Chords New M -Scaler

[austinpop]

I posted this on head-fi over the weekend, but replicating it here for the CA community.

 

I am so glad I had other plans that brought me to London this weekend, allowing me to attend CanJam, reconnect with my good friend @romaz , and meet CA'ers like @str-1, and @TheAttorney.

 

It was also very nice to spend some time with, and get to know @Rob Watts. Since the lines at the HMS/TT2 stations were unrelenting, Rob was kind enough to set me up with his own personal units in a quiet space, so I could have a listen.

 

Please note that these are just first impressions, and subject to change after I have a chance to evaluate in my own home and system.

 

The Hugo TT2 easily drove my HD800 (SD mod) as loud as I could want them, even with the SE output, that Rob says is the most transparent. Luckily I had a female XLR adapter that allowed me to also try the rear XLR balanced outputs, and there is a ton of power in this thing.

 

Since I didn't have a lot of time, I listened to the HMS/TT2 as a combo. I was very impressed. Not only was I hearing excellent SQ, it was without the need for my expensive Cavalli LAu in the picture. This is an intriguing value proposition, as the sale of my amp could mitigate the daunting cost of this combo. What I heard was very clean, punchy, incisive, and balanced. There was no hint of harshness, and a wealth of micro- and macro-detail. There was a rightness to the sound that I very much want to explore further. Instruments like piano and female voices sounded natural and just "right." Bass was prodigious, which is not something I have been able to say with any component other than my Cavalli driving my HD800s.

 

However what blew my mind, and really is the highlight, is the MScaler. I am not putting down the TT2 at all, since it clearly contributes to the whole. Certainly, with more evaluation time, I'll have a better chance to characterize it relative to other DACs I've known and loved. Some of you may recall me saying how impressed I was with the Blu2 demo at RMAF last year. That was in a crowded demo room, listening for all of 5-10 minutes. Here, with a little more time, and full control of the music selection, I have to say I felt that same wow factor from Denver.

 

The HMS has a scaling button that cycles between pass-through, and upsampling to 192, 384, and 768. The upsampling to 192 and then 384 are each a nice jump in SQ. But I was amazed at how big a leap in performance occurred between 384 and 768.

 

I'll reserve more prose for when I can actually review this combo, but I will say if you're in the market and funds permit, this combo merits a careful listen.

[TheAttorney]

12 hours ago, ecwl said:

Since I have a Blu2, and @Miska has said that we can do similar and superior filtering using PCs with HQPlayer, I am excited to try HQPlayer out to:

1) See if I have a personal preference for specific filters, maybe even over Chord's

 

 

I did a HQP vs Chord DAVE upsampling comparison a while back, with HQP on i7 W10 laptop -> mR -> DAVE -> headphones. Always with redbook FLAC files.

 

Previously, I had heard the benefits of HQP's upsampling and filters when driving my iFi iDSD, and I particularly liked HQP's conversion to the higher DSD rates with this DAC, although my laptop struggled to reliably convert to DSD256 and above.

 

However, with DAVE (and its 167k taps), I preferred HQP to have all its upsamplig and filters switched off, i.e. let DAVE do all the upsampling. This was compared to HQP's highest upsampling and best filters. Not a night-and-day difference, but as it meant that I could keep my laptop's CPU usage to a minimum, it was a good result for me.

 

Much later on, I tried the M-scaler part of Blu2 on a weekend home trial, this time with Roon on my laptop, with the same redbook FLAC files. I was one of the few people who where somewhat underwhelmed by the Blu2: yes it sounded better, but not £8k better to my ears, particularly as I didn't need, or like, its CD section. However, I'll be giving the much lower cost Hugo M-scaler another go when it comes out, as the VFM is now much better for those that don't need the CD section. 

 

The interesting thing about my Blu2 experience was that there was no diminishing returns regarding the number of taps: I felt that going from DAVE's 167k taps to M-scaler's 250k or 500k was unremarkable. The biggest SQ boost came when going from 500k to 1M taps (but still not worth £8k for me at the time).

 

So, with some extrapolation of the above simple tests, I would say that any setting on HQP would struggle to compete with the M-scaler.

