What you cannot see is that there is a 1-2 dB resonance peak. The frequency response rolls off in a controlled manner, i.e 6 dB/octave in the lateral plane. Instead, the roll-off is steady and controlled. The lateral f-res has a vanishingly low Q (low amplitude) and is virtually non-existent (on my own Ortofon Jubilee/red armtube the rise in amplitude at the resonant frequency is 1-2 db). This means that the bass response is phase-coherent and linear.
The second graph shows an approximate 5 dB rise, but in terms of bass performance tis should be discarded: there is no vertical information cut into the record at that freqency.
I would imagine that a phase-coherent, linear response down to 19 Hz is a good thing.
And, of course, our whole discussion began as a debate about Korf’s interesting claim that all cartridges behave the same regardless of their compliance.
The second graph shows an approximate 5 dB rise, but in terms of bass performance tis should be discarded: there is no vertical information cut into the record at that freqency.
I would imagine that a phase-coherent, linear response down to 19 Hz is a good thing.
And, of course, our whole discussion began as a debate about Korf’s interesting claim that all cartridges behave the same regardless of their compliance.
Last edited:
I have to clarify two basic facts first.
Resonant frequency ALWAYS exists. The question is at what frequency it occurs.
The ideal resonant frequency should be outside of the music range, i.e., 20 Hz - 20 kHz, ideally around 9-11 Hz.
I don't know what method that guy used to test his tonearm's resonant frequency. I suspect that his tonearm resonant frequency is within 20 Hz- 20 kHz. The input signal is already at about 90 dbs and the resonant frequency is hiding under the normal input signal. In other words, The design of his tonearms is flawed.
I am not an expert on cutting vinyl, but I highly suspect the correctness of your statement. In my resonant frequency tests, I used Cardas Frequency Sweep record. One is 65.4 Hz lateral modulation and another is 65.4 Hz vertical modulation. If you are correct that there is no difference between lateral modulation and vertical modulation under 125 Hz, why did Stan Ricker, who cut the Cardas Test LP, need to cut two different tracks for different modulations?
I would imagine that a phase-coherent, linear response down to 19 Hz is a good thing.
I am not sure what you meant by phase-coherent, linear response while we are talking about resonant frequency. 19 Hz is too close to 20 Hz, the music range, and is not desirable. From the diagram, I really don't think he could conclude that his tonearm resonant frequency was 19 Hz.
Last edited:
@super10018
Indeed, the f-res always exists. In the case I quote it is at 19 Hz. (f.res is the frequency at which the signal begins to attenuate).
If the resonant frequency has a low Q (ie, it does not rise in amplitude) it means that the system’s overall frequency response is linear.
When I said ‘phase-coherent linear response down to 19Hz’ I referred to the overall linearity of the entire reproduced frequency range: I think we both agree that it must be a good thing that the 19Hz signal is at the same amplitude as the 1 kHz signal?
The measurements were done in the usual way: a low-frequency sweep plodded with amplitude as the y-axis and frequency as the x-axis.
The measurements were (as far as I can see from the graphics) made with Feickert’s Adjust+. (My own measurements with the same system, are in broad alignment with the results I reproduced)
The frequency at which low frequency signals are added together is decided by the mastering engineer. It varies from case to case and is also applied gradually, much like the hi-blend filters that reduce hiss in FMtuners.
Stan Rickert, I am sure, did his vertical modulation because he could do it. What purpose it serves, I cannot say.
I did not say that there is no difference between lateral and vertical modulation inder 125 Hz, I said that there is no (i.e. very little) vertical modulation at low frequencies.
Indeed, the f-res always exists. In the case I quote it is at 19 Hz. (f.res is the frequency at which the signal begins to attenuate).
If the resonant frequency has a low Q (ie, it does not rise in amplitude) it means that the system’s overall frequency response is linear.
When I said ‘phase-coherent linear response down to 19Hz’ I referred to the overall linearity of the entire reproduced frequency range: I think we both agree that it must be a good thing that the 19Hz signal is at the same amplitude as the 1 kHz signal?
The measurements were done in the usual way: a low-frequency sweep plodded with amplitude as the y-axis and frequency as the x-axis.
The measurements were (as far as I can see from the graphics) made with Feickert’s Adjust+. (My own measurements with the same system, are in broad alignment with the results I reproduced)
The frequency at which low frequency signals are added together is decided by the mastering engineer. It varies from case to case and is also applied gradually, much like the hi-blend filters that reduce hiss in FMtuners.
