Head-Fi.org › Forums › Equipment Forums › Sound Science › Ken Ishiwata on CMOS and the modern PCB
New Posts  All Forums:Forum Nav:

Ken Ishiwata on CMOS and the modern PCB

post #1 of 29
Thread Starter 

Hi All,

 

    My understanding of electronics is very basic, but Ken seems adamant that CMOS circuitry cant generate enough current for audio - the design actually impedes current generation to save power. He claiims - and I dont want to get dragged into this - that 24-bit DACs often sound worse to his golden ears than older 16-bit designs. Granted, Ken is an aging holdout from the analog era, but he has spearheaded the design teams behind some very sweet gear in the years since Marantz abandoned the vacuum tube. No shortage of Head-Fiers who swear that the headphone stage in older Marantz receivers is on par with the best dedicated headamps out there - I cant comment on that personally, just putting it out there. I've seen some of the really old boards out of mainframe computers, and they were seriously chunky pieces of silicon running up huge power bills, but what's good for computing and the environment may not be ideal for audio - hopefully there is a middle ground. 

 

   Happy to hear from those with a better technical understanding than myself, If its back to point-to-pont wiring, there's always Bottlehead.  biggrin.gif

 

Tks,

 

estreeter

post #2 of 29

Could you please post any links you might have regarding this subject?

 

Thanks!
 

post #3 of 29

One thing is the current needed to drive even low impedance headphones.  Quite another is the current needed to drive a typical loudspeaker.

post #4 of 29
Quote:
Originally Posted by Mauricio View Post

One thing is the current needed to drive even low impedance headphones.  Quite another is the current needed to drive a typical loudspeaker.

 

In any case, a CMOS DAC does not have to directly drive headphones, let alone speakers, so its limited current output has little if anything to do with sound quality.

post #5 of 29
Thread Starter 

Mauricio, I think I know where you are going with this, and I'm not going there.

 

For the rest, here is the Youtube interview - have at it. 

 

post #6 of 29
Quote:

Originally Posted by estreeter View Post


No shortage of Head-Fiers who swear that the headphone stage in older Marantz receivers is on par with the best dedicated headamps out there - I cant comment on that personally, just putting it out there.

 

Really? My Marantz receiver doesn't even have a dedicated headphone stage. It simply uses resistor drops from normal speaker outputs to the headphone jack.

post #7 of 29
Thread Starter 

You need to  take  that up with Skylab and  the other folk in the Vintage Receivers thread - the 'afterthought' approach you describe is more common in modern AV receivers and entry-level stereo amps. Move up the pricing ladder and I believe it gets better, but even my old CA 340SE gave me a better result from the headphone out than a $500 DAC/amp I purchased shortly after joining Head-Fi. I know it wasnt placebo because the Yamaha HT receiver I had in the same period had a woeful headphone out - a complete afterthought on a $400 box of silicon. 

post #8 of 29
Quote:
Originally Posted by estreeter View Post

You need to  take  that up with Skylab and  the other folk in the Vintage Receivers thread - the 'afterthought' approach you describe is more common in modern AV receivers and entry-level stereo amps. Move up the pricing ladder and I believe it gets better, but even my old CA 340SE gave me a better result from the headphone out than a $500 DAC/amp I purchased shortly after joining Head-Fi. I know it wasnt placebo because the Yamaha HT receiver I had in the same period had a woeful headphone out - a complete afterthought on a $400 box of silicon. 

 


The Marantz receiver that I'm talking about is a vintage Marantz 2220, manufactured in 1973 IIRC. Not a high end unit by any means (none of the 2xxx series were to be fair), but it sounds great with my loudspeakers. The headphone jack was obviously an afterthought though, but it functions well enough.

