The issue here syn08 is that you say there is no justification for the 20ms GD figure but offer no data for your counter claim. Did Holman offer any evidence for his claim? He plotted a number of limits in the famous figure 7 and did some listening tests and made a recommendation.
Other than refute his claim and kick up a fuss, what have you done to support your claim? do you have a paper you can point us to? some math?
I see you are using the same tactics as our dearly departed Mr. W on the CFA thing.
Other than refute his claim and kick up a fuss, what have you done to support your claim? do you have a paper you can point us to? some math?
I see you are using the same tactics as our dearly departed Mr. W on the CFA thing.
Last edited:
He is right, there is no evidence. At this point the sensible thing to do is agree with the caveat that you can get low GD therefore why not and park it as it would need a new thread to thrash it out as well as a literature search. Maybe there is something published in the last 50 years that advances the knowledge. Sub 50Hz GD certainly does not keep me awake at night.
Did I claim anything except of "no support data"? Should I prove that Mr. Tomlinson's 20mS criteria is wrong? That's a Jakob(x) type of FUD argument.
Since when can a negative be proven, and where is the burden of proof for any extraordinary claim, clearly colliding with the existing body of knowledge? Hint: Mr. Tomlinson being absent, the burden of proof is on those unconditionally supporting his claim. Do you have the supporting data? No? Too bad.
I'll mention your name next time I'll meet Mr. W, I am sure you would not want to hear about his reaction.
P.S. Nice to hear you now appreciate math as a valid proof. Last times you encountered it (here and in your beloved CFA thread) you royally ignored/dismissed/snubbed all the written math proof.
Last edited:
Whatever group delay variations exist, they'll be a sum of the contributions of the fundamental resonance, the preamp's high-pass filter, and whatever occurred making the record. A fancy-enough DSP could maybe back out of the first two if really important. FM problems excepted.
To me the only important thing is to properly place the F-res and its Q. Measuring this is possible by driving the stylus with a variable frequency motor. I tried once (the proverbial 40 years ago) using a tweeter with its outer suspension cut away lots, which was talked about at the time, but had dangerous magnetic field issues. Don't try this at home kids.
All good fortune,
Chris
To me the only important thing is to properly place the F-res and its Q. Measuring this is possible by driving the stylus with a variable frequency motor. I tried once (the proverbial 40 years ago) using a tweeter with its outer suspension cut away lots, which was talked about at the time, but had dangerous magnetic field issues. Don't try this at home kids.
All good fortune,
Chris
I tried once the same using a piezo speaker I pulled from a dead quartz watch. Don't remember any relevant results.
I’ve asked you for the ‘existing body of knowledge ‘ on GD but you just get angrier and more irrational (eg claiming/imagining things that never took place on the CFA thread) by the minute. You are in FUD overdrive and it’s only Tuesday.
In the absence of any guidance or considered and reflective input on the subject from you I’ll have assume TH’s views to be valid for the time being.
That would certainly solve the magnetic death grip issue, but I'd worry about backlash in the Rube Goldberg parts. Or does that only apply to the old Motorolas? I'll look into it; thanks.
All good fortune,
Chris
Be my guest. Since nobody has proof that alien anal probing didn’t occur, you may want to assume this view as well.
Last edited:
Look at (surprise, surprise) Fig. 7. It clearly shows the “limit of audibility” (labelled 1) and the “German Broadcast standard” (labelled 2) while the text provides a credible reference about, numbered 16a. More than one order of magnitude over TH “recommendation”. That’s pretty much “existing body of knowledge” and can’t do much better than rubbing your nose on it.
Since you seem to be into the authority belief system, may I suggest getting some Bybees. They are recommended by the famous JC and is also colliding with syn08.
Last edited:
Here is some more “existing body of knowledge``.
Quote:
'Back to frequency-dependent delay…
The subjective effect of excessive group delay is a “loosening” of the bass or a “less dry” bass quality. Currently there is insufficient psychoacoustic research on the threshold of group delay at low frequencies. One value is known: 2.5 ms at 100 Hz ....'
end quote
This is a factor 10 lower than the curves in the TH paper show. If we take a factor of 10 lower, then the 20ms requirement is not an arbitrary number anymore.
I found this on a web page of the Neumann company (yes, the Neumann company beeing known for their studio microphones and monitoring speakers). The article is about group delay at low frequencies for studio monitor speaker systems.
Here the link: Georg Neumann GmbH - Professional Monitoring
Assuming these guys are knowledgable, this is confirming my suspicion that the 'audible limit' and 'german post and broadcast standards' by far are not valid for HiFi reproduction. These standards have been established in the 50's for acceptable limits for telephone landlines which they named 'audible limit' (at that time the german post was a federal agency and they were the 'gods' setting telecommunication transmission standards). The broadcast people (broadcasting at that time was also a federal affair - no private stations) simply had to accept this - live transmissions in the 50`s and 60's from a stadium or a concert hall had to use landlines to transmit signals to the broadcast station.
The word HiFi was not even invented. The standard curves all end at 50 Hz - simple reason: no signals below 50 Hz were ever transmitted. The later german DIN standard for HiFi systems required a bandwith of 50 Hz to 12.5 kHz with +/-3dB variation allowed. These were the limits the upcoming FM broadcasting just could garantee. Such a system by today's standard would be named 'crap'.
Quote:
'Back to frequency-dependent delay…
The subjective effect of excessive group delay is a “loosening” of the bass or a “less dry” bass quality. Currently there is insufficient psychoacoustic research on the threshold of group delay at low frequencies. One value is known: 2.5 ms at 100 Hz ....'
end quote
This is a factor 10 lower than the curves in the TH paper show. If we take a factor of 10 lower, then the 20ms requirement is not an arbitrary number anymore.
