[x RELIC x]

He claims tap quality and implementation are equally or more important than numbers.

I really don't want to get too deep in to this conversation, BUT, Rob hasn't ever claimed 'just numbers' are important. In fact he says that his TAPs with his WTA algorithm are what make it good. On the other hand, Rob has pointed out that other filters are 'just numbers' without the proper algorithm.

Looks like each side is accusing the other side of the same thing, LOL.

 

[Foxman50]

You are probably right. But only as far as DAVE/M Scaler are concerned imho.
And personally I can only say it is the best commercially available combo I have heard so far no more!

The other Chord DACs have some real serious competition to fight against imho.

And it is always interesting to read about what other competitors are claiming and doing too.

This guy Jussi also claims Chord's and Rob's claims regarding tap numbers are just " techno babble" in one post.
He claims tap quality and implementation are equally or more important than numbers.
He also claims to have refuted some of Rob's claims. Not sure which though?
Rob never posts or responds at CA as far as I know.
On the other hand the DAC Miska refers to in the link you provided, doesn't even seem to exist in the physical realm at all, so of course I take what he claims with more than "a pinch of salt".

Game Rob, it seems, his dacs exist!

But nor do I swallow everything any other digital GURU,even Rob claims without listening for myself when I can.

The only way to know for yourself which is best is to audition and compare different DACs and other links in a reproduction chain.
Currently I am quite happy with Qutest/MScaler as my travel combo. It takes quite a good ,but far from perfect DAC on its own, to another clearly higher level.
But I also know for sure that a DAVE/MScaler is even better and reproduces acoustic music closer to how I am used to hearing it live both at concerts and rehearsals and recording sessions, which are my only reference points.
The problem for me and many other music lovers is that Rob's tech at its best without commercial compromises thrown in still comes at costs levels unobtainable for most.

Qutest is an intentionally compromised product geared towards an intended market section.
Some of the compromises are very irritating to me on a daily basis.
I hate the fact that it can't turned off without plugging it out and the micro usb power port is a really bad choice imho.
Luckily I am in one and the same Condo since a month and a half, and at least I don't neeed plug in and out the power micro usb at all for the time being.
Just seeing which is up and down on that tiny thing is troublesome in anything but good light, not to mention all the wear involved.
On my old HUGO 1 the supplied usb micro cable stopped working within months for all the wear and tear involved in daily use.
When travelling with a Qutest or H2 and moving around a lot it is a PITA with Qutest too.
Unfortunately almost the same applies to M Scaler as far as lack of an on and off button is concerned.
But at least it has got a proper power connection point.
What were they thinking of when designing their products this way?
Are there no thunder storms in the UK???
Oh yes there are!
And even more of them in the tropics!

I pull the plug from the socket every day to turn M Scaler on and off and since the battery powered LPS I am using with Qutest actually has got an on and off switch I don't need to unplug Qutest at the silly end!


Personally I could upgrade to a DAVE/M Scaler but I am also although my own tech knowledge is very limited compared to many here and Rob himself of course,also clearly aware that DAVE is NOT the best that CAN be done.
It may have been but isn't any longer.

Why invest such really BIG money in something that is surely going to be redundant as Benchmark before long?

Whatever Rob or Chord claim here higher resolving even more transparent to the source Digital Toys will come probably both from Rob and maybe others too before long.

In theory Rob could easily build a DAC that would outperform both TT2 and DAVE simply by combining the best of both and adding say 40 elements as obviously planned for Davina if it ever materializes?
Or why not 64 if even ESS chips can do that if elements are so important???

And also KEEP IN MIND so far even via Robs DACs we are ALL listening to music recorded with or digitalized from analouge with other ADCs!!!
DAVINA seems to have slipped into Sleeping Beauty mode for MUCH longer than I had originally hoped!

Imho and without any malicious intent or purpose, my advice to Rob as far as his ADC project is concerned would be to skip the decimation problem question, for the time being, and get down to actually delivering the very best possible digital recording and reproduction device without any compromises involved at all.

