Article: My Quest for a New DAC, Part 5: Chord Electronics Hugo M Scaler & Hugo TT 2

[mozes]

No need to spend $$$ to upgrade the performance of the TT2 by investing in a PSU a la PH SR7. I found a clearly noticeable improvement by replacing the stock PSU of the TT2 with a 50,000 mah laptop powerbank at 15V. I don't know why, but maybe less noise from the SMPS. All of my gear are plugged to one dedicated AC circuit.

P.S. I use the TT2 to directly drive my Omega speakers.

[ted_b]

Rajiv,

As others have already said, this review (and the review cycle in general) is a model for what audio reviews should include.  It is also gave me, personally, a real benchmark for any additional reviews I would do in the future (when my system is back up).  As some may know, I have reviewed (very positively) Chord products in the past, have a good relationship with Rob Watts, yet have not listened to their offerings in a couple years.  This review got my Chord-ism rekindled.   

 

Regarding the M Scaler vs HQP, I am VERY interested in someone doing some relatively equivalent comparisons.  Whereas I trust Rob's heightened approach to "temporal coherence" and his focus on noise floor modulation, I also respect Jussi's software-based approaches to filters, modulators and letting the DAC do as little heavy lifting as possible, given the price/performance of massive hardware/software vs the $5 upsampling chips in most dacs.  The relevance of a comparison between these two SOTA approaches (M Scaler and HQP) seems to be a no-brainer.  Anyone?

 

Net/net, great job Rajiv.  You've raised the bar.

[Archimago]

14 hours ago, audio.bill said:

This has been discussed by Rob Watts (Chord's digital designer) over at the head-fi forum where he claimed that his hardware solution cannot be equaled in software like HQPlayer. As you'd expect HQPlayer's designer (Miska here) claimed otherwise, that their solution implemented on a suitably powered PC can actually exceed the performance provided by Chord's HW solution. That's a somewhat simplified view of their relative positions from my recollection. Guess it's up to each listener to decide which is best in a given system for themselves as the ultimate test. 

 

Would be nice to see Rob Watts try proving that claim.

 

Doubt his rationale given that computing hardware gets cheaper and faster over time and the software can certainly target and exceed the computational precision.

[austinpop]

 

50 minutes ago, ted_b said:

Rajiv,

As others have already said, this review (and the review cycle in general) is a model for what audio reviews should include.  It is also gave me, personally, a real benchmark for any additional reviews I would do in the future (when my system is back up).  As some may know, I have reviewed (very positively) Chord products in the past, have a good relationship with Rob Watts, yet have not listened to their offerings in a couple years.  This review got my Chord-ism rekindled.   

 

Thanks!

 

50 minutes ago, ted_b said:

Regarding the M Scaler vs HQP, I am VERY interested in someone doing some relatively equivalent comparisons.  Whereas I trust Rob's heightened approach to "temporal coherence" and his focus on noise floor modulation, I also respect Jussi's software-based approaches to filters, modulators and letting the DAC do as little heavy lifting as possible, given the price/performance of massive hardware/software vs the $5 upsampling chips in most dacs.  The relevance of a comparison between these two SOTA approaches (M Scaler and HQP) seems to be a no-brainer.  Anyone?

 

 

While it seems simple, try defining an experiment where these are the only independent variables!

 

Both approaches are tied to DACs that are conducive to their scaling. I've tried HQPlayer with DACs that didn't benefit at all, although I readily believe others who say it helps with certain DACs. Similarly M Scaler really works best with Chord DACs with the 2-stage WTA architecture.

 

 

I'd say the only way to compare is to assemble an entire digital chain, from music server to DAC, and compare the entire chains.

 

 

[austinpop]

2 hours ago, chrille said:

Hello austinpop, 

and thanks for your rapid response.

I think I 'll stick with the ones I have .

Vänskä's Mahler recordings  from Minnesota have received   quite mixed responses from some reviewers and I am not really in desperate need of another M2.

 

From a performance perspective I prefer Fischer's  M2 from Budapest on Channel Classics and it is also  a  good native  DSD 64 recording.  Jared Sacks balances very naturally and realistically compared to DGG's normal  style of a bit  too much of  multimiking. Kaplan's  VPO M2 though available as SACD in DSD is actually a native  24/96 PCM recording. Quite good by DGG standards ,but at least via my SACD players not as good as Channel Classics M2 on SACD. Or for that matter  when played as DSD 64 download via Q/HMS.