Stan Rickert, I am sure, did his vertical modulation because he could do it. What purpose it serves, I cannot say.
I did not say that there is no difference between lateral and vertical modulation inder 125 Hz, I said that there is no (i.e. very little) vertical modulation at low frequencies.
Although I have many disagreements with your reply, I will focus on the main topics.
Here are the results of my tests using two different pink noise sweeps in lateral and vertical modulations. You can see the overall frequency response, in the meantime, you can see resonant frequencies clearly.
A good tonearm design doesn't try to smooth out the resonant frequency, but to keep it within a certain range. In my opinion, at least, it should be lower than 11 Hz.
No. It is not correct. If the resonant frequency has a Q so low the overall frequency response is linear. It means there is no resonant frequency. It also means the tonearm design is flawed.
I don't understand why he used Adjust+ to test resonant frequency. All he needs is an oscilloscope or software such as Adobe Audition with a test LP. The frequency response generated by Adjust+ software is too smooth to see the truth. Such a smooth frequency response for a cartridge isn't possible. I don't know what the purpose of such frequency response in Adjust+ is. There are many ways to test resonant frequency. I don't think Adjust+ is one of them.
Here are the results of my tests using two different pink noise sweeps in lateral and vertical modulations. You can see the overall frequency response, in the meantime, you can see resonant frequencies clearly.
A good tonearm design doesn't try to smooth out the resonant frequency, but to keep it within a certain range. In my opinion, at least, it should be lower than 11 Hz.
Last edited:
No. I am just questioning his methodology. I didn't draw any conclusion about his tonearm yet.
Let's take a look at his analysis.
He did a resonant frequency test for his DP-6. He posted his results as follows.
The result is completely consistent with my tests. The resonant frequencies were about 12-13 Hz for lateral and vertical modulation with damping in both planes. By the way, I disabled the damping device when I did my resonant frequency tests.
However, he switched to different software to do the same test for his DP-8.
He stated that Typical DP-8 lateral resonance. Be aware that at very low frequencies where the modulation is horizontal, there is no fundamental resonance. The amplitude just rolls off downwards.
I don't understand why he switched to different software. Isn't the method for testing DP-6 good enough? Then, he stated that there is no fundamental resonance. Why? Is it because of heavy damping on DP-8? The amplitude of the resonant frequency had been completely wiped out by the damping. I really doubt his conclusion and method. I understand that damping may reduce the amplitude of the resonant frequency. Can damping completely eliminate resonant frequency? I don't know, but his method is questionable.
Let's assume that heavy damping can eliminate resonant frequency completely. But what are the effects on playing back? In my experience, heavy damping is not desirable.
Furthermore, he thought the vertical resonant frequency was 19 Hz based on his test result. I disagree. I think there is not enough evidence to support his claim by looking at the chart generated in Adjust+.
In any case, if you choose to withdraw from this discussion, it is fine with me.
Let me start by correcting you: you wrote 'I didn't draw any conclusion about his tonearm yet’.
But you have already twice concluded that Mørch’s DP-8 design is flawed: you wrote 'In other words, The design of his tonearms is flawed.’(yesterday 11:10) and 'If the resonant frequency has a Q so low the overall frequency response is linear. It means there is no resonant frequency. It also means the tonearm design is flawed.’(yesterday 16:09)
In regard to the more substantial matters:
You are comparing apples to pears.
The DP-6 measurements (made with 1970s Brüel & Kjær equipment) show Left and Right channels subsonic frequency response. They show the groove content at +45°/-45° . These frequencies reflect what an unidentified ‘typical’ cartridge picked up in 1982 when the DP-6 came on the market.
The DP-8 measurements (made with Adjust+) are of L-R (vertical) and L+R (lateral) subsonic frequency response. This reflects in-phase and out-of-phase subsonic frequency response. The Adjust+ software calculates the f-res automatically based on where the largest amplitude is located.
The two measurements show different parameters and cannot be compared.
The amplitude of the input signal is relative (Feickert is is to blame for this: he called the y-axis ‘dB', but in fact it is a percentage. 100% is the maximum level available in the program and depends on the soundcard. The 100% level is set manually before measurements are conducted).
You speculate that the arm is too heavily damped: The DP-6 and the DP-8 are identical in their bearings (and, yes there is some damping in the lateral bearing). If the lack of a pronounced f-res in the DP-8 is to blame for this (desirable) feature, it should also be present in the DP-6.
