John Curl's Blowtorch preamplifier part IV

Status
Not open for further replies.
It seems to me impossible to fulfill this requirement as GD differences are a *consequence* of other factors.

Yet, you didn't hesitate to make the statement;

"I dont care of the papers of one or another, I trust my ears, taking great care of Group delay in speakers since decades and I can assure-you it makes a difference.
One thing you can verify by yourself, comparing a digital equaliser with an analog equivalent. Reason why so many sound engineers continue to use big analog mixing desks for music..."

It don't sum up...

//
 
Yet, you didn't hesitate to make the statement;
I do not hesitate to set a correction on an instrument during mixing session neither. Usually, it sound better after ;-)

I said I do believe GD matters and indicate where I supposed-it to affect our perception. i' was yet interested by this question in the 70s concerning speakers design.

Of course, if, to make your mind, you listen with a speaker which is not time aligned and soft like butter in the sun, if you do not focus on the qualities of transents, you'll have few luck to figure-it out.

Anyway, I think that trying to find an explanation on something I have experienced and try to find ways to improvements is more productive than reading papers about something I don't, with a believer attitude.
 
Clarification on the above post - I am talking about 'differential' GD at 20 Hz compared to GD in the pass band above the filter. Absolute GD in the passband is 18 ms
I don't think we are so sensible to group delays at low frequencies. And our speakers are not flat here, anyway. May-be a correlation with the frequencies where we are the most sensible for localisation: low and high medium ? And, may-be the treble when they can increase or decrease a transient peak? All this seems logical.
 
But if the GD is constant between two channels as delivered by the recorded material, how would you know or have a problem? I get that in speakers you may want the drivers time aligned as this impacts on stage depth/width etc, but again, pretty easy to resolve to acceptable standards during the design process.

Re the HPF, my concern was that I’d ruin the bass ‘sound’ I’d the GD at LF was markedly different with the filter switched in. Subjectively I found is does not affect the bass sound badly at all.
 
But if the GD is constant between two channels as delivered by the recorded material, how would you know or have a problem? I get that in speakers you may want the drivers time aligned as this impacts on stage depth/width etc, but again, pretty easy to resolve to acceptable standards during the design process.

Re the HPF, my concern was that I’d ruin the bass ‘sound’ I’d the GD at LF was markedly different with the filter switched in. Subjectively I found is does not affect the bass sound badly at all.

My feeling is that we are not on the same line.
Afair, the discussion that emerged was because Pavel was supposed to hear a difference when a 100Khz preamp filter was switched on and off.
My interpretation was that this difference was detected while listening to two channels, but I have no further details.

Comment made by Scott amongst others was that 100Khz alone is not enough information , because a 100Khz filter already starts effecting phase way down in the audio range, but that FR and GD should also be known.
Mentioned was that according to Blauert&Laws threshold of audibility at 8Khz is a GD of 2msec.
But there are several other reasons thinkable why the difference was detected, like distortion caused by the filter in the audio range because of mediocre components or a semiconductor used as a switch etc, etc,
All in all, it's hard to impossible to draw any conclusion on this experience.

And yes, concerning your rumble filter, at 20Hz GD threshold seems to be much higher at ca.30msec, but at 2Khz, apparently the most sensitive frequency, threshold seems at 1msec.

Hans
 
Ok, good clarification and accepted.

I’d be very surprised if someone could hear a 100 kHz -3dB filter switched in and out assuming that the only difference was the very slight drop off in level in the upper octave between 10 and 20 kHz a d not some other confounding parameter.

Do we know how the speaker behaved when fed HF? Might be IMD for example.
 
Last edited:
My feeling is that we are not on the same line.
Afair, the discussion that emerged was because Pavel was supposed to hear a difference when a 100Khz preamp filter was switched on and off.
My interpretation was that this difference was detected while listening to two channels, but I have no further details.
I take it you are referring to his "two preamp hearing test"? Can you hear a difference between 2 solid state preamps?
 
Comment made by Scott amongst others was that 100Khz alone is not enough information , because a 100Khz filter already starts effecting phase way down in the audio range, but that FR and GD should also be known.
Mentioned was that according to Blauert&Laws threshold of audibility at 8Khz is a GD of 2msec.
But there are several other reasons thinkable why the difference was detected, like distortion caused by the filter in the audio range because of mediocre components or a semiconductor used as a switch etc, etc,
All in all, it's hard to impossible to draw any conclusion on this experience.
Not to forget IM or TIM reduction in the power amp, reason why we near all set this kind of input filter at their inputs.
I could add that, as far as I am concerned, what would have surprised me is that two preamps, even very good in the audio bandwidth ... sound the same. Even the 2 sames with different power supplies. Dynamic behavior is something we do not measure, usually.
 
Last edited:
It's the thread I have linked to that Mark was referring to

Thank you for the link, I didn't know that thread.
Scrolling quickly through the postings I found this one from Pavel without any further detail, confirming what Mark mentioned:

" I built different preamp with similar BW (about 1MHz) which was extremely detailed and revealing. Putting and additional LP filter at the input (100kHz/-3dB) made the sound "smooth", however much less revealing and with less impact. It seems that even small linear changes in the transfer function may be quite audible and after some training detectable even in the ABX."

Can you hear a difference between 2 solid state preamps?

Hans
 
I make no claims about temperature rise. I don't know how to translate from pressure to temperature, and I'm not certain that the guys back in the day really could either. But I'd trust their guesses over my own. My case goes:


Peak (each channel) stylus velocity at NAB 0VU = 5 cm/s. At +20 VU peak acceleration is then about 3km/s/s or about 320 G. Not all recordings are so conservative, and peaks of +30 VU must be expected, or a little more. Lets call it 1000 G's for maximum drama. (But real.)


An excellent, perhaps no longer available, stylus might have an effective moving mass of .25mg - we'll use that as a conservative best case. So the vinyl surface must exert a peak force on the stylus of 2.5 mN.


Assuming a contact area of 25um ^2 (ballpark for modern shapes - no fricking idea how they come up with it) peak pressure at the contact points is 2.5^-3 / 2.5^-5 = 100 Pascals = 689500 psi = 344 tons per sq inch.


My argument goes that we can't apply day-to-day rules of thought to vinyl playback. It's more profound than meets the eye, and more interesting!


Much thanks, as always,
Chris

Nice calc except the stylus is moving not just against the sidewalls but also the length of the groove so the contact area is much larger.

Although there certainly are cases of misadjusted parameters where a single play can damage the record. Usually when the stylus is tracking light and lops off the top of the groove. Other case is tracking too heavy and scores the groove. Both clear cases of misadjustmemt but do show the vulnerabities of vinyl records.

Overview of materials for Vinyl Ester Sheet Molding Compound
 
Last edited:
Status
Not open for further replies.