Chord launches new M Scaler for Hugo TT

TOBYJUG posted:

What if one of those taps goes faulty and starts dripping noise. Will be a buggeration for the tech docs to find it.

Lot of washers to replace if they leak...

But so what, if it sounds good, reraely goes wrong, and is fixed or replaced if it does go wrong. I’ve yet to hear the MScaler (or blu), but I couldn’t care less about yhe technology inside, only the sound, and Dave alone is sublime, so if the M-Scaler really improves it to a noticeable degree (and I am wary of the hyperbole endemic in this forum), then it must bring te nearest thing to heaven on earth if the rest of the system is up to it, which I thought Dave was already!

Well, I've watched the video of Mr Watts trying to explain himself, as I have several other similar videos, and as is pretty much always the case, I'm totally confused.  In a talk he did about DAC design, he speaks of the WTA reconstruction filters and taps, and I kind of understood what he was on about.  Here though, there is a reconstruction filter pre-DAC.  Err....so what does the DAC do when of these up-scaling things is attached?  Fine, there's a million taps in the Blu2 and Hugo thing, but the DACs fed by it have their own filters, and with a fraction of that.  How does that all fit together?  Aargh,  Brain hurty!

Peder posted:

Simon about Rob Watts...

🔹◾even if he couldn’t at the time totally explain why.◾ 

And it is this that is so fascinated with this interest,everything can not be explained.!

But so it is in "real life too",everything you can not explain,if we could we would not have any problems😉.

Audioquest says themselves that they do not know why their ehternet cables sound different,...just that they do it.

That is,why it is so important to test and compare,...regardless of technical explanation.

/Peder 🙂

To a point, our conversation was more exploring  it was potentially addressing some unknown flaw or sub optimisation elsewhere... and we discussed what had been measured and assessed..but yes he was happy this product was subjectively improving the output of some of his DACs and he was pleased with it.... but it might not have been specifically related to the oversampling... hence my comment about hidef audio... all is not what it appears... to which I am sure he won’t mind me saying we both shared a smile.

BTW upsampling is not that good, as it introduces mathematical errors..however over sampling is different as it usies whole multiples of Fs and using zero value sample insertion is generally perceived as the best way as it produces no specific digital distortion/mathematical errors.

 

TOBYJUG posted:

No one mentions the TToby.   

Looks a very neat system M Scaler, TT2 and a pair of Toby's. Rob has mentioned some new digital amps.

one thing i wonder is why have Chord designed the case in TT style, possibly a Dave version of M Scaler with 2M taps in the future?

SongStream posted:

Well, I've watched the video of Mr Watts trying to explain himself, as I have several other similar videos, and as is pretty much always the case, I'm totally confused.  In a talk he did about DAC design, he speaks of the WTA reconstruction filters and taps, and I kind of understood what he was on about.  Here though, there is a reconstruction filter pre-DAC.  Err....so what does the DAC do when of these up-scaling things is attached?  Fine, there's a million taps in the Blu2 and Hugo thing, but the DACs fed by it have their own filters, and with a fraction of that.  How does that all fit together?  Aargh,  Brain hurty!

A reconstruction filter is an intrinsic part of the DAC and the digital to analogue signal process. Effectively the reconstruction filter processes the digital values so as make the discrete samples appear as close as possible to infinitely narrow points in time with infinite energy so they can more accurately represent a continuous signal... which of course is not physically possible hence the approximation of the reconstruction filter. Typically the sample values are passed through a digital Sinc filter (ie 1/sin(x) ) to achieve this. The larger the kernel of this filter (ie the more samples or ‘tap’ values it contains) the better.

Simon-in-Suffolk posted:
SongStream posted:

Well, I've watched the video of Mr Watts trying to explain himself, as I have several other similar videos, and as is pretty much always the case, I'm totally confused.  In a talk he did about DAC design, he speaks of the WTA reconstruction filters and taps, and I kind of understood what he was on about.  Here though, there is a reconstruction filter pre-DAC.  Err....so what does the DAC do when of these up-scaling things is attached?  Fine, there's a million taps in the Blu2 and Hugo thing, but the DACs fed by it have their own filters, and with a fraction of that.  How does that all fit together?  Aargh,  Brain hurty!

A reconstruction filter is an intrinsic part of the DAC and the digital to analogue signal process. Effectively the reconstruction filter processes the digital values so as make the discrete samples appear as close as possible to infinitely narrow points in time with infinite energy so they can more accurately represent a continuous signal... which of course is not physically possible hence the approximation of the reconstruction filter. Typically the sample values are passed through a digital Sinc filter (ie 1/sin(x) ) to achieve this. The larger the kernel of this filter (ie the more samples or ‘tap’ values it contains) the better.

Thanks.  In theory then this upscaled digital feed should give the DAC an easier task?  I'd formed the impression that Chord DACs were pretty well optimised for redbook, so feeding them 768khz (not sure of the bit-depth) conflicts with that, doesn't it?  Apologies for being such a dunce.

Nyquist  theorem relates to encoding a continuous signal as a series of discrete samples, and it states to accurately capture the continuos signal by sampling it, the frequency rate of the series of samples must be at  least twice that of the highest frequency in the continuous signal to be captured.

Oversampling is for playback, not encoding, and it allows simplified or more accurate low pass filters to be incorporated into the DAC in terms of performance and reduced artefacts across the sampled frequencies (pass band),  by creating ‘headroom’ above the sampled frequencies where there should be no meaningful data. It also allows the reduction of timing noise (Jitter) in the range of sampled frequencies in a ADC to DAC chain.

SongStream posted:

Thanks.  In theory then this upscaled digital feed should give the DAC an easier task?  I'd formed the impression that Chord DACs were pretty well optimised for redbook, so feeding them 768khz (not sure of the bit-depth) conflicts with that, doesn't it?  Apologies for being such a dunce.