Why did Mørch change methodology? You imply some sinister motive, but the answer is that better and less expensive soft- and hardware allowed it. The Brüel & Kjær equipment used cost more than 100,000 Danish Kroner in 1978 when the arm was first prototyped and the equipment could only be accessed in two locations in Copenhagen. The Feickert Adjust+ cost €250 when it was last available and runs on (almost) any PC.
However, there are also sound engineering reasons for the change: low frequencies are cut laterally, and thus the lateral and vertical frequency response tell us more about the arm’s behaviour and stability than R/L measurements.
But you have already twice concluded that Mørch’s DP-8 design is flawed: you wrote 'In other words, The design of his tonearms is flawed.’(yesterday 11:10) and 'If the resonant frequency has a Q so low the overall frequency response is linear. It means there is no resonant frequency. It also means the tonearm design is flawed.’(yesterday 16:09)
In regard to the more substantial matters:
You are comparing apples to pears.
The DP-6 measurements (made with 1970s Brüel & Kjær equipment) show Left and Right channels subsonic frequency response. They show the groove content at +45°/-45° . These frequencies reflect what an unidentified ‘typical’ cartridge picked up in 1982 when the DP-6 came on the market.
The DP-8 measurements (made with Adjust+) are of L-R (vertical) and L+R (lateral) subsonic frequency response. This reflects in-phase and out-of-phase subsonic frequency response. The Adjust+ software calculates the f-res automatically based on where the largest amplitude is located.
The two measurements show different parameters and cannot be compared.
The amplitude of the input signal is relative (Feickert is is to blame for this: he called the y-axis ‘dB', but in fact it is a percentage. 100% is the maximum level available in the program and depends on the soundcard. The 100% level is set manually before measurements are conducted).
You speculate that the arm is too heavily damped: The DP-6 and the DP-8 are identical in their bearings (and, yes there is some damping in the lateral bearing). If the lack of a pronounced f-res in the DP-8 is to blame for this (desirable) feature, it should also be present in the DP-6.
Why did Mørch change methodology? You imply some sinister motive, but the answer is that better and less expensive soft- and hardware allowed it. The Brüel & Kjær equipment used cost more than 100,000 Danish Kroner in 1978 when the arm was first prototyped and the equipment could only be accessed in two locations in Copenhagen. The Feickert Adjust+ cost €250 when it was last available and runs on (almost) any PC.
However, there are also sound engineering reasons for the change: low frequencies are cut laterally, and thus the lateral and vertical frequency response tell us more about the arm’s behaviour and stability than R/L measurements.
Last edited:
Let's cut to the chase. I want to ask you two questions.
1. For his DP-8, he stated that there is no fundamental resonance anymore. Why? Is it because of damping? Where is the resonant frequency? We all agree that resonant frequency exists no matter what you do.
2. So, for his DP-8, its lateral resonant frequency is about 27 Hz and its vertical resonant frequency is about 19 Hz. However, the consensus for resonant frequency is ideally between 9-11 Hz. Don't you think it is problematic with such high resonant frequencies?
1. For his DP-8, he stated that there is no fundamental resonance anymore. Why? Is it because of damping? Where is the resonant frequency? We all agree that resonant frequency exists no matter what you do.
2. So, for his DP-8, its lateral resonant frequency is about 27 Hz and its vertical resonant frequency is about 19 Hz. However, the consensus for resonant frequency is ideally between 9-11 Hz. Don't you think it is problematic with such high resonant frequencies?
RE: 1.: We should probably blame that on the fact that Hans Henrik’s native language was not English or the statement was not properly proof-read. I think it should have read: ' Be aware that at very low frequencies where the modulation is horizontal, there is no fundamental resonance peak.’
I must admit that I cannot adequately explain where the f-res has disappeared to (and I cannot ask Hans Henrik who died two or three weeks ago… he was 92). But I would suggest that what we may find is that it is hiding in the area of 2-3 Hz and thus does not show up in traditional measurements. (you will not find any test records that go so low. You could try to measure those frequencies with a stationary transducer, but that then means that you are measuring the static behaviour of the system and that therefore we are back to square one). and of course, we have no way of knowing what kind of filter is applied in computer sound-cards or if the system feeds into an amplifier that follows the recommendations of the 1973 RIAA curve (6 dB/octave cut-off from 50 Hz). What is certain is that no domestic amplifier is linear down to DC.
