New R2R Holo DAC coming the KTE Spring-2

[eltech]

37 minutes ago, davewantsmoore said:

You bang on relentlessly in the Class-D thread that this is totally unacceptable filtering performance ..... Yet it is fine here (?!)    I'm confused.

A very legitimate question. IMO

[georgehifi]

8 minutes ago, eltech said:

What class D amp doesn't use an output filter?

None as far as I know. I told Dave to have a look.

8 minutes ago, eltech said:

I've been asking you for a photo of a 15khz sine wave on the output of the Holo, but you don't want to show it, or discuss what you see. 

Find your own, I have never taken a pic of one or commented on one. It's you that are "banging on" about it

Edited December 3, 2018 by georgehifi

[davewantsmoore]

1 hour ago, georgehifi said:

second about the phase shift

Exactly.   I just showed a picture of the phase shift  (implied by the amplitude response).

1 hour ago, georgehifi said:

On the scope go and have a look at the switching noise on the Class-D output without output filter.

I didn't ask about the noise.   I asked about the frequency response (ie. the amplitude response, and phase response).

 

You say this problem is unacceptable in a class-D amplifier .....  yet you say this problem doesn't exist in this DAC (it clearly does).    So, as I said, I'm just really confused as to why you would say these things.

 

... and why you would go an do a big "goal post shift", and start talking about noise.   [ie. you just answer a completely different question to the one I ask]

1 hour ago, georgehifi said:

get out of the wrong side this morning did we?

No.... everything was fine, until I got here.

1 hour ago, rocky500 said:

Nos mode to me just sounds more real and makes the Holo stand out in being more enjoyable than from the other Dacs I have tried.

I agree.

 

As I said before, it is interesting that correcting for the filter response (ie. removing the amplitude and phase error which encroaches into the audible band), removes a lot of the subjective difference ....  although not all IME.

 

It's also interesting that this works the other way too  (applying the error to a device which doesn't have it.... alters the sound subjective towards which a more typical NOS sound is).

 

 

There's a lot of theories around why this stuff happens... and I personally suspect that MQA and those working in similar areas are very much on the right track about this.

1 hour ago, georgehifi said:

You obviously don't have ears Dave. It's not a difference in FR at all, get the wax cleaned out and listen again.

I would not typically bother posting something which was "just my own experience" ..... 'cos why would I expect other to believe that.

 

What I am talking about is the many studies that have concluded the correcting for the amplitude and phase errors introduced in for a NOS approach ..... removes a very big (if not all) proportion of the reported subjective difference (ie. they then sound very close to the same).

1 hour ago, eltech said:

What class D amp doesn't use an output filter?

I'm pretty sure he means without a filter that if often included in the measurement system to reduce the very high frequency noise in the output.

 

However, the problem is I asked nothing about such noise.

 

Either George doesn't understand that's what I was asking....   or he does understand, and has purposefully answered a totally different question.    I've asked to clarify.

1 hour ago, eltech said:

I've been asking you for a photo of a 15khz sine wave on the output of the Holo, but you don't want to show it, or discuss what you see. 

If you look at the measurements of the Holo DAC (and similar DACs) you can see what will happen.

 

There will be both linear and non-linear distortion.

 

The non-linear distortion looks quite bad, compared to what most people are used to seeing.....  eg. up around 0.1% ....  but this shouldn't be audible, and will be much less with real world signal levels (HF are not usually loud).

 

The linear distortion is reasonably significant.    Like all (invariant) linear distortion, it can be corrected simply.

 

 

 

So, to be clear (as it seems I could be taken out of context) .... I don't think any of this is a significant problem (the Holo performs fine for a playback system) .......    but I can't understand how in this thread it's fine .... but in the class-D thread, it's discussed as a seriously significant problem.

 

More perplexing.... is that we have the same person acting as the cheerleader for it being fine here... and a big problem, over there.

[davewantsmoore]

1 hour ago, georgehifi said:

"bang on" really Dave

Yes, really.

 

You go on at great length (of which "bang on" is a colloquialism for) in the class-D thread(s) how the impact of the amplitude and phase errors in the audio-frequency band are unacceptable.