I found this on a web page of the Neumann company (yes, the Neumann company beeing known for their studio microphones and monitoring speakers). The article is about group delay at low frequencies for studio monitor speaker systems.
Here the link: Georg Neumann GmbH - Professional Monitoring
Assuming these guys are knowledgable, this is confirming my suspicion that the 'audible limit' and 'german post and broadcast standards' by far are not valid for HiFi reproduction. These standards have been established in the 50's for acceptable limits for telephone landlines which they named 'audible limit' (at that time the german post was a federal agency and they were the 'gods' setting telecommunication transmission standards). The broadcast people (broadcasting at that time was also a federal affair - no private stations) simply had to accept this - live transmissions in the 50`s and 60's from a stadium or a concert hall had to use landlines to transmit signals to the broadcast station.
The word HiFi was not even invented. The standard curves all end at 50 Hz - simple reason: no signals below 50 Hz were ever transmitted. The later german DIN standard for HiFi systems required a bandwith of 50 Hz to 12.5 kHz with +/-3dB variation allowed. These were the limits the upcoming FM broadcasting just could garantee. Such a system by today's standard would be named 'crap'.
Last edited:
You are right
we should stop this in my view useless discussion as nobody here wants to build excessive filtering into the MC headamp. Setting up TT, tonearm and cart correctly is by far the best measure to avoid infrasonic signals leading to modulations of the wanted signals.
And as already mentioned several times: no filter can remove the distortions if they already are produced in the cartridge itself.
If somebody is really interested in digging into the GD field for low frequency he should start a new thread. Available program material containing real music notes below 30 Hz is in the promille range. And reproduction systems going that low in normal living rooms are also rare
we should stop this in my view useless discussion as nobody here wants to build excessive filtering into the MC headamp. Setting up TT, tonearm and cart correctly is by far the best measure to avoid infrasonic signals leading to modulations of the wanted signals.
And as already mentioned several times: no filter can remove the distortions if they already are produced in the cartridge itself.
If somebody is really interested in digging into the GD field for low frequency he should start a new thread. Available program material containing real music notes below 30 Hz is in the promille range. And reproduction systems going that low in normal living rooms are also rare
Last edited:
I fully agree that it is a non issue and should not be discussed in this thread to such a length.
If you want to stop your membranes from flapping and prevent IM modulation because of that, use a rumble filter, period.
In all cases a 3rd order is good enough and that's all there is to know.
When you have full range Electrostatic speakers, you won't even need a rumble filter.
Hans
Speakers and turntables, apples and oranges. But enough of this, agreed. Bonsai is right and I am wrong, ok?
bill, I should caution that this is appears to be the lesser Leach variant from #311 rather than true Duraglit. But it's a simple (?!!) matter to remove 6 bits and change 2 resistors to get the, at least 3dB, Vni improvement of true Duraglit.It's likely PSR is also improved but I would want to confirm this in real life. There is no evidence that lesser variants give the advantages for World Peace and better sex life of true Duraglit.
My apologies. syn08's variant is in #375 not #311.
syn08, are you referring to yet another variant?. #375 shows an "adapted current mirrored version of the Leach". It is certainly possible to simplify & improve that by moving to true Duraglit ... especially with a real life MC cartridge
Your lesser version is the likely reason you have not experienced improvements in World Peace and a better sex life.
at the risk of generalising ... the audible effect of mistracking is very different from amp clipping. Those of us who played with the original Shure Obstacle Course records will have some experience of this.
________________
I've done a lot of work on the audible effect of Group Delay.
AES E-Library >> Is Linear Phase Worthwhile?
Surprisingly, 40yrs on, there is only one new item of data to add to my references ...
AES E-Library >> On the Audibility of Midrange Phase Distortion in Audio Systems
Double Blind Listening Tests on stuff like this is often a 2 stage process. First you determine if the effect is reliably audible .. with loadsa caveats about the conditions, the people who can truly tell the difference vs da wannabe Golden Pinnae bla bla
ONLY if the answer is 'yes' to this first question, do you ask the second, which is "Which do you prefer?". The answer isn't always expected.
There was a lot more work on 'linear phase' which isn't in my paper but (generalising) loadsa group delay is often preferred by those who can tell the difference.
However, as a speaker man, the audible benefits of vinyl playback HP filtering are certainly real for da Xmax issues. I would ALWAYS build in drastic HP filters for vinyl playback.
In da early 70s, the Radford Transmission Lines were the first domestic speakers with true flat response below 30Hz. ALL their amps from that period had 40Hz HP filters which couldn't be switched off
cos customers complained about dishes rattling on warps.
My apologies. syn08's variant is in #375 not #311.
syn08, are you referring to yet another variant?. #375 shows an "adapted current mirrored version of the Leach". It is certainly possible to simplify & improve that by moving to true Duraglit ... especially with a real life MC cartridge
Your lesser version is the likely reason you have not experienced improvements in World Peace and a better sex life.
Jeeeeezzzz Richard, can't you just LOOK at the #375 schematic? The right side transistors have absolutely nothing to do with noise. They are there only for the purpose of defining the left side (gain) complementary devices DC bias. The right side Q101/Q103 could be the noisiest transistors known to mankind, their overall noise contribution will still be zero. The main advantage of this approach is that matching (for beta, Vbe) the low noise complementary devices is not required, also the bias is beta independent. The only extra noise contribution is half of the left side pair emitter resistor, that is 0.5ohm. I think we can agree this is a negligible noise price to pay.
Last edited: