Chord Electronics Owners & Discussion Thread

[Ittaku]

39 minutes ago, rocky500 said:

I see now, they were smart to use dual BNC as it means it locks you into one of the great Chord Dacs instead of using another brand to get the major benefits of the m scaler.

 

Not really, as very very few DACs support sample frequencies above 382 via USB anyway, but sure their target market will always be enhancing their own DACs.

[rocky500]

53 minutes ago, Ittaku said:

Not really, as very very few DACs support sample frequencies above 382 via USB anyway, but sure their target market will always be enhancing their own DACs.

The Holo Dac seems to be able on USB only to accept 768kHz.

Edited January 17, 2019 by rocky500

[Ittaku]

6 minutes ago, rocky500 said:

The Holo Dac seems to be able on USB to do 768kHz.

Nice! Mine supports 3072 but I don't know of a single device that outputs that.

[Guest]

@Ittaku that’s just bragging

[Ittaku]

1 minute ago, Sime V2 said:

@Ittaku that’s just bragging

You better believe it! I need to stroke it after spending so much money damnit!

[davewantsmoore]

3 hours ago, Cardiiiii said:

So basically if I listen to my Spring DAC is NOS mode (44.1), I will get no benefit from this device?

No, there'll be no major penalty.   The DAC will receive audio from the scale which has been oversampled to ~384khz ....and then your DAC would do no more oversampling.

Edited January 17, 2019 by davewantsmoore

[agisthos]

5 hours ago, legend said:

 

I do not know much about the theory of the M-scaler (having only just heard about it from Simon) but wonder if talk about the million taps relates to FFT and so whether the aural improvements are due to improvements in timing - in much the same way as MQA claims?

This is exactly what is going on with the M-Scaler. Despite not knowing much about the technology you have picked up on it. 

According the Rob Watts, the ear is much more susceptible to timing transients than generally realized. With their WTA filter, the more taps they implement the better the timing, and this results in improved depth perception. But how WTA and MQA do it is very different apparently.

[davewantsmoore]

17 hours ago, legend said:

I do not know much about the theory of the M-scaler (having only just heard about it from Simon) but wonder if talk about the million taps relates to FFT and so whether the aural improvements are due to improvements in timing - in much the same way as MQA claims?

Sort of.    In so far as it is about get the final waveform out of the converter more accurately.

 

The Chord approach is a little more "brute force", and less "unique" (not that I want to call MQA literally unique) than MQA.

 

As you would know, the sampling theorem requires the use of infinites.... infinitely long impulses, based on the infinite history of the waveform ..... and an obvious way to better approximate this is to use longer digital filters.

[davewantsmoore]

As I mentioned before, this is a common approach that people have been doing for a long time with computers.

 

From Chords own tech paper:

 

Quote

But there is another way, which is a lot cheaper than acquiring a collection of Chord hardware and much easier than taking the very major step of programming an FPGA. That’s to perform sinc interpolation offline, in software. I first wrote a software utility to do this over 10 years ago – and compared to FPGA programming it’s an absolute doddle. The problem is, the program takes ages to run with anything longer than a very short audio file because it requires the calculation of (U-1) × N2 sin(x)/x values, where U is the oversampling factor and N is the number of samples in the file. This is for each channel. But, like I say, it’s easy, it’s cheap, and it allows you to generate and listen to a file that’s been oversampled using full sinc interpolation, in which respect it’s even better than the M Scaler.

These days you don't have to do it offline, and it doesn't take ages.    It can be done with a similar level of delay (less than a few seconds) to the M Scaler.

[legend]

2 minutes ago, davewantsmoore said:

As I mentioned before, this is a common approach that people have been doing for a long time with computers.

 

From Chords own tech paper:

 

These days you don't have to do it offline, and it doesn't take ages.    It can be done with a similar level of delay (less than a few seconds) to the M Scaler.

Many thanks for the theoretical info.  Any suggestions/recommendations for particular programs/software to do it?

[rand129678]

44 minutes ago, davewantsmoore said:

From Chords own tech paper:

 

Is that from Chord's own tech paper?

 

Looks to be an article by Keith Howard for HiFiCritic magazine... interesting article still.

 

https://www.moon-audio.com/mwdownloads/download/link/id/159/

Edited January 18, 2019 by Music2496

[rand129678]

42 minutes ago, legend said:

Many thanks for the theoretical info.  Any suggestions/recommendations for particular programs/software to do it?

Jussi Laako's HQPlayer is one, as mentioned below:

 

https://www.stereo.net.au/forums/topic/260051-chord-hugo-m-scaler-discussion-and-impressions-thread/?do=findComment&comment=4049319

 

Website:

 

https://www.signalyst.com/consumer.html

Edited January 18, 2019 by Music2496

[Cruncher]

I use Audirvana + and it can do up-sampling. I can set 'filter max length' up to 2 Million using a slider in the GUI. It makes a difference, but nothing mind blowing. Up-sampling did use more a lot more CPU ( I have a very old MacBook. It has a built in CD player). I got the SOX option free in an upgrade I can't say I would have paid for it. Hence I am struggling to my experience with up-scaling to the mscaler reports.

 

I did have a delta-sigma DAC and I felt it did a better job with DSD than PCM. Using Audirvana I used to upscale and convert everything and send it to the DAC as DSD.

 

Linux has some free upscale

[rand129678]

15 minutes ago, Cruncher said:

I did have a delta-sigma DAC

 

Just in case anyone was interested, Rob Watt's Chord DACs are delta sigma also...