[georgehifi]

You two are a pair, and are "banging on" about nothing, everyone I know who has heard the Holo Spring prefers it in NOS, and there is no penalty in HF detail and or decay, there is just more body to everything above the bass, so go and have a listen instead of "banging on" about 1 or 2db difference at 20khz or 15khz sine waves, which btw all I know can reproduce perfectly. if you guys want to put **** on the Holo, go start your own thread.

Edited December 3, 2018 by georgehifi

[sir sanders zingmore]

I think we have a late contender for the Art Vandalay avoiding-the-question award 2018

[davewantsmoore]

4 minutes ago, georgehifi said:

"banging on" about nothing

I asked a pretty specific, and potentially important question.   It's not "nothing".

4 minutes ago, georgehifi said:

everyone I know who has heard the Holo Spring prefers it in NOS

So do I .... so what?!

4 minutes ago, georgehifi said:

and there is no penalty in HF detail and or decay

You said there was no rolloff in the HF  (ie. amplitude and phase errors).    There IS a rolloff

 

You said the rolloff (amplitude and phase error) is a massive problem for a class D amplifier .....  Yet, it's no problem here.  Why?

 

 

I mentioned interesting results about what happens when the rolloff is corrected for.....   which confirms the idea that the rolloff can be significantly audible....   So, I very much agree that they can be significantly audible in a class-D design...... I just strongly disagree that the only solution is to move the switching frequency higher  (there are other trivial solutions).

 

So why aren't those same issues audible here?    How can the same error be "preferred" for the DAC, but be the source-of-all-evil in a class-D amplifier ?

 

4 minutes ago, georgehifi said:

instead of "banging on" about 1 or 2db difference at 20khz 

I'm asking YOU why you say it's a problem in a class-D amplifier.... but not a problem here.

[davewantsmoore]

14 minutes ago, georgehifi said:

so go and have a listen

?    

 

I have, and preferred NOS, both when the amplitude/phase errors were corrected, and when not corrected  .... but how do you think listening or not is going to change my question?

[davewantsmoore]

2 hours ago, georgehifi said:

You obviously don't have ears Dave

My hearing has nothing to do with any of this mess   

[sir sanders zingmore]

2 minutes ago, davewantsmoore said:

have, and preferred NOS, both when the amplitude/phase errors were corrected, and when not corrected  

Was that blind, level matched etc Dave?

[davewantsmoore]

1 minute ago, Sir Sanders Zingmore said:

Was that blind, level matched etc Dave?

No and yes.   It was just quick mucking about, with some other similar DACs in the mix.

 

As I mentioned after correcting for the major (the linear distortion) difference between the two modes .... nearly all of the subjective difference went away, but I still thought NOS had something.

 

 

 

Also as mentioned..... lots of people have been working on answers to this issue over the past 20 years (eg. MQA).    The answers they've come up with run fairly counter to typical audiophile accepted wisdom.

 

In short.   NOS has significant distortion problems.... but the linear can be corrected easily, and the non-linear isn't very audible.    OS in theory solves lots of problems....  but the filtering we can get running in real time, up until recently (and even now in most cases) has not been of sufficient quality.

 

 

 

[georgehifi]

54 minutes ago, davewantsmoore said:

I'm asking YOU why you say it's a problem in a class-D amplifier.... but not a problem here.

As I told you take the LP output filter off both and see what you see on the scope, and which needs the LP filter more, and you'll answer your own question.

Edited December 3, 2018 by georgehifi

[davewantsmoore]

1 hour ago, georgehifi said:

As I told you take the LP output filter off both and see what you see on the scope, and which needs the LP filter more, and you'll answer your own question.

It doesn't answer either question:

 

1.   Why do you say these amplitude and phase errors are totally fine for a DAC, but a big problem for a class-D amplifier?

 

2.   Why did you say "there was no extra noise or rolling off of the highs with NOS" ?      This is just straight up false.   Every measurement of this DAC shows (as expected) there is more noise and a frequency response roll-off in NOS mode.

[georgehifi]

Told you what to do, if you don't want to do it that's your problem, by-by

Edited December 3, 2018 by georgehifi

[davewantsmoore]

3 hours ago, georgehifi said:

You two are a pair, and are "banging on" about nothing, everyone I know who has heard the Holo Spring prefers it in NOS

Wonderful.  So do I.

 

Hence the question.    Why do you say that exactly the same symptoms (phase and amplitude shift in the audio band) are unacceptable in a class-D amplifier?

