John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
AX tech editor
Joined 2002
Paid Member
My posts are falling behind, but I might pose an interesting scenario for some audiophiles. Let us say that you have a home playback system, composed entirely of vacuum tubes. Of course, that would be impossible for digital, but let's say you have a good vinyl record collection and truly enjoy it. Now, somewhere, at sometime, you found an electronic crossover with IC's in the high frequency thru-path, however it does virtually everything you could ask for in an electronic crossover. Do you think that you could detect a SINGLE IC/channel added in series with your all tube design, while playing an early Joan Baez recording? Let's ask, everyone. ;-)

Well I think it is very likely that it is measureable. Whether it is audible depends on the quality of the tube amp but my gut feeling says its also likely.
Although I wouldn't go for Joan Baez; I didn't find her records very well recorded ;)

jan didden
 
My opinion is that 16 bits is not enough to cover the normal (Whatever that is) human hearing range. I think it really takes 26 bits.

Why? 96dB of dynamic range isn't enough?

If you want to repeat Schroeder's experiment you can create .wav files that contain sine waves and harmonics. Be sure to keep the amplitude constant as you create files with different phase shifts. With a good monitor speaker you can see how much phase shift you can perceive.

I don't listen to sine waves.

Do you?

se
 
Hearing noise at the listening position is rather obvious.

I don't listen to sine waves.

Do you?

se

Even power line noise with the dominant energy at 50 or 60 hertz just at or below the threshold of hearing there are artifacts that creep into the midrange. Swapping the mains plug is noted by many as improving the midrange! You are welcome to come up with any explanation you like to explain the results.

Yes I do listen to sine waves and their fourier combinations.
 
Well that's no problem at all. Measured specs can be compared and repeated and are a relatively 'objective' measure of how clean and transparent an amp processes it's signal. 'Sounds much better' is a very personal and taste-dependent (and about 100 other factors) so really says nothing about the 'quality' of the amp.

Indeed, sound quality is personal taste dependent and also, sometimes, system dependent. That is, the sound quality of a power amp may be dependent upon the speakers it feeds. Therefore, there is no scientific measure for "sound quality".

However, it seems that most people choose speakers for their sound system, between various options within their budget, according to how the speakers sound, not how they measure.

To me and to some other people I know, the same goes for amplifiers, cables, power conditioners and you name it.

The fact that there is no scientific measure for sound quality shouldn't dictate to me that I should choose amplifiers for my personal use only according to how they measure.

My personal sound system serves me to enjoy better reproduced music, not as a basis for scientific paper.

Placebo Effect? Yes, it is there. Yes, I know how to neutralize it (for myself).
 
Even power line noise with the dominant energy at 50 or 60 hertz just at or below the threshold of hearing there are artifacts that creep into the midrange.

If you can't hear that noise at the listening position, then what's the relevance?

Swapping the mains plug is noted by many as improving the midrange! You are welcome to come up with any explanation you like to explain the results.

People note all sorts of similar things after doing all sorts of silly things, like putting photographs of themselves in their freezers, or placing magic crystals next to their speaker terminals, etc.

How do you explain those results?

Yes I do listen to sine waves and their fourier combinations.

You miss the point.

What we can detect with single tones and what we can detect when listening to music are not one and the same.

And what bout my question concerning 96dB of dynamic range? Is that not enough? What is the effective dynamic range of human hearing? Is it even as high as 96dB?

se
 
Let us say that you have a home playback system, composed entirely of vacuum tubes. Of course, that would be impossible for digital, but let's say you have a good vinyl record collection and truly enjoy it. Now, somewhere, at sometime, you found an electronic crossover with IC's in the high frequency thru-path, however it does virtually everything you could ask for in an electronic crossover. Do you think that you could detect a SINGLE IC/channel added in series with your all tube design, while playing an early Joan Baez recording? Let's ask, everyone. ;-)

Do your think you could detect the 400 transistors the signal went through between the time it left the microphone and the time it was engraved on the vinyl? Imagine how many there are just in DolbyA professional. For those who know the old saw about how you woudn't want to see sausauges or laws being made, you can add that you wouldn't want to see how phonograph records were made if you are a fringe audiophile. The mess would shock you.
 
jan,

Which begs us to wonder why no one has tried seriously tried to identify mechanisms and measurements that identify bad sound. I find distortion and frequency response measurements at probably the bottom of the list.
We have had a few stabs at TIM, PIM...........and Mr.Curl's comments on higher order distortion products.............but to this day no one has put it all in a meaningful manner, and I bet a few artifacts or distortions have not been identified yet.