Personally I am currently trying to figure out why DAVE sounds so much better with M Scaler than both TT2 and Qutest or H2 with M Scaler?
It obviously has got nothing whatsoever, or at least, little to do with the number of taps employed, since with all these other Chord dacs, the inherent ones are bypassed anyway.
I see that as a strange WASTE!
Why throw away 164000 or 90000 taps into slumber if they are so important?
I personally see little reason to even consider a TT2 for my own personal use and needs on its own.
Its only advantage for me would be its superior anlogue section/ output, over say a Qutest as I understand this, maybe wrongly?

I know it sounds very good in combination with M scaler via high quality headphones.

But on its own it is also a clearly intentionally 10 elements only, compromised product which good as it sounds is clearly outperformed by both Qutest /MScaler and of course DAVE on its own.

For me nearfield speaker listening is basically of no interest at all.
I have little interest in hearing a symphony orchestra delivered that way.

I would much rather,and do so on a daily basis while travelling, use headphones for such compromised listening conditions than a mini orchestra with compressed dynamics for lack of power.

Sorry about the overlong rant from me again, tounge in cheek, hoping that Rob will use boxing gloves and not bare fists this time around if we meet at Singapore Canjam in a few weeks time again.
I am not expecting anything new of interest for me there personally from Chord but Benchmark's HPA 4 sounds tempting on paper at least and there are some headphones I would like to see and hear.
Cheers Controversial Christer

Classic case of over analysing if you ask me.
If you can afford it and like it then buy it. Chord will always create compromised products at various price points, its called business.
As for the power plug issue, unplug it at the wall if you feel the need.
Relax and enjoy the music.

 

[Christer]

At the end of the day all this tech talk doesn't matter, your ears will tell you which is best. Regardless what you buy it will be outdated, that's life.
Don't worry about the elements too much, they run at different frequencies/implementation etc, not always comparable. Anyway, the pure 1-bit DSD dacs are best and multilevel is crap supposedly.
There are many ways to skin a cat, none are perfect. In hindsight, where we ended up may not be the ideal... we can say that about a lot of things though.

Interesting take on these things.
But personally , and without much tech knowledge I still have an interest in the possible progress and advances of digital recording and reproduction that's why I rant away here at times. Rob probably would say :"No there is only one way to skin this particular cat" but I also want to hear what others have to say on the subject.

Furthermore you can both "skin a cat" and "skim" it I don't like being skimmed.
Or suspect that I might be skimmed.
There are claims that the skinning of the cat can be be done as well or better than Rob's way and at much lower costs too.
I have yet to hear any that does but does not rule out the possiblity of it happening.

As far as the DSD/PCM fight goes although I personally know some pros in the classical world who were firmly in the DSD camp and their recordings can sound excellent .But I tend to suspect that Rob might be absolutely correct about at least DSD 64.
Things may even out and possibly be somewhat different from DSD256 and up to DSD 512 and recently at least in theory DSD 1024 I don't know for sure more than that there are contrary claims from both camps.
I have but one native raw masterfile from one such source in both DSD 64/128 and 256 and DXD and to me the DXD sounds slightly closer to how I have remember that orchestra sounded in that hall.
Via M Scaler it can sound very lifelike indeed.
But what I look forward to hearing asap is native 768KHz against native DSD512 or DSD 1024. That would be interesting I think.



Even via my relatively humble travel kit Qutest/M Scaler/Sennheiser HD800 via battey powered headphone amp,as good as some DSD 64 recordings I have actually been to myself can sound, really well recorded hi res pcm can sound more transparent than DSD 64.

And I am talking about often large scale very complex symphonic and operatic material,not the best DSD or PCM version of Hotel California or such,which some use as a reference in such SQ comparisons.
The more complex the scoring the more obvious the differences are to my ears.
100 instruments and voices is a lot more to handle than any band IMO.
Some cats are actually bigger than others!
Cheers Controversial Christer

 

[Christer]

At the end of the day all this tech talk doesn't matter, your ears will tell you which is best. Regardless what you buy it will be outdated, that's life.
Don't worry about the elements too much, they run at different frequencies/implementation etc, not always comparable. Anyway, the pure 1-bit DSD dacs are best and multilevel is crap supposedly.
There are many ways to skin a cat, none are perfect. In hindsight, where we ended up may not be the ideal... we can say that about a lot of things though.

Interesting take on these things.
But personally , and without much tech knowledge I still have an interest in the possible progress and advances of digital recording and reproduction that's why I rant away here at times. Rob probably would say :"No there is only one way to skin this particular cat" but I also want to hear what others have to say on the subject.

Furthermore you can both "skin a cat" and "skim" it I don't like being skimmed.
Or suspect that I might be skimmed.
There are claims that the skinning of the cat can be be done as well or better than Rob's way and at much lower costs too.
I have yet to hear any that does but does not rule out the possiblity of it happening.

As far as the DSD/PCM fight goes although I personally know some pros in the classical world who were firmly in the DSD camp and their recordings can sound excellent .But I tend to suspect that Rob might be absolutely correct about at least DSD 64.
Things may even out and possibly be somewhat different from DSD256 and up to DSD 512 and recently at least in theory DSD 1024 I don't know for sure more than that there are contrary claims from both camps.
I have but one native raw masterfile from one such source in both DSD 64/128 and 256 and DXD and to me the DXD sounds slightly closer to how I have remember that orchestra sounded in that hall.
Via M Scaler it can sound very lifelike indeed.
But what I look forward to hearing asap is native 768KHz against native DSD512 or DSD 1024. That would be interesting I think.



Even via my relatively humble travel kit Qutest/M Scaler/Sennheiser HD800 via battey powered headphone amp,as good as some DSD 64 recordings I have actually been to myself can sound, really well recorded hi res pcm can sound more transparent than DSD 64.

And I am talking about often large scale very complex symphonic and operatic material,not the best DSD or PCM version of Hotel California or such,which some use as a reference in such SQ comparisons.
The more complex the scoring the more obvious the differences are to my ears.
100 instruments and voices is a lot more to handle than any band IMO.
Some cats are actually bigger than others!
Cheers Controversial Christer

 

[Christer]

I really don't want to get too deep in to this conversation, BUT, Rob hasn't ever claimed 'just numbers' are important. In fact he says that his TAPs with his WTA algorithm are what make it good. On the other hand, Rob has pointed out that other filters are 'just numbers' without the proper algorithm.

Looks like each side is accusing the other side of the same thing, LOL.

Yes it does look like they do and I find that a bit contradictory and would like to know if both can be right or not?
Cheers Controversial Christer

 

[mlxx]

Yes it does look like they do and I find that a bit contradictory and would like to know if both can be right or not?

Quality of taps always matters. The number of taps depends on what you are trying to do and also what is possible with the processing power available.
For what the mscaler is trying to do, we need an infinite sinc tap filter ideally, so we need quality and number here.

 

[Triode User]

Quality of taps always matters. The number of taps depends on what you are trying to do and also what is possible with the processing power available.
For what the mscaler is trying to do, we need an infinite sinc tap filter ideally, so we need quality and number here.

. . . . and Rob's algorithms (which others of course do not have).

 

[mlxx]

Lets simplify things and compare to DSD1024...
You can think of the pulse array dac as a "better" DSD dac. The pulse array already runs at 2048FS, that is like DSD2048. Elements are not like bits, they are levels, going from 32e to 64e is like adding 1 more bit, so from 5 bits to 6 bits. Dave has 20e running at twice the frequency of DSD1024. The mscaler is as close to perfect in its reconstruction that we can currently get in realtime. Plus the noise shapers, amp section etc, technically, I don't see it being any contest (and the measurements show it). I also think mDave sounds the best. Yeah it costs a lot and will be outdated, be glad you didn't get an Abyss Phi CC recently.

 

[Christer]

Quality of taps always matters. The number of taps depends on what you are trying to do and also what is possible with the processing power available.
For what the mscaler is trying to do, we need an infinite sinc tap filter ideally, so we need quality and number here.

But what is quality of taps?
I remember a thing someone said about number of taps comparing the number of taps to the number of measuring points on a chart or map of a coastline. The more measurement points and there are along the coastline the more accurate the chart or map. Sounds rather logical doesn't it?
But quality of taps?
Cheers Controversial Christer

 

[Christer]

Lets simplify things and compare to DSD1024...
You can think of the pulse array dac as a "better" DSD dac. The pulse array already runs at 2048FS, that is like DSD2048. Elements are not like bits, they are levels, going from 32e to 64e is like adding 1 more bit, so from 5 bits to 6 bits. Dave has 20e running at twice the frequency of DSD1024. The mscaler is as close to perfect in its reconstruction that we can currently get in realtime. Plus the noise shapers, amp section etc, technically, I don't see it being any contest (and the measurements show it). I also think mDave sounds the best. Yeah it costs a lot and will be outdated, be glad you didn't get an Abyss Phi CC recently.