 

 

[romaz]

On 7/23/2018 at 6:41 AM, mansr said:

I'm certain that nobody can hear the difference between 100k and 1M taps. 

 

I suppose it depends on how you define a "tap."  

 

This is how Rob Watts defines a "tap."  

 

# of taps = (input sample rate frequency in Hz) * (delay time in seconds) * (oversampling rate) * 2

 

For Hugo 2, as an example, sample rate frequency is 48,000 Hz, delay time is .032 seconds, and oversample rate is 16.

 

And so # of taps for Hugo 2 = 48,000 * 0.032 * 16 * 2 = 49,152 taps

 

Feel free to listen to a Hugo 2 with and without an M-scaler.  A partially deaf person should be capable of easily hearing the difference because the difference is not subtle.

 

 

[romaz]

On 7/25/2018 at 3:18 PM, Miska said:

P.S. I just tested setting one of my filters to million taps and upsampling to 16x PCM, CPU load on my oldish quad-core Xeon E5 consumes less than 5% of CPU time.

 

 

Very impressive.  

 

FYI, the new Chord Hugo M-Scaler consumes just 13 watts and after running all day long in its fanless aluminum chassis, it barely feels warm to the touch.

[romaz]

Lots of chest thumping and absolute claims about a product by people who have not yet even heard the product.  Debating theory is fine.  Looking at measurements and plots are important.  As a physician and clinical researcher, this is what I do every day and it is essential in my pursuit to improve the lives of my patients, however, in my pursuit of the enjoyment of music, the most important metrics are not quantitative ones but a single qualitative one.  The system that engages me the most to the music I love wins.  The system that most capably transports me to a different time and place wins.  End of argument.  Go ahead and tell me about how your latest DSD filter is supposed to be the best there is but if I listen to it and compare it and prefer something else, then there's nothing more to say.  My point is that you have to listen and compare and decide for yourself because there's simply no accounting for personal tastes.

 

I own HQPlayer.  I find it to be a powerful, versatile and affordable tool.  It definitely has merit.  I have heard many fine upsampling HQP servers with the SGM2015 perhaps being the best sounding of all.  I get together routinely with a group of audiophiles and we listen and compare different pieces of equipment.  With certain DACs, I'm happy to bring out my PC and upsample to DSD512 with HQP because it sounds better.  With other DACs, upsampling to high-res PCM sounds better.  And with Chord DACs, I find that HQPlayer doesn't add anything at all.  With poor recordings that are harsh and bright, I like what DSD does.  With well recorded music, especially unamplified acoustical recordings (i.e. large orchestral performances), I personally struggle with the softer transients of DSD and its relative lack of depth precision.  Go ahead and tell me I'm wrong.  I've listened and compared enough to know what I prefer.  In the end, it's not about the theoretical superiority of any one approach, it's about what sounds better and this is a very personal thing.  As someone who values transparency and resolution above all, I have yet to hear anything better than a Chord DAVE combined with an M-scaler.

 

[rayl1234]

One other comment don’t want to get lost is that philosophically the HMS direction is actually consistent with the views expressed by the non HMS group here - that of separating DSP from the DAC. Chord does intend to introduce “digital amps’ which will leave the DSP work to HMS and focus on the DAC combined with amp.

 

To that extent, HMS has outputs with additional metadata to pass along volume control.

 

So I see philosophical alignment in viewing DSP as a discrete step— but the distinction is on a piece of hardware or on th PC. 

 

When the digital amps get realized, the relative distribution of value in their stack will shift, making HMS will more compelling as it brings down the cost of the post-HMS chain. But me, I thought it was worth it at almost 2x the cost in the form of Blu2. (I own only 7 CDs, so the transport didn’t factor and I also knew standalone Mscaler was expected within 18 months tops. I had never heard of Chord the company before last October and ended up replacing 1/2 the cost of my system after hearing Blu2DAVE. I did at least catch the lower pre-Jan price list which was about 15% lower, probably thanks to the marked down post Brexit pound at the time.)