Via my home system I can play the channel M2 louder without congestion than the DGG/ Kaplan which has a tendency to congest and harden a bit  at climaxes  via my electrostatic speakers.

But the Channel Classics M2  sounds really impressive at  very loud levels too.

But I may change my mind on the Kaplan one  if I get to hear it in its native 24/96 version via M Scaler.

I'll start by playing the rbcd layer of my SACD via M Scaler  when I get home again to hear how that fares. And if it sounds good enough I may go for the 24/96 download.

Good as 16/44.1 can sound via M Scaler it is still not hi res imho.

M Scaler has changed my opinion of the SQ levels  on quite a few recordings. Mainly for the better. But not always. Revealing as it is,it can also make  the bad aspects of a recording  more noticable. 

 

PS a bit OT maybe but how did you like the Maggie 3.7s?

 

IMHO Maggies can sometimes  be a very  good alternative to large electrostatics for large scale symphonic music.

Enjoy your TT2 /M Scaler Cheers Chrille

 

I just haven't heard Fischer's 2nd, although if his monumental 3rd (which I bought in DXD) is any indicator, it will surely be great. I'll give it a listen.

 

The Maggie 3.7s are not mine, but my friend's. He has taken great care to set them up, with proper positioning and room treatments. They are stunning for the price! Especially when matched with good electronics like the Ref 6 and Hegel power amp.

[austinpop]

2 hours ago, str-1 said:

Excellent review, Rajiv.

 

Thanks,

 

2 hours ago, str-1 said:

As an owner of both M Scaler and TT2 I was particularly interested in your comments on use of the SR7.  I was not surprised by your findings with the TT2 given the role of the super-capacitors but was interested to learn how important running the SR7 at 15V was to getting a significant bump in performance with the M Scaler.  You didn't mention so I guess you did not try the SR4 used elsewhere in your chain, but from your experience would you hazard a guess that it (the 5V-12V version) probably wouldn't deliver a significant improvement over the supplied smps?  I wonder if the 9V-19V version might be a better bet, but probably too soon for anyone to be able to try that.

 

Yes, after my finding with 12V robbing dynamics, I didn't' even bother with the SR-4 set at 12V. I would love to know what the 19V version of the SR-4 (set to 15V) would sound like.

 

2 hours ago, str-1 said:

Also, I take from your equipment list that you used the Cardas Clear with the M Scaler's smps.  If that was the case, did you notice an appreciable difference compared to the supplied power cable?  I'm wondering if this would be a better and cheaper way for many to further raise the M Scaler's performance.

 

No, actually I used the supplied power cable. The SMPS for the HMS and TT2 has a 2-pin C7 input, and all my power cables are the standard IEC. Good question - hard to say.

[austinpop]

2 hours ago, TheAttorney said:

An excellent review @austinpop. So clearly presented, with all the comparison information one could wish for.

 

I've had my own HMS for a few weeks now (added to DAVE) and my own thoughts broadly agree with yours. In particular:

 

1. The quality of source and all upstream components still matters.

 

2. My single rail SR7 (set to 15v)  gave an incremental improvement, in paricular giving a more natural presentation over stock power supply, which in comparison sounded a bit "etched".

 

3. Tweaking the stock BNC cables gave incremental improvements: With only one cable connected at a time (i.e. not using the full 1M taps), I tried multiple clip-on ferrites (17) and JSSG 360. They were both better than stock, but I couldn't decide which was best, but it didn't matter because of the next change:

 

4. After some research and PMs, I tried a pair of Blaxius^2D BNCs as being a relatively low cost upgrade (around 250 Euro each + VAT in EU countries). I'm nowhere near the allegedly 150 hour burnin required for these, but even from the first day I was taken aback by the improvement they gave. My subjective initial reaction was that they doubled the improvement given by the stock HMS. One of those "Whoa, I didn't see that coming!" moments. I'll report back after some more burn-in when I'm out of the honeymoon period.

 

So I'm curious how the Blaxius would compare to the other cables you tested. Any chance you could get a set to review?

I've been following the Habst reviews with interest, but they are just soooo expensive.

 

One word of warning is that the Blaxius is thick and very stiff, with a large radius of curvature, so choosing the right length is important.

