hypex ncore

[Julf]

Originally Posted by barrows

you mean to suggest, Julf, that you cannot remember the sound of your mother's (or sister's, daughters', brother's, etc) voice on the telephone (not exactly high res sound there either) unless you just talked to them 10 seconds ago?

Of course I can. As you probably know all too well, telephone is 8 bits at a sample rate of 8 kHz, and uses a tiny crappy mic and a not-much-larger speaker driven by a digital amp with 10% distortion. The fact that I still recognize their voices so well is great proof of the ability of my brain to only extract and remember the essential (for a human) parts of the sound, while blocking out any shortcomings in the reproduction.

Great for recognizing voices, less great for acting as an absolute reference.

In any case, it would not be too hard to do a mono test to AB and even X if one really needed to.

Still in the spirit of fun - did you watch the Penn & Teller video I linked to?

Here it is again:

Penn & Teller: ********! - The truth about bottled water

If you don't have the 8 minutes to watch it, it shows how, when people were served regular tap water, but in fancy bottles and with fancy descriptions and price tags, they described in rather poetic ways how much clearer, brighter, deeper, more resolving, whatever the "fancy" water tasted compared to tap water. And the most expensive water usually tasted the best.

Considering how strong a factor perceptual bias is, how would you guard against it without a double-blind ABX test?

OK, even this thread seems to have turned into one of those endless and circular "objective" vs "subjective" debates - should we just get back to discussing hypex amps?

[Esperado]

Originally Posted by Julf

The SMPS600 has synchronous rectification.

You mean there is a connection to synchronize the oscillator of the SMPS with the one of the amp ?

[Julf]

Originally Posted by Esperado

You mean there is a connection to synchronize the oscillator of the SMPS with the one of the amp ?

At this point, when this thread is 561 pages long, I think it would be pretty cruel to ask you to read the whole discussion, so here are a couple of pertinent quotes from Bruno:

"There is not a single regulator on the SMPS600. Its main benefit, as I have stated several times in this thread, lies in the synchronously rectified secondary side. The UcD and NC amps have symmetrical PSRR. Since the synchronous rectifiers force the ripple to be symmetrical, it cancels in the amp. As a result, the benefit of regulation (reduced AC mains components in the output signal) is obtained without the trouble of regulation."

"The SMPS puts out a 100Hz sawtooth just like an ordinary supply. It is not clean and it has lots of harmonics. The output of the amplifier is the sum of both rails divided by PSRR. PSRR is about 80dB and the SMPS is synchronously rectified so the ripple is symmetrical to begin with. That's all there is to explain!"

[Ouroboros]

Originally Posted by Esperado

You mean there is a connection to synchronize the oscillator of the SMPS with the one of the amp ?

Synchronous rectification in an SMPSU means that the output rectifier diodes are replaced by MOSFETs. The gates of the MOSFETs are driven by signals from the SMPSU controller chip so that the correct MOSFET gets turned on when forward conduction is required to perform the rectifying action.
The main advantage is a lower forward voltage drop than a diode would have. Sync rectification helps efficiency a lot if the output voltage of the supply is low.

[Julf]

Originally Posted by Ouroboros

Synchronous rectification in an SMPSU means that the output rectifier diodes are replaced by MOSFETs. The gates of the MOSFETs are driven by signals from the SMPSU controller chip so that the correct MOSFET gets turned on when forward conduction is required to perform the rectifying action.
The main advantage is a lower forward voltage drop than a diode would have. Sync rectification helps efficiency a lot if the output voltage of the supply is low.

Indeed, but in this case the even more important benefit is that the ripple from the PSU is symmetrical - so if the amp has good common mode PSRR, the ripple gets cancelled out.

So in this case we have a design where the properties of the SMPS and the amp support each other.

[tony399]

Nicely pointed out Julf

[mr_push_pull]

Originally Posted by barrows

you mean to suggest, Julf, that you cannot remember the sound of your mother's (or sister's, daughters', brother's, etc) voice on the telephone (not exactly high res sound there either) unless you just talked to them 10 seconds ago?

straw man argument. voice recognition is a different matter. you recognize a person's voice based on specific "features" (this is exactly what they're called in voice recognition algorithms jargon). those features are not only based on tone but also on other things like intonation etc which to an extent have a limited amount of variability in each person. and (unlike many things in audio) those features are measurable and quantifiable, otherwise voice recognition would not work. so the idea is that your brain uses all the data available. with audio, intonation (or whatever its correspondence in music) for instance does not vary, the "sense of rhythm" that many reviewers talk about is not related to that, as there aren't significant variations in say frequency of a DAC's clock, compared to speed variations found in human speech.
take part of those features away (alter them in software for instance) and suddenly the voice of your sister becomes unrecognizable. it has many times happened to me that close friends called from another phone than the one they normally use and I wasn't able to recognize them. I could almost swear that someone is pulling jokes on me. I'm sure it happened to you too.

also, just as there are visual illusions, there are auditory illusions too. one of them, related to speech perception (I can't remember the name right now) goes basically like this: sound "1" is played, you look at a face pronouncing sound "2" and you hear sound "3". occurrence varies from language to language and from person to person. and I think everyone knows how much perception changes when one listens with eyes closed.

see, it's not that simple. you can be easily tricked. I remember a Nordost demo (available on video somewhere) where they were telling the audience what improvements they were going to hear with their cables before music even started. I'm not sure refraining from doing so makes a demo like that a "blind test", if that's what they wanted to avoid.

Originally Posted by barrows

The idea that sonic memory is so short is flawed.

why? because someone with perceived authority (who sometimes happens to try to sell something) said so? or because someone actually proved it?

see, we're bound to run in circles. subjectivist says "auditory memory is flawed", subjectivist says "DBT is the way to go" etc. it goes on forever.

Originally Posted by barrows

Also, listening is a skill which can be made better through practice. I would entirely discount any studies made on the "population at large", the same way I would not suggest that the average driver could step into an F1 car and be expected to equal Lewis Hamilton's lap times.

this time I agree and IMO this is why ABX/DBT gets a bad rep. every time I read a study which says that subjects were chosen from among students, I hit ALT+F4. unfortunately many do so because it's obviously the cheapest way.
but it doesn't have to be like that. it's automatically implied that DBT is done with untrained listeners, that inadequate or unfamiliar music is used, that too short song samples are used etc.
but it can be done with trained listeners, with familiar music of their choice, switching can be done whenever the listeners choose so, prior sighted evaluation of tested equipment allowed (and advised). it can also be done in many sessions to avoid fatigue etc. the problem is that it is expensive to do and unless there's something to gain from it it's unlikely that anyone will set up a test of the kind just for the heck of it. H/K do it but they're designing products that are supposed to bring profit.

but, at the same time it should be noted that audiophiles many times report extraordinary (or replace with whatever audiophile review jargon word) improvements. extraordinary somehow implies that those differences should remain obvious even with less than ideal conditions.

PS: there was a thread discussing ABX/DBT on hydrogenaudio (yes, many will shudder with disgust at the name of that objectivist hive) where the issue of auditory memory has been brought up by people who actually studied psychoacoustics and if I recall correctly they go into more detail, I think it'd be easy to find

PS2: and instead of idle, endless speculation and reinventing the science, maybe we first understand what has already been studied: http://en.wikipedia.org/wiki/Echoic_memory
many times you'll find that issues that seem novel are just things that are periodically resurrected since they were first studied, usually long time ago.

[Julf]

Originally Posted by tony399

Nicely pointed out Julf

Thanks - that is why I am just somewhat confounded as to why some people insist on wanting to use a linear supply - so far I have not seen any objective evidence or rational as to why a linear supply would be better than the purpose-designed SMPS600. I can only attribute it to some audiophile mythology about "linear good, switching bad".

Actually, no. That is unfair. I guess blaming it on audiophile mythology is too simplistic. This is diyaudio after all, and people are here because they want something more than what you can buy in any audio store - either something that is "custom" or something they at least to some extent created themselves.

So, what sounds better, "I got these really great ready-to-use modules and put them in a box", or "These amps are really revolutionary, but the guy who designed them, while brilliant, seems too obsessed with switchmode supplies (probably for commercial reasons), and we all know that that is not optimal, so *I* created this totally unique amp by combining these state-of-art modules with some really-state-of-art rectifiers and capacitors I soldered together"? We are all humans, after all.

[Julf]

Originally Posted by mr_push_pull

yes, many will shudder with disgust at the name of that objectivist hive

Ah, you mean the fact-obsessed bunch?

[mr_push_pull]

Originally Posted by Julf

Ah, you mean the fact-obsessed bunch?

yes, "small, joyless people" as Mr. Atkinson once called them which refuse to give in to the proven healing qualities of high-end audio shopping.