 

"My pulse array (5 bits, 2048FS) does not show this behavior on simulation - a -60dB step change has no consistent delay compared to 0dB step change. This is due to the very high speed of operation, the 5 bit resolution, and the fact that one can properly dither a 5 bit delta sigma"

 

and

1. It's delta sigma but with 300 dB noise shapers running at 104MHz."

 

He explains some big differences with other DACs below:

 

https://www.head-fi.org/threads/hugo-tt-2-by-chord-electronics-the-official-thread.879425/page-142#post-14557160

 

https://www.head-fi.org/threads/watts-up.800264/page-34#post-13934433

 

Edited January 18, 2019 by Music2496

[Ittaku]

On linux I use pulseaudio with sox-vhq filter built in, but I also hacked it myself to use the 32bit resampling (instead of 28) that isn't normally even used in the soxr library. Doesn't make an audible difference but it makes me feel better knowing everything is set to 11.

Edited January 18, 2019 by Ittaku

[Cruncher]

@Music2496 I should have said a DAC based on the ESS Sabre32 chip.

[Steveo888]

@legenddid you happen to listen to the mscaler with your deqx int he chain? Did you happen to listen to your Tikandis with it?

[legend]

Unfortunately we ran out of time to try the M-Scaler on the DEQX and Tikandi.  We were using a modified version of the (passive) Kantus with a Be tweeter so had the same drivers as the Tikandi Grandes but without the subs.  I am not sure that using the M-Scaler's upsampling with the DEQX processor would have been useful given the lower audio sampling rates of the DEQX (limited by all the Xover and other DSP work they have to do), 

[davewantsmoore]

2 hours ago, legend said:

Many thanks for the theoretical info.  Any suggestions/recommendations for particular programs/software to do it?

In short, anything which uses an FIR type of filter....  by using a longer impulse for the filter it increases the computational intensity, and the latency.....  but makes the ripple less, and the transition to stopband faster.

 

Essentially it is making the "brickwall filter" .... more like it is supposed to be.....   and then to the degree to which you aren't "perfect", shaping the sinc function to account for this.

 

Examples of players which look down this rabbit hole are HQplayer and XXhighend ....  but any type of FIR filter will benefit from this - as you'll see in any filter designer  (longer filters = more precision).

[davewantsmoore]

2 hours ago, Music2496 said:

Is that from Chord's own tech paper?

Yes, it was a direct quote from the Chord paper on the M Scaler.

 

 

EDIT....

Wow.   Well, aren't I an idiot.    It is not from a paper written by Chord, but by "Keith Howard" ......  The article is published on the Chord website.

Edited January 18, 2019 by davewantsmoore

[davewantsmoore]

19 hours ago, Music2496 said:

 

Jussi Laako (of HQPlayer) had a comment related to this:

 

"looked from two extremes, both Chord talking about transient accuracy with extremely long filters and MQA talking about transient accuracy with extremely short filters are both right in a way, but only looking at things from one point of view while ignoring others. As usual in life, truth is somewhere between the extremes..."

 

Heh, very much so.   Although I don't see the approaches are mutually exclusive - You can use both.... so it isn't like you need to do a compromise between one and the other.

 

The approach to use a long filter isn't at all new or unique .... it's the fundamental basics of how these types of filters work.    The things which is new is the ability to practically implement them, with anything approaching acceptable latency, for even the hifi playback system market (which is relatively limited) ..... they're obviously unacceptable for many 'pro' uses.

[rand129678]

 

3 hours ago, davewantsmoore said:

Yes, it was a direct quote from the Chord paper on the M Scaler.

EDIT....

Wow.   Well, aren't I an idiot.    It is not from a paper written by Chord, but by "Keith Howard" ...

Yes, as I mentioned

 

https://www.stereo.net.au/forums/topic/260051-chord-hugo-m-scaler-discussion-and-impressions-thread/?do=findComment&comment=4049644

 

 

3 hours ago, davewantsmoore said:

The article is published on the Chord website.

Bizarre. 

 

The dead giveaway that they didn't check with Rob before publishing that article (and you quoted earlier...):

 

"But, like I say, it’s easy, it’s cheap, and it allows you to generate and listen to a file that’s been oversampled using full sinc interpolation, in which respect it’s even better than the M Scaler."

Edited January 18, 2019 by Music2496

[davewantsmoore]

1 hour ago, Music2496 said:

Bizarre. 

Why?

1 hour ago, Music2496 said:

The dead giveaway that they didn't check with Rob before publishing that article (and you quoted earlier...):

 

"But, like I say, it’s easy, it’s cheap, and it allows you to generate and listen to a file that’s been oversampled using full sinc interpolation, in which respect it’s even better than the M Scaler."

Why?   It is the truth.

 

If Chord are serious in their approach to infinite sinc and perfect brickwall filters ..... then they can't deny that solutions which implement it better than they do aren't superior.

 

Of course, I they wouldn't think that people who are going to "do it themselves" are really their target market ..... so it's unlikely to hurt sales if they admit this.

[Guest]

@davewantsmoore I agree, the M-Scaler is the easier solution, but an expensive one. 

[rand129678]

12 minutes ago, davewantsmoore said:

Why?   It is the truth.

I agree it is the truth (I use HQPlayer...) but anyone that's spoken to Rob (directly) will know he's not a fan of any software up-sampling with his DACs...

 

" in which respect it’s even better than the M Scaler."

 

Not a chance he would ever agree to this publicly... so to have that article on his M-Scaler page tells me they didn't check with him 

 

One example (and there are many):

 

https://www.head-fi.org/threads/chord-electronics-dave.766517/page-693#post-14110123

 

So yeh, that was the dead giveaway for me, that what you quoted wasn't a Chord paper, even before finding the article and reading it... strange they would add  that article to his M-Scaler's page.

 

Edited January 18, 2019 by Music2496