 

 

As I wrote, my intent was just to see how much impact, if any, different BNC cables made between HMS and TT2. Being an owner of both Lush and Lush^2 cable, I know how good Peter's cables can sound. I'd certainly like to try the Blaxius^2's at some point.

[austinpop]

2 hours ago, barrows said:

Hey Rajiv,

 

Thanks for the very informative and in depth review.  I have not heard the TT2 myself, so my main reference for Chord DACs is DAVE, which (like most) I like very, very much.  It is good to know that the TT2 gets close to that level of performance, and also to know how good it is at driving (even demanding) headphones directly.  I am not surprised that you found some improvement with a good linear power supply, I designed a linear supply for my Brother's Qutest, and he notes increased smoothness and better dynamic shadings with it versus the OE switcher.

Also, thanks for reminding me about the Gorecki #3, I had forgotten about this wonderful piece of music and ordered up a DSD (SACD) version last night.

 

Best,

 

Barrows

 

I have this DSD version: http://lipinskisound.com/products/sacd-lrf-105-gorecki-/ which was conducted by Górecki himself. Unfortunately this version appears to be out of print, but if you can find it, I recommend it.

[austinpop]

The other thing I tried in this review was to publish a playlist on Qobuz of the tracks mentioned in the review, along with others I listened to in the course of the evaluation. Please let me know if you were able to access the playlist, and if helped with the review.

[TheAttorney]

1 hour ago, ted_b said:

Regarding the M Scaler vs HQP, I am VERY interested in someone doing some relatively equivalent comparisons...  Anyone?

 

I have used (and still have) HQP on laptop to drive my portable iFi iDSD DAC/amp. Leaving the weird BugHeadEmperor aside for a moment, it was the best sounding s/w app I knew of at the time, and the upsampling and filters etc made worthwhile subtle improvements in that environment. A bit extra detail/dynamics/smoothness etc - but the original sound stayed fundamentally the same.

 

It's a huge apples/oranges shift to try to compare that to HMS/DAVE, but nevertheless I feel confident that the subtle upsampling effects that HQP gave would never compare to the transformational impact the HMS has on DAVE.

 

I say "transformational" because of, for example, the way HMS adds significant image height across the whole width of the soundstage. The HQP upsampling didn't even begin to be able to pull off a trick like that. I should add that I find HMS transformational in some ways and rather subtle in others. When I listened to a HMS/TT2 in noisy meet conditions, I noticed the difference the HMS made, but it seemed rather underwhelming under such conditions. I agree with those who say it takes some time to tune in to the full range of benefits.

 

BTW, I made a typo in my last post that I can't now correct: I described my SR7 as single rail. What I meant to say was that it is single regulated (with 3 rails).

 

 

[mozes]

2 hours ago, austinpop said:

I'd certainly like to try the Blaxius^2's at some point.

You should! I use these cables and they really showed what HMS/TT2 is capable of. Highly recommended. 

[str-1]

2 hours ago, austinpop said:

No, actually I used the supplied power cable. The SMPS for the HMS and TT2 has a 2-pin C7 input, and all my power cables are the standard IEC. Good question - hard to say.

Yes, I imagine most of us have few if any good C8 cables to try. 

 

The purists will scoff at use of such cheap adapters but I found this £6 investment to be worthwhile - https://www.amazon.co.uk/gp/product/B07BRTN7YZ/ref=ppx_yo_dt_b_asin_title_o06_s00?ie=UTF8&psc=1

I’ve yet to listen critically with this adapter but I already know from a quick tryout while changing things around in comparing different bnc cables that there are differences to be heard with the M Scaler’s smps.

[ted_b]

Just now, TheAttorney said:

 

I have used (and still have) HQP on laptop to drive my portable iFi iDSD DAC/amp.

 

Thanks.  Although I respect your opinion, I have to say that I don't think HQP was being used to anywhere its fullest here.  Why?  Twofold (at least):

1) the iFi dac does not get out of the way, entirely.  That is to say, it is not an NOS dac, nor has an NOS setting (to my knowledge) so the full extent of HQP's upsampling to one's sweetspot is difficult, if not impossible, to achieve. 

2) HQP on a laptop does not represent anywhere near a fair fight against HMS!    My HQP i6700k server environment is tweaked to the max, with some very heavy lifting occurring.  And we have not begun to discuss any possible advantages of an NAA in this comparison. 