Re: 2.: I do not understand how you arrive at 27 Hz as the lateral resonant frequency. I can see that there is a green line at that frequency, but this is an artefact of the Feickert Adjust+. As I said above it finds the frequency where there is the highest amplitude and marks that as the f-res. This works well when you have a ‘normal’ arm, but not in such a linear system as the DP-8 where the an increase in amplitude is in the region of 1-2 db.
The recommendation of f-res in the region 9-11 Hz is supported by the work of Happ & Karlov (1973) and Rother (1977) but dates from a much earlier time. Personally, I do not bother with those recommendations when speaking of the DP-8. The lateral f-res has virtually no amplitude and the vertical f-res is in a frequency range where there is no signal.
In my own system (DP-8 and Ortofon Jubilee) the vertical f-res is at 16 Hz with a 5-6 dB rise in amplitude. In practical terms, that means that at 20 Hz the resonance is absent. so, no problems there.
I must admit that I cannot adequately explain where the f-res has disappeared to (and I cannot ask Hans Henrik who died two or three weeks ago… he was 92). But I would suggest that what we may find is that it is hiding in the area of 2-3 Hz and thus does not show up in traditional measurements. (you will not find any test records that go so low. You could try to measure those frequencies with a stationary transducer, but that then means that you are measuring the static behaviour of the system and that therefore we are back to square one). and of course, we have no way of knowing what kind of filter is applied in computer sound-cards or if the system feeds into an amplifier that follows the recommendations of the 1973 RIAA curve (6 dB/octave cut-off from 50 Hz). What is certain is that no domestic amplifier is linear down to DC.
Re: 2.: I do not understand how you arrive at 27 Hz as the lateral resonant frequency. I can see that there is a green line at that frequency, but this is an artefact of the Feickert Adjust+. As I said above it finds the frequency where there is the highest amplitude and marks that as the f-res. This works well when you have a ‘normal’ arm, but not in such a linear system as the DP-8 where the an increase in amplitude is in the region of 1-2 db.
The recommendation of f-res in the region 9-11 Hz is supported by the work of Happ & Karlov (1973) and Rother (1977) but dates from a much earlier time. Personally, I do not bother with those recommendations when speaking of the DP-8. The lateral f-res has virtually no amplitude and the vertical f-res is in a frequency range where there is no signal.
In my own system (DP-8 and Ortofon Jubilee) the vertical f-res is at 16 Hz with a 5-6 dB rise in amplitude. In practical terms, that means that at 20 Hz the resonance is absent. so, no problems there.
Last edited:
Well, if it is not there, it means it is not there no matter how you flap your gum. Again, there is no resonant frequency in the 2-3 Hz area at all. If there is no resonant frequency, there are two possible reasons. 1) The testing method is not correct. 2) The design of the tonearm is flawed. It is as simple as that.
As you said, Adjust+ will find the peak frequency automatically. 27 Hz is the peak frequency for the lateral resonant frequency. The amplitude of the resonant frequency is always the peak frequency because it is input signal + resonant signal=resonant frequency. So, it is always the highest amplitude.
I would like to summarize my view again. The producer of DP-8 used an incorrect method to test the resonant frequency of DP-8 tonearm. In addition, he explained the results incorrectly.
What worries me about your underlying assumption is that you reject the measurements which are conducted with a method that produces the expected shape of f-res with the DP-6 out of hand when it is applied to the DP-8. The outcome of the measurement of the lateral frequency response of the DP-8 is different, but not flawed. But the whole idea is precisely that the DP-8 produces a linear frequency response in the bass.
So, let us just agree to disagree: our major disagreement is about the interpretation of the measurement that Mørch presented and the definitions of f-res.
You believe that all cartridges must have a notable rise in amplitude around the resonant frequency, and I believe that the fundamental resonance is defined by a well-defined roll-off below the f-res . In your view the 1 dB rise in frequency around 27 Hz (which is similar to a flexing mode and not present in all Mørch’s published measurements) is the fundamental resonance. Feickert’s program displays a green line at 32 Hz for the Dynavector DV17 and 20 Hz for the B&O MMC1 and on that basis I conclude that the green line is a red herring (sorry about that metaphor).
In my view the beginning of the well-defined roll-off of 12 dB/octave visible in the measurements defines the f-res.
Each to his own….But is it makes you any happier, my own measurements of the Ortofon Royal N in a blue arm-tube shows a 3 dB rise at 12 Hz lateral and a perfect +/-0.5 dB deviation 20-40 Hz (in other words a very linear frequency response in the bass).
So, let us just agree to disagree: our major disagreement is about the interpretation of the measurement that Mørch presented and the definitions of f-res.
You believe that all cartridges must have a notable rise in amplitude around the resonant frequency, and I believe that the fundamental resonance is defined by a well-defined roll-off below the f-res . In your view the 1 dB rise in frequency around 27 Hz (which is similar to a flexing mode and not present in all Mørch’s published measurements) is the fundamental resonance. Feickert’s program displays a green line at 32 Hz for the Dynavector DV17 and 20 Hz for the B&O MMC1 and on that basis I conclude that the green line is a red herring (sorry about that metaphor).
In my view the beginning of the well-defined roll-off of 12 dB/octave visible in the measurements defines the f-res.
Each to his own….But is it makes you any happier, my own measurements of the Ortofon Royal N in a blue arm-tube shows a 3 dB rise at 12 Hz lateral and a perfect +/-0.5 dB deviation 20-40 Hz (in other words a very linear frequency response in the bass).
It struck me that there is a simple reason why f-res does not show in the graph: inertia (nb: Inertia is not the same as damping). Let’s consider an analogy: a child's garden swing with a light plastic seat. A gust of wind will set off the swing at the f-res. But if you place a block of concrete on the seat, the same gust of wind will not make the seat move.
In the case of the DP-8 we have the concrete-block scenario: the lateral signal does not contain enough energy to make the high lateral mass of the arm move (resonate). So what we are seeing in the graph is the pure electrical output of the cartridge. Since the RIAA curve specifies a roll-off of 6 dB/octave, that is what we see. (The resonance that I reported in the case of the Ortofon Royal was caused by a misalignment in the centre bearing in my Oracle record player. Going through my files, I see that subsequent measurements with the DP-8 mounted on the Technics SP10 showed a 6 dB/octave roll-off. The Oracle turntable has since been serviced by Oracle in Canada and no longer shows this resonance peak).
So forget all my waffle about 20 Hz as the f-res: the subsonic response observed by Mørch/Feickert is a function of the RIAA curve.
You can read more about the subsonic frequencies known colloquially as Wow and Flutter and their interaction with measurements in Ladegaard's AES paper, pp. 6-9.
Keine Hexerei.
In the case of the DP-8 we have the concrete-block scenario: the lateral signal does not contain enough energy to make the high lateral mass of the arm move (resonate). So what we are seeing in the graph is the pure electrical output of the cartridge. Since the RIAA curve specifies a roll-off of 6 dB/octave, that is what we see. (The resonance that I reported in the case of the Ortofon Royal was caused by a misalignment in the centre bearing in my Oracle record player. Going through my files, I see that subsequent measurements with the DP-8 mounted on the Technics SP10 showed a 6 dB/octave roll-off. The Oracle turntable has since been serviced by Oracle in Canada and no longer shows this resonance peak).
So forget all my waffle about 20 Hz as the f-res: the subsonic response observed by Mørch/Feickert is a function of the RIAA curve.
You can read more about the subsonic frequencies known colloquially as Wow and Flutter and their interaction with measurements in Ladegaard's AES paper, pp. 6-9.
Keine Hexerei.
Last edited:
I don't care what method he used. I only want to see if there is a peak under 20 Hz. If I don't see it, something is wrong. For the purpose of determining the resonant frequency of a particular tonearm, I don't care if there is a difference between lateral and vertical resonant frequency.
Again, I don't care if the frequency response is linear or not. I only care for the peak and its location within the overall frequency range. The frequency response is irrelevant here. The resonant frequency of a tonearm MUST be under 20 Hz so the tonearm can be qualified as a good design. Otherwise, it is flawed. I keep saying I want to see the peak. But you kept saying that its frequency response is linear. It doesn't matter here at least, for our discussion.
Let's assume your test is 100% correct. This is another piece of evidence that the method used by Mørch is problematic. Again and again, frequency response is irrelevant here.
Last edited:
The DP-8 is simly an arm with different effective mass in the vertical and lateral dimensions. It does so by heavy mass very close to the pivot point for vertical motion and a heavy lateral mass as well. So resonance may well be 16 Hz vertically while being much lower horizontally. This follows the music info in LPs (mono bass etc) and the out of music error signals (0.55 Hz horizontally 4-10 Hz vertically).