[Sloop John B]

2 hours ago, romaz said:

Lots of chest thumping and absolute claims about a product by people who have not yet even heard the product.  Debating theory is fine.  Looking at measurements and plots are important.  As a physician and clinical researcher, this is what I do every day and it is essential in my pursuit to improve the lives of my patients, however, in my pursuit of the enjoyment of music, the most important metrics are not quantitative ones but a single qualitative one.  The system that engages me the most to the music I love wins.  The system that most capably transports me to a different time and place wins.  End of argument.  Go ahead and tell me about how your latest DSD filter is supposed to be the best there is but if I listen to it and compare it and prefer something else, then there's nothing more to say.  My point is that you have to listen and compare and decide for yourself because there's simply no accounting for personal tastes.

 

I own HQPlayer.  I find it to be a powerful, versatile and affordable tool.  It definitely has merit.  I have heard many fine upsampling HQP servers with the SGM2015 perhaps being the best sounding of all.  I get together routinely with a group of audiophiles and we listen and compare different pieces of equipment.  With certain DACs, I'm happy to bring out my PC and upsample to DSD512 with HQP because it sounds better.  With other DACs, upsampling to high-res PCM sounds better.  And with Chord DACs, I find that HQPlayer doesn't add anything at all.  With poor recordings that are harsh and bright, I like what DSD does.  With well recorded music, especially unamplified acoustical recordings (i.e. large orchestral performances), I personally struggle with the softer transients of DSD and its relative lack of depth precision.  Go ahead and tell me I'm wrong.  I've listened and compared enough to know what I prefer.  In the end, it's not about the theoretical superiority of any one approach, it's about what sounds better and this is a very personal thing.  As someone who values transparency and resolution above all, I have yet to hear anything better than a Chord DAVE combined with an M-scaler.

 

 

Indeed it is strange that for some enthusiasts the proof of the pudding is not in the eating but rather in assessing the recipe. 

 

.sjb

[mansr]

2 hours ago, romaz said:

I suppose it depends on how you define a "tap."  

There is only one definition.

[auricgoldfinger]

1 hour ago, Sloop John B said:

 

Indeed it is strange that for some enthusiasts the proof of the pudding is not in the eating but rather in assessing the recipe. 

 

.sjb

 

+1

[mav52]

2 hours ago, Sloop John B said:

 

Indeed it is strange that for some enthusiasts the proof of the pudding is not in the eating but rather in assessing the recipe. 

 

.sjb

 

Very true. But, for some the recipe is as exciting as the actual listening. I for one go the listening route but I do appreciate knowledge base of others in helping to explain how something works.   Like my wifes mother ( who is an award winning baker) tells her , there is more ways to bake a cake, its all in the recipe and the methods used.   I guess we can say the same about dac design and engineering as there are lots of dacs out there using different designs.

 

One thing I do like about this M-Scaler idea,  it can be used with other dacs not just Chords products

[mansr]

8 minutes ago, mav52 said:

Very true. But, for some the recipe is as exciting as the actual listening. I for one go the listening route but I do appreciate knowledge base of others in helping to explain how something works.   Like my wifes mother ( who is an award winning baker) tells her , there is more ways to bake a cake, its all in the recipe and the methods used.

If a cake recipe calls for vinegar, I suspect we can all say something about how it is likely to taste without actually eating it.

[The Computer Audiophile]

8 minutes ago, mav52 said:

One thing I do like about this M-Scaler idea,  it can be used with other dacs not just Chords products

 

Perhaps @austinpop can touch on this. It may not work like many people think it will work when using other DACs.

[mav52]

Just now, The Computer Audiophile said:

 

Perhaps @austinpop can touch on this. It may not work like many people think it will work when using other DACs.

 

I guess it will depend on the other dacs and their inputs and how the other dacs handle decoding.

[mansr]

28 minutes ago, mav52 said:

One thing I do like about this M-Scaler idea,  it can be used with other dacs not just Chords products

What is the output format?

[The Computer Audiophile]

Inputs:

2x BNC

2x optical

1x galvanically isolated USB Type-B compatible with DSD and PCM data (up to DSD 256); with DSD upsampled to PCM

Outputs:

1x optical

1x S/PDIF

1x galvanically isolated dual BNC (enables upscaling to the maximum resolution (768kHz from 96kHz data)

 

[The Computer Audiophile]

Not many DACs have dual BNC cable of receiving this.