 

I'm just trying to understand the notion that these vintage units had a good headphone output. As far as I know, using resistor drops was common technique, but maybe some other vintage units actually did have quality headphone outputs.

post #9 of 29

After watching that video all I can think is that was ten minutes of my life I'll never get back.  Seemed like pure marketing speak to me: all noise and no signal.

post #10 of 29

I was talking about amplifiers, components that may be required to put out large currents.  High current is not really a requirement for pre-amplification, line-level microelectronics.


Edited by Mauricio - 6/11/12 at 8:51pm
post #11 of 29
Thread Starter 
Quote:
Originally Posted by BlindInOneEar View Post

After watching that video all I can think is that was ten minutes of my life I'll never get back.  Seemed like pure marketing speak to me: all noise and no signal.

 

I suspect that its a fairly old vid - Ken refers to 8GB portable players as though that was a big deal, and the CMOS reference seemed equally dated - havent modern PCs been using CMOS for decades ?? As for the  marketing, prior to retirement Ishiwata was their Brand Ambassador - a title that reeks of something Marketing came up with. Still, 10 minutes of your life - clearly you kept watching beyond the intro !  biggrin.gif

post #12 of 29

Funniest thing is that he says that test signals are constant and music changes every second and is more dynamic. t_t

I guess he never heard of noise, sweeps, MLS or the like.


Edited by xnor - 6/10/12 at 5:10pm
post #13 of 29
Thread Starter 

I think he was referring to the way measurements are published, but if you read any decent analysis of 'proper' lab measurements there is usually an attempt to acknowledge that by pinpointing different readings across the frequency curve / power band. When was the last time you saw a news report on an earthquake or big wave surfing which mentioned anything other than the peak measurement ?

 

All of that aside, Ken's 'creations' seem to measure very well ('superbly' to quote John Atkinson) in the reviews I've read. 

post #14 of 29

Actually, I think I spent more than ten minutes on that video, having to replay a few bits to grapple with Ken's accent.  Even so I'd still not feel comfortable attempting to paraphrase much of what he said.  However I would like to point out a few things. 

 

First, there are acknowledged tests, accepted industry wide, for measuring the performance of an audio component.  If Ken is going to dismiss these tests out of hand as being meaningless or incomplete then it is incumbent upon him to explain and show why they are deficient and to fully describe his proposed alternative.  In that video he did neither.  Sorry, sighted listening tests are not an acceptable substitute for valid science.

 

Secondly, at one point he seems to be saying that mid and high frequencies (in audio terms) pass through a circuit more quickly than lower frequencies.  I am completely at a loss as to how this can happen.  I submit that an audio circuit neither knows nor cares whether it is passing a 100 hz signal or a 1,000 hz signal.  If a midrange signal does pass through a circuit faster than a bass signal, how is it no one noticed it before Ken?  Why hasn't Ken published a paper describing this effect?  Why hasn't anyone else?  This would seem to be a significant issue that circuit designers would have to address, yet I'm unaware of any reputable discussion of the matter.

 

Ken struck me as being an amiable guy and an inveterate schmoozer.  Good for him that he's got such a good job.  However, I'd be very surprised if he actually spent much time at the test bench while any of his "creations" were being designed. 
 

post #15 of 29
Quote:
Originally Posted by BlindInOneEar View Post

 

 

Secondly, at one point he seems to be saying that mid and high frequencies (in audio terms) pass through a circuit more quickly than lower frequencies.  I am completely at a loss as to how this can happen.  I submit that an audio circuit neither knows nor cares whether it is passing a 100 hz signal or a 1,000 hz signal.  If a midrange signal does pass through a circuit faster than a bass signal, how is it no one noticed it before Ken?  Why hasn't Ken published a paper describing this effect?  Why hasn't anyone else?  This would seem to be a significant issue that circuit designers would have to address, yet I'm unaware of any reputable discussion of the matter.

 

 

 

This is a well-known phenomenon.  It is often called group delay and a related phenomenon is the the concept of phase difference.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Ken Ishiwata on CMOS and the modern PCB