 

 

3 hours ago, georgehifi said:

which btw all I know can reproduce perfectly

Well, clearly it doesn't .....  as it has phase and amplitude errors below 20khz.

 

Just to be very clear .... I don't think this is a big problem, and if someone wants it can be corrected for .....  hence the question, why do you say it IS a big problem in a class-D amplifier?

[davewantsmoore]

2 minutes ago, georgehifi said:

Told you what to do, if you don't want to do it that's your problem, by-by

Wow.

 

 

... 'bout that time then, eh.

[georgehifi]

 

This thread is about the new up coming KTE Holo Sping 2 dac, not for you to try to denigrate it and me for posting it before anyone has had a chance to listen to one, so I ask you politely go away and come back when you've slept it off. 

Edited December 3, 2018 by georgehifi

[davewantsmoore]

3 minutes ago, georgehifi said:

This thread is about the new up coming KTE Holo Sping 2 dac, not for you to try to denigrate it

I'm doing no such thing.   It seems like a fine DAC.

 

 

 

Based on this comment it would seem you misunderstand my question(s).    You know I'll happily restate them, etc. ...... so I suspect the misunderstanding is wilful on your part...   In which case?!

[sir sanders zingmore]

3 hours ago, georgehifi said:

everyone I know who has heard the Holo Spring prefers it in NOS, 

I know a few people who prefer to upsample externally and feed that signal to  their Holo in NOS mode

[Guest Muon N']

The Holo is not in my budget, but still interesting to read about, thanks for posting, George.

[eltech]

5 hours ago, Sir Sanders Zingmore said:

I know a few people who prefer to upsample externally and feed that signal to  their Holo in NOS mode

Or they can play native high sample rate recordings without any up or oversampling which is in my view one of the best uses for a NOS DAC.  

But for the record, I have no issues with 44.1khz recordings that are upsampled with good oversampling chips.

And furthermore whether it be with my own recordings or commercial releases, I'm yet to hear any high sample rate recording that sounds significantly better than the 44.1 version. I do hear differences sometimes, but I don't think they are significant, or in other words deal makers or breakers. 

8 hours ago, davewantsmoore said:

and the non-linear isn't very audible

 I'd like to provide two examples which I think easily highlight the audible differences between NOS and  OS.

 

I hear that NOS thickens the upper midrange in an audibly similar way to how second order distortion sounds (if you're familiar with that?).

NOS makes audio sound thicker than it actually is (mostly in the upper midrange and highs) which is caused by aliasing.

 

Two really easy ways to hear this are; 

to listen to hi hats. With NOS they sound very thick and brassy, which initially I thought sounded more realistic, but upon swapping back to OS, it was apparent to me that the thickness is caused by aliasing distortion. With OS, hi hats sound more accurate and clearer.

 

On Foreigner - "I want to know what love is", with oversampling the choir in the chorus is far more clear and each voice can be heard , but with NOS it's difficult to pick out each voice. 

I ought to point out that this NOS characteristic was observed by me on a variety of NOS DACs, but I haven't heard the Holo.

 

 

 

 

Edited December 3, 2018 by eltech

[georgehifi]

8 hours ago, eltech said:

but I haven't heard the Holo.

All that waffle and you haven't even heard it, that says it all.

Better body to the upper mids and highs does not mean thicker.

If you wish to use those terms, then I'd have thicker over thinner any day of the week.

Why when owners of dac's or cdp's that can switch between NOS or OS go for NOS every time.  

Edited December 3, 2018 by georgehifi

[sir sanders zingmore]

31 minutes ago, georgehifi said:

 

Why when owners of dac's or cdp's that can switch between NOS or OS go for NOS every time.  

1. They don’t. Not every time. 

2. Most DACs are not NOS. Most people who buy a Holo probably buy it because they want a NOS DAC, not because they want one that can do both. They’d probably lose very few sales if they removed the OS option. 

If you bought it as a NOS dac obviously you are more likely to use it as a NOS dac. 

[sir sanders zingmore]

39 minutes ago, georgehifi said:

All that waffle and you haven't even heard it, that says it all.

 

Have you heard this particular Holo dac ?

[georgehifi]

24 minutes ago, Sir Sanders Zingmore said:

Have you heard this particular Holo dac ?

I have heard the Level 2 Holo Spring Dac,

This Spring 2 in good authority is said to be even better and cheaper, I wait to hopefully find out on day.