 

My comments about a comparison between HQP and HMS rely heavily on the reviewer doing a complete, i.e all-in, comparison..taking full advantage of each "upsampler" architecture.  As I like to say in a baseball analogy, the goal is to get the batter out; not comparing just speed:  Nolan Ryan's fastball to Greg Maddox's off speed stuff.  Ryan wins every time.  This likely means two different signal paths, with HQP on a robust server talking to an NOS-style dac, while HMS runs dual BNC to a Chord dac of similar pedigree (read: total price?).  This is less an experiment (more than one variable) than it is a comparison of two upsampling approaches.  The TT2/HMS solution in one column, the HQP/server/NAA (if needed)/NOS dac in the other column. 

[Jud]

6 minutes ago, ted_b said:

That is to say, it is not an NOS dac, nor has an NOS setting (to my knowledge) so the full extent of HQP's upsampling to one's sweetspot is difficult, if not impossible, to achieve. 

 

True, although if you send it DSD512, what I think you are mostly hearing is the relatively inexpensive parts in the final analog filter (edit: and output circuitry), more than any DSD filtering.

[Always.Learning]

4 hours ago, austinpop said:

 

I just haven't heard Fischer's 2nd, although if his monumental 3rd (which I bought in DXD) is any indicator, it will surely be great. I'll give it a listen.

 

FWIW, the Fischer/Budapest Mahler 2 also happens to be my favorite. You can check it out on Tidal in 16/44.1, and the goodness really shines through.

[sdolezalek]

1 hour ago, ted_b said:

Thanks.  Although I respect your opinion, I have to say that I don't think HQP was being used to anywhere its fullest here.  Why?  Twofold (at least):

1) the iFi dac does not get out of the way, entirely.  That is to say, it is not an NOS dac, nor has an NOS setting (to my knowledge) so the full extent of HQP's upsampling to one's sweetspot is difficult, if not impossible, to achieve. 

2) HQP on a laptop does not represent anywhere near a fair fight against HMS!    My HQP i6700k server environment is tweaked to the max, with some very heavy lifting occurring.  And we have not begun to discuss any possible advantages of an NAA in this comparison. 

 

Ted: Although it may not be possible, my original question would essentially say "What would it take, in combination of computer/HQP/NAA outputting the same signal to a Hugo TT2 to equal what an HMS does.  In other words, leave the DAC static.  

 

A separate test would be to say what computer/HQP/NAA/NOS DAC combo might equal the HMS/TT2 combination in sound, but that introduces a bunch of DAC variables on top of the upsampling.  

 

Lastly, I have found that my taste in filters might not be the same as that of others.  Software gives me the option of choosing the one I like best; hardware tends to lock me into the filters the designer likes best.

[zettelsm]

Wonderful review, Rajiv. In fact, a model of what a review should be and sadly too often isn't. You could give lessons to more than a few of the print reviewers.

 

I'm in the process of moving on from my DAVE and Blu MK II, but that in no way invalidates the fact that you got the essence of the Chord sound spot-on. For those that have been eagerly anticipating the TT2/HMS release your review should be required reading.

 

Best regards,

 

Steve Z

[Superdad]

7 hours ago, austinpop said:

Yes, after my finding with 12V robbing dynamics, I didn't' even bother with the SR-4 set at 12V. I would love to know what the 19V version of the SR-4 (set to 15V) would sound like.

 

Great review Rajiv.  And I bet it drives a lot of people to try out the much more affordable and well-admired Chord Qutest.

 

Just FYI:

Last month we did a customized JS-2 for a client with one of its outputs offering a 15V setting (he traded away the 7V setting of one rail).  On the bench I had it cranking out 4.8 amps continuous at 15V (versus our usual 7.5A/12V).  All it took was changing one voltage set resistor.  [But people should not ask us for 19V.  Tried that, and with the current R-core trans barely made it to 1A.]

We could send you a JS-2 with 15V available from both rails and it would easily handle both the TT2 and the HMS.

[Superdad]

1 hour ago, zettelsm said:

In fact, a model of what a review should be and sadly too often isn't. You could give lessons to more than a few of the print reviewers.

 

No shit! Big time. 

(And this was really two product reviews, so I hope Chris paid you double...)

[austinpop]

17 minutes ago, Superdad said:

(And this was really two product reviews, so I hope Chris paid you double...)