Till then I suppose we have to trust or ears or Mr.Curl.:D

Regards,

Jam
 

Attachments

  • EAR-CARTOON.jpg
    EAR-CARTOON.jpg
    4.8 KB · Views: 175
  • 010806curl.jpg
    010806curl.jpg
    54.7 KB · Views: 194
Last edited:
My posts are falling behind, but I might pose an interesting scenario for some audiophiles. Let us say that you have a home playback system, composed entirely of vacuum tubes. Of course, that would be impossible for digital, but let's say you have a good vinyl record collection and truly enjoy it. Now, somewhere, at sometime, you found an electronic crossover with IC's in the high frequency thru-path, however it does virtually everything you could ask for in an electronic crossover. Do you think that you could detect a SINGLE IC/channel added in series with your all tube design, while playing an early Joan Baez recording? Let's ask, everyone. ;-)

Even worse, I recently bought a Korneff 76 preamp to mate with my DIY 45 amp, and the best phono preamp I have is a Behringer Ultramatch/Ultracurve combination. I put the RIAA curve into the graphic equalizer and modify it with the parametric equalizer. I have been listening to digitized vinyl, 96/24 quality, and thoroughly enjoying it! My re-equalized Deutsche Gramaphones have never sounded better.

As an aside, I recently nabbed the Vendetta SCP-2b advertised on Audiogon. I don't have it yet but I can hardly wait! Researching that unit led me to this forum. This is my first post. Hello everyone.
 
My opinion is that 16 bits is not enough to cover the normal (Whatever that is) human hearing range. I think it really takes 26 bits.

You won't find much hard evidence to support that.

Also I feel that some folks can hear 5 degrees of phase shift at 20khz as reported by R. Neve and M. Schroeder. I don't know of any audio gear that meets these design goals.

Does this require your head in a vice and micrometer placement of your speakers? That's about 240kHz single pole and there are a lot high-speed op-amps that can do that at 26dB of gain easily.

If you want to repeat Schroeder's experiment you can create .wav files that contain sine waves and harmonics. Be sure to keep the amplitude constant as you create files with different phase shifts. With a good monitor speaker you can see how much phase shift you can perceive.

Phase shifts of what relative to what?
 
Last edited:
Hello Jan.
So if I was to send you two 15 second 96k/24bit wav files that sound subtly different, you can document that difference ?.

Dave.

If they are actually verified to sound different (i.e., differences detected in a properly controlled ears-only test, as opposed to "my wife who was in the next room could hear it!"), then yes he can, I can, you can, anyone can.
 
Here is a paper by Bob Stuart about how much bits and sampling rate we need.
http://www.meridian-audio.com/w_paper/Coding2.PDF

Interesting, though the lossless coding discussion lost me for a while. Lossless is lossless and more computer horsepower will bring the answer closer to the (theoretical?) limit at the speed needed. Supposedly IBM has patents on lossless arithmetic coding that has a huge compression compute load but decompresses fast.

I tried the experiment of taking a .wav and making a low quality mp3 and subtracting the two and losslessly compressing the difference, no surprise the worse the mp3 the bigger the other got and no win.

Something that still baffles me is that a low quality mp3 of 60Hz for 10 seconds and 60Hz minus one cycle in 10 seconds come out perfect, i.e. exactly one cycle was missing in the output.
 
Last edited:
Here is a paper by Bob Stuart about how much bits and sampling rate we need.
http://www.meridian-audio.com/w_paper/Coding2.PDF

I don't know about that.

I found this rather misleading:

An externally hosted image should be here but it was not working when we last tested it.


If taken literally, then yes, the dithering noise at 16 bits would be audible.

But the hearing threshold shown in the plot is based on anechoic conditions after the person has become acclimated and doesn't show the thresholds under normal conditions.

The ambient noise in a typical quiet home is about 40dB SPL. A recording studio about 30dB SPL. I doubt you'd be able to hear that noise under those conditions.

And of course we're not sitting in quiet homes listening to nothing more than dither noise. We're listening to music at levels far greater than ambient and that reduces our sensitivity all the more.

That's why I was asking Ed earlier what the effective dynamic range of our ears is.

Sure, we can hear from the thermal noise limit of the air itself up to over 120dB.

But NOT AT THE SAME TIME.

se
 
My view on it is simple. CD quality is very good but just not good enough to be absolutely perfect. It certainly is good enough to beat my analog rig in terms of speed stability and noise so most of my younger friends that have grown up with CD prefer my CD rig. Analog is still more emotional to me but the reason may simpy be that i have learned to supress it´s flaws and enjoy it´s advantages. It is simply a fascinating exprience to me. We are developping a Server right now and we got really good sound with HD downloads so i hope that this can make the analog crowd happy too. I am still not prepared to sell my records and analog rig. It is part of my life and always will.
 
Analog is still more emotional to me but the reason may simpy be that i have learned to supress it´s flaws and enjoy it´s advantages.

Perhaps its vinyl's flaws that ARE its advantages in terms of what one prefers subjectively.

I've never really bought into the trope that vinyl sounds good in spite of its flaws rather than because of them.

se
 
Status
Not open for further replies.