Thanks that explains things more clearly to me.
And tallies well with what I keep hearing via M Scaler in comparison with native DSD via some other DACs. I was under the obviously wrong impression that at least DSD 1024 was a more advanced "faster" higher sampling rate than M Scaler.
Thanks for educating me.
It is quite comforting to be told that what I hear from my M Scaler obviously also is the most advanced tech in spite of what some other sources claim?
Abyss warning?
Well I only bought the then still very expensive HEKV2 a couple of years ago but I am not so sure any longer that it is really that much better than my old Sennheiser HD 800 now that I have the M Scaler.
I don't travel with my set this winter but during my auditions of M Scaler with quite a few high quality headphones it was not as convincing as I had expected it to be any longer.
And Abyss,well that is a heavy beast I don't want to wear more than a few minutes.
I will put my HEKV2 through tough listening tests later this spring.
I have to confess I am quite happy with my old HD 800 with M Scaler/Qutest for now.
I don't really miss the HEKV2 much.
But I am beginning to miss my usual weekly live classical concerts.
I have attended over 20 live concerts and some rehearsals this winter in Asia.
But I have been in Thailand over a month now without any concerts. I don't visit Bangkok I don't want to breathe Bangkok's air.
I won't get to hear an orchestra live until mid March again.
On the other hand I can swim one k daily in warm clear blue water and breathe fresh air and listen to my favourite music via M Scaler and headphones.


Cheers Controversial Christer

 

[mlxx]

But what is quality of taps?
I remember a thing someone said about number of taps comparing the number of taps to the number of measuring points on a chart or map of a coastline. The more measurement points and there are along the coastline the more accurate the chart or map. Sounds rather logical doesn't it?
But quality of taps?
Cheers Controversial Christer

For the mscaler, it is a sinc function as per the sampling theory, its a formula you can use to get the tap values and a window function. For other situations it is not as clear, you have to design to meet specs and resources. Like the video filter on the mscaler, it has to have low latency so you have to compromise here, it becomes harder to design and which is better becomes debatable. A digital filter is like a mathematical transformation, what needs to be done to a set of numbers to achieve what we want.

 

[Christer]

For the mscaler, it is a sinc function as per the sampling theory, its a formula you can use to get the tap values and a window function. For other situations it is not as clear, you have to design to meet specs and resources. Like the video filter on the mscaler, it has to have low latency so you have to compromise here, it becomes harder to design and which is better becomes debatable. A digital filter is like a mathematical transformation, what needs to be done to a set of numbers to achieve what we want.

Thanks again, Yes to me it seems the real Achilles heel of M Scaler tech seems to be this latency problem.
I haven't really used the video function/ setting much since I am on very slow wifi here and can't stream any music videos .
But when I could it was strange to see how obvious the latency problem was without the filter set for video.
I normally watch a lot of classical concerts too and it is weird when the conductor's beat is off or lip synch is also off or a pianist's fingers on the keyboard and the notes heard are not in synch.
In an ideal world I would like all the 1 M taps fully at work even with videos. But that seems impossible to achieve?
Cheers Controversial Christer

 

[mlxx]

Yes for the ideal sinc FIR case... with long tap length, 1M taps, you have to wait for half the samples because it looks into the future, so 500k samples which is too long for video. IIR has no delay but no good for this case. Rob did an asymmetric FIR to reduce it, has less taps for future.

 

[dawktah2]

Dear (dawktah2),

Thank you for your order from Moon-Audio.com! We wanted to let you know that your order (#10000XXXXX) was shipped via FedEx, FedEx Ground® on 2/19/2019. Whoo Hoo!

 

[ZappaMan]

Dear (dawktah2),

Thank you for your order from Moon-Audio.com! We wanted to let you know that your order (#10000XXXXX) was shipped via FedEx, FedEx Ground® on 2/19/2019. Whoo Hoo!

Sweet