 

Now why didn't I think of that! 😛

[bobfa]

Rajiv,

 

This is one of the really nice touches in your review.  This covers a very broad range of great material to listen to.  I am playing with it on my systems this weekend.

 

Bob

 

 

Review Playlist
 
HMS/TT2 Review Playlist on Qobuz (US)
 
To enable you to listen to the same tracks (see caveats) I did, I created a public playlist on Qobuz USA. This playlist includes the tracks mentioned in this review, as well as others I listened to in the course of this evaluation. Please note that in some cases, the Qobuz track will only stream at 16/44.1, whereas I may have used a local hi-res version. Still, this gives you a sense for the music I listened to for evaluation.

[luisma]

12 hours ago, austinpop said:

 

 

Thanks!

 

 

 

While it seems simple, try defining an experiment where these are the only independent variables!

 

Both approaches are tied to DACs that are conducive to their scaling. I've tried HQPlayer with DACs that didn't benefit at all, although I readily believe others who say it helps with certain DACs. Similarly M Scaler really works best with Chord DACs with the 2-stage WTA architecture.

 

 

I'd say the only way to compare is to assemble an entire digital chain, from music server to DAC, and compare the entire chains.

 

 

Thank you Rajiv for bringing closure, I have been following. This comparison indeed would be awesome, just that there would be more variables to consider on the software upscaling side that on the closed fixed mscaler ecosystem, some parametrization would be required 

[luisma]

12 hours ago, austinpop said:

The other thing I tried in this review was to publish a playlist on Qobuz of the tracks mentioned in the review, along with others I listened to in the course of the evaluation. Please let me know if you were able to access the playlist, and if helped with the review.

I just finished listening to your entire playlist and it is great, this is something very helpful to have on these reviews, like it all with the exception of the Gorecki track, not one to my liking. Call me simplistic but the quality of daft punks Within I was far surprised by it. 

[luisma]

9 hours ago, ted_b said:

Thanks.  Although I respect your opinion, I have to say that I don't think HQP was being used to anywhere its fullest here.  Why?  Twofold (at least):

1) the iFi dac does not get out of the way, entirely.  That is to say, it is not an NOS dac, nor has an NOS setting (to my knowledge) so the full extent of HQP's upsampling to one's sweetspot is difficult, if not impossible, to achieve. 

2) HQP on a laptop does not represent anywhere near a fair fight against HMS!    My HQP i6700k server environment is tweaked to the max, with some very heavy lifting occurring.  And we have not begun to discuss any possible advantages of an NAA in this comparison. 

 

My comments about a comparison between HQP and HMS rely heavily on the reviewer doing a complete, i.e all-in, comparison..taking full advantage of each "upsampler" architecture.  As I like to say in a baseball analogy, the goal is to get the batter out; not comparing just speed:  Nolan Ryan's fastball to Greg Maddox's off speed stuff.  Ryan wins every time.  This likely means two different signal paths, with HQP on a robust server talking to an NOS-style dac, while HMS runs dual BNC to a Chord dac of similar pedigree (read: total price?).  This is less an experiment (more than one variable) than it is a comparison of two upsampling approaches.  The TT2/HMS solution in one column, the HQP/server/NAA (if needed)/NOS dac in the other column. 

This is exactly my point, speaking lightly and without proof of course I would say a comparison between these 2 would have to be defined in advance on the HQP side, I strongly believe that implementation, hardware, software, tuning, os etc can lead to HQP to sound even better and also worse. All depends on the testing environment definition. 

[ajay556]

I would recommend to compare the new Hegel mohican CD player $5k.. Add a $1200 power cord and $500 isolation feet. Burn all discs on ultra disc gold CDs and watch the music unwind like a sweet vinyl sound with great dynamics. I am reveling a secret sauce here but i know it will not be tasted since its not a computer sourced system. I have tried the schitt yagadrasil, PS audio stream, Chord Qutest and finally the Innuos Zenith with new MSB discrete DAC. None of these are in the league of the Hegel mohican. Though i am sure the TT2/HMS must be excellent too. I am a fan of chord electronics  -  I do own the hugo DAC. I believe Chord have very distinctive sound - transparency and speed are they forte. It can be very seductive initially, but wear with time. Fluidity and liquidity (sorry best way to describe live music with two words :-)) is what are brains truly associated with live music.