[mav52]

5 minutes ago, mansr said:

What is the output format?

 

 

 

I see Chris the dual BNC's

[rayl1234]

3 minutes ago, mansr said:

What is the output format?

Output SPDIF via BNC up to 384, dual BNC up to 768 (one per channel, that's how Chord does it), optical up to 192...

 

I suspect Chris' quote is a marketing copy bug bec I believe it is 768 max from 48 (705.6 max from 44.1).  At least that's what Blu2 supports and I understand the code to be the same.

[guiltyboxswapper]

5 hours ago, romaz said:

Lots of chest thumping and absolute claims about a product by people who have not yet even heard the product.  Debating theory is fine.  Looking at measurements and plots are important.  As a physician and clinical researcher, this is what I do every day and it is essential in my pursuit to improve the lives of my patients, however, in my pursuit of the enjoyment of music, the most important metrics are not quantitative ones but a single qualitative one.  The system that engages me the most to the music I love wins.  The system that most capably transports me to a different time and place wins.  End of argument.  Go ahead and tell me about how your latest DSD filter is supposed to be the best there is but if I listen to it and compare it and prefer something else, then there's nothing more to say.  My point is that you have to listen and compare and decide for yourself because there's simply no accounting for personal tastes.

 

I own HQPlayer.  I find it to be a powerful, versatile and affordable tool.  It definitely has merit.  I have heard many fine upsampling HQP servers with the SGM2015 perhaps being the best sounding of all.  I get together routinely with a group of audiophiles and we listen and compare different pieces of equipment.  With certain DACs, I'm happy to bring out my PC and upsample to DSD512 with HQP because it sounds better.  With other DACs, upsampling to high-res PCM sounds better.  And with Chord DACs, I find that HQPlayer doesn't add anything at all.  With poor recordings that are harsh and bright, I like what DSD does.  With well recorded music, especially unamplified acoustical recordings (i.e. large orchestral performances), I personally struggle with the softer transients of DSD and its relative lack of depth precision.  Go ahead and tell me I'm wrong.  I've listened and compared enough to know what I prefer.  In the end, it's not about the theoretical superiority of any one approach, it's about what sounds better and this is a very personal thing.  As someone who values transparency and resolution above all, I have yet to hear anything better than a Chord DAVE combined with an M-scaler.

 

 

My experience with DSD upsampling is quite different to yours.  Softer transients on DSD?  Lack of depth precision?  Guess it'll depend on the DSD implementation on the DAC, but that's certainly the opposite I've experienced.

 

I actually had the DAVE around for the 2nd time, as I was looking forward to dropping HQPlayer (can be a bit unreliable in Windows) and going back to basics (i.e. input - DAC does the work - output). 

 

Treble energy on the DAVE was almost too much in my setup (KEF Blades, NC1200s, room treated) to the point of pain and that's despite the NC1200s running at 11db gain.  Soundstage width was noticeably narrower, as was sound stage depth.   Had lost a bit of impact/dynamics, along with arguably a lower register in bass response.    

 

I was quite looking forward to removing HQPlayer from the chain and simply keeping the DAVE, but low and behold HQPlayer + T+A chipless DSD at 512 rates did it again.  I have to admit if it wasn't for the searing treble I may have overlooked some of these minor differences and gone with the DAVE anyway.  But I certainly don't agree with the idea of a good DSD implementation being transiently weak or lacking soundstage precision - infact I find in native DSD mode its much more effortless.

 

I didn't bother trying HQPlayer with the DAVE when its already doing something similar onboard, plus didn't want the hassle.

[guiltyboxswapper]

One thing certainly worth mentioning - its important to pay attention to the quality of the USB feed to the DAC in DSD mode.   Using a dedicated USB card really was an eye opener, when combined with HQPlayer.  Jitter & USB noise at DSD 512 rates can certainly take the edge off the performance.

[ecwl]

23 minutes ago, guiltyboxswapper said:

My experience with DSD upsampling is quite different to yours.  Softer transients on DSD?  Lack of depth precision? 

