that, right?
Thanks for the correction - my head was elsewhere.
Yes, I do see that - but my real point that lay behind - is how easy a trap it is, to directly-compare 'specifications', when the reference is not explicitly stated. Apples vs oranges, all day long. Also - what went Rong in the '%thd wars.' of +30yrs ago / ongoing to this day.


(& to be clear- abs not a barb at ASR: frankly I'm glad to see anyone invest so much into 'measurement' this late in the day, and give the results up freely for discussion)
 
I understand and partly agree with the arguments Chord make for using SMPS, but the amount of leakage current going into the chassis of the TT2 and Mscaler really should be sorted. One hand on the mains chassis grounded Sugden amp and then lightly touching the Chord product with the other hand givers that finger tingling/static feeling which tells you the PSU leakage is rather higher than it should be in products of this cost. They all do it - TT2, Mscaler, Huei, Anni, Qutest.
 


advertisement


Back
Top