Easier might be one way to call it, but effectively allows the response of the filter to be more accurate and consistent across the audio band .. 

To your observation, well that moves into areas specific to Chord DACs... and it would appear in those designs a separate oversampler may subjectively provide a better result... and this may be because of sensitives due to system crosstalk and transport stream quality.... but this moves into speculation on things that are proprietary and specific to Chord designs.

tonym posted:
Gazza posted:

Tony, presumably same sound quality without a CD player should be lower price, has the same number of the “watts” taps for conversion, so could be happyxmas.

Yes, I hope so Gazza, and that was my understanding, although reading through Chord's blurb it states the device is optimised for the TT, and doesn't mention DAVE.

Checkout the discussion on headfi. Rob makes it pretty clear that an M Scaler + DAVE is better than an M Scaler + TT2, though the latter is arguably better than a DAVE on its own. 

French Rooster posted:

if the nd555 , around 21k source, can be outperformed by a 4k source, naim will be considered as a crap audio company or chord will the miracle in audio history.

I don’t believe in any of this possibilities .

M Scaler is not a source. It still needs some kind of streamer before it and a DAC after it.

I suppose by the time you have taken care of those, the proposition gets fairly costly.

feeling_zen posted:
French Rooster posted:

if the nd555 , around 21k source, can be outperformed by a 4k source, naim will be considered as a crap audio company or chord will the miracle in audio history.

I don’t believe in any of this possibilities .

M Scaler is not a source. It still needs some kind of streamer before it and a DAC after it.

I suppose by the time you have taken care of those, the proposition gets fairly costly.

the source is the 4k combo of m scaler/ hugo tt. I invite you to read above what Hifi Nut wrote and what Emre asked.  My response was related to that.

French Rooster posted:
feeling_zen posted:
French Rooster posted:

if the nd555 , around 21k source, can be outperformed by a 4k source, naim will be considered as a crap audio company or chord will the miracle in audio history.

I don’t believe in any of this possibilities .

M Scaler is not a source. It still needs some kind of streamer before it and a DAC after it.

I suppose by the time you have taken care of those, the proposition gets fairly costly.

the source is the 4k combo of m scaler/ hugo tt. I invite you to read above what Hifi Nut wrote and what Emre asked.  My response was related to that.

I did read it. The M Scaler is 3.5k on it's own. With a TT the total is 7.5k. Or with a DAVE, 11.5k. And it still needs a transport of some description.

Frank Yang posted:

Should Naim hire this guy Rob Watts?

I think you will find that Naim do not like the sound of what Rob Watts designs, they would at least heavily modify any FPGA design he would come up with. So perhaps not a marriage made in heaven. I think Naim would/have said the Chord approach gives a different presentation, but not how they feel music should sound. The punter needs to decide between them, not try and make Naim sound the same as Chord or Linn etc.

Sloop John B posted:

Rob explains how it works here. 

https://youtu.be/VfscfTkHgM4

He he seems to me to suggest that the mathematical formula used is proven to perfectly recreate the original audio signal. 

.sjb


Very interesting video !

Rob Watts explains the timing problem in digital music and its solution in a very straightforward way.

Makes me wonder how or if Naim did solved this essential problem in it's new ND555.


Gazza posted:
Frank Yang posted:

Should Naim hire this guy Rob Watts?

I think you will find that Naim do not like the sound of what Rob Watts designs, they would at least heavily modify any FPGA design he would come up with. So perhaps not a marriage made in heaven. I think Naim would/have said the Chord approach gives a different presentation, but not how they feel music should sound. The punter needs to decide between them, not try and make Naim sound the same as Chord or Linn etc.

That's a strange one Gazza. To my old, Naim-battered lugholes, I'd say Chord DACs have all the strengths of the sound, and then some. I'm afraid my cynical nature might think "Well, they would say that, wouldn't they?", but then I discount pronouncements from Naim or Chord until I've heard things myself.

From my purely personal point of view, the main, really interesting,  comparison will be Melco/Scaler/DAVE vs. ND555/Melco, and as I've only heard these two setups at my dealers I wouldn't like to pick a winner until they've both performed in my own system for a while.

To me the people buy Chord over Naim because they prefer some aspect. Roy George is still at Naim 3 days a week listening and voicing the product, so no change likely. If Naim ever did go down the FPGA route I doubt it would sound like Dave. I have no axe to grind here as my next step or choice is NDX2 vs ND555. Whichever I prefer I will demo out against Blu2/Dave. If I like the latter I will hold on until the mscaler arrives. They are all great products, just depends on ones own lugholes for final approval to reach for wallet. I also need a preamp, probably 252, so big outlay.

We listened to a technical review from Steve Sells and even despite the old dac technology they do believe this is their best product. Having heard it twice now the jury s still out for me, the NDX2 and 555ps i suspect may have a punchier sound, need to hear them both in more ideal conditions.

TT2 + Mscaler + ND5xs2 you are still under 10K vs 20K of ND555.......

Or Dave + Mscaler + ND5xs2 your are at 14K vs 20K and got a world class headphone amp as well...

If you have 555ps idle is a different story..... 

İf you dont have TT2/Mscaler is a serious competition against both Nd555 and NDX2...

Many people are using Hugo with their expensive naim streamers.... 

 

Folks, I had a home demo of BLU2 + Dave, the music was absolutely awesome. I hadn't played any CD for a long time but it was definitely very impressive. Interestingly, the CD sounded better than via the Melco USB out. I had to use the NDS digital out to match the CD sound quality. Now waiting for an M-Scaler demo while I'm experimenting with the Dave direct to the 300DR.

×
×
×
×