 

Treble energy on the DAVE was almost too much in my setup (KEF Blades, NC1200s, room treated) to the point of pain and that's despite the NC1200s running at 11db gain. 

I think system synergy sometimes matters more. With class D amplifiers, the switching frequencies are usually in the 400-1000kHz range. If DAVE's transient timing is more accurate because of the upsampling to 768kHz 24-bit, that transient timing accuracy is going to be slightly lost due to switching frequencies of the class D amplifiers.

[mansr]

26 minutes ago, rayl1234 said:

Output SPDIF via BNC up to 384, dual BNC up to 768 (one per channel, that's how Chord does it), optical up to 192...

Dual S/PDIF is kind of rare, but at least it's a standard.

[TheAttorney]

1 hour ago, mav52 said:

One thing I do like about this M-Scaler idea,  it can be used with other dacs not just Chords products

 

Not really. The DAC needs to have dual BNC inputs, working in parallel, to get the full 1M taps effect.

Without that, the best you can get is 0.5M taps, which as has been said earlier, is nowhere near as good.

 

So in practice, only the most recent Chord DACs will get the full benefit of the M-scaler.

[austinpop]

1 hour ago, The Computer Audiophile said:

 

Perhaps @austinpop can touch on this. It may not work like many people think it will work when using other DACs.

 

1 hour ago, The Computer Audiophile said:

Not many DACs have dual BNC cable of receiving this.

 

1 hour ago, rayl1234 said:

Output SPDIF via BNC up to 384, dual BNC up to 768 (one per channel, that's how Chord does it), optical up to 192...

 

I suspect Chris' quote is a marketing copy bug bec I believe it is 768 max from 48 (705.6 max from 44.1).  At least that's what Blu2 supports and I understand the code to be the same.

 

First, each of the BNC S/PDIF outputs is capable of up to 384kHz sample rate, which Chord DACs like the DAVE, Hugo 2, and now TT 2, support. Here is the spec, from the DAVE manual:

•  75Ω SP/DIF (sic) BNC Coax 44.1KHz - 384KHz

With regard to the dual-BNC signaling, I was curious about this myself, so asked Rob if this was a proprietary format. His response:

Quote

It uses the mono data format on AES3, in order to double the sample rate, so this is an industry standard. But I don't know any DACs that use it today.

 

I also asked him why he didn't have the output available on USB. His response:

Quote

I have not tested the USB TX device yet!

 

So bottom line, while you could get HMS upsampling to 192kHz to most DACs, that only gives you access to 1/4 million taps. Should a DAC be capable of 384kHz on its S/PDIF input, that would give you 1/2 million taps.

 

In my (admittedly limited) listening, I found a profound jump in SQ from 1/2 million to 1 million taps, so for the moment, I only see the true value of HMS as a companion to the H2, TT2, and DAVE DACs.

 

Perhaps in the future, Rob will consider delivering the full 1 million tap output at 768kHz on USB, which should be widely compatible. However, I can imagine that being a business decision for Chord, as HMS could drive a lot of Chord DAC sales.

[Confused]

It occurs to me that there are some similarities between the M-Scaler and the dCS Vivaldi Upsampler.  The dCS uses dual output, albeit AES/EBU, but can be set to output via single AES/EBU, with lower output rates to suit whatever DAC is used. 

 

In terms of the speculation about using the M-Scaler with non Chord DACs, I have heard the dCS Upsampler with a Devialet D800, running 24/192 PCM and also DSD64.  The results were very impressive, but the cost of the dCS Upsampler is so high that it is almost impossible to justify running in a compromised mode that is required to suit something like a Devialet.  The dCS Upsampler is supposed to be at it's best when converting / upsampling to DSD, so this contrasts to the M-Scaler which I think is basically for PCM.  So comparisons between the two are likely to be very different depending on the DAC used.

 

So over time I guess we will see a few people experiment with the M-Scaler and various different DACs.  The dCS Upsampler is about £16K in the UK (Or £26K with the Vivaldi Master Clock), so in comparison the M-Scaler looks cheap!  (Although still very expensive versus HQPlayer, of course.)  It will be interesting to see if anyone gets good results with the M-Scaler and a non Chord DAC.