U67 de-emphasis network

[Paul W]

I don't see a straw man at all... And I don't think the (you) speaking for the (we) thing is valuable at this point. Just speak for your self as I dont think you agree with one another 100% anyway.

I wasn't saying we agree on everything. I was only saying that we agree about the fact that there should be two different mechanisms of low-pass filtering at work: a relatively minor effect of inhibiting the motion of the diaphragm, and the relatively major effect of electrical cancellation.

You straw-manned that position as denying that the diaphragm would move like a speaker at all, and argued against a position none of us held, several times, despite our clarifying repeatedly that yes, we agree that the capsule will act as an inverse transducer to some extent.

That's why I resorted to all caps to get you to stop it. It was getting really tiresome.

for instance do you %100 agree with the statement below.?

I never claimed that we all agree on everything. That was never the point, either, as I think should have been pretty clear from what I actually wrote.

(Kingkorg's response sounds pretty good to me, now that you ask. But that's a different claim than I was defending.)

 

[GeorgeToledo]

You are either not reading what i'm posting, or you have strong convictions that are difficult to change even with posted measurements.

The improvement in THD achieved by using NFB is negligible. Even in extreme conditions it is just 0.5% (@20K) of improvement in u87. And even then THD at LF is dominated by THD caused by the transformer. Even at lower levels. Neumann chose to use NFB instead of simple filter is because why add THD (by using simple filter) when you can reduce THD by using NFB.

To anyone reading this, If you are a kind of person that hears 0,5% THD diference at 20Khz, while at the same time 2nd harmonic rises at those very frequencies, i apploud you. In that case you have to buy the real deal because there's no clone that will ever fulfill all the expectations. That is also if you record at above 130db which is what we are talking about here. And yes, you are probably a bat.

My statement there comes straigt from testing result where i record same exact source with NFB engaged, and disengaged on a u87. I then apply Eq curve that is measured by REW by injecting the signal and exporting that curve into Voxengo Curve EQ on disengaged NFB file. The two files cancel out in null test. I admit i didn't test this with a u67, but i don't see why i would get different result.

 

[kingkorg]

I haven't tested this with wind, nor plosives. I am honestly interested in real world applications, and i have pointed that i have no idea what might happen in extreme conditions. I still don't believe, whatever that article states, the amount of the level in the feedback can counter the plosive coming into the diaphragm.

I managed to put about 22Vpp out of a k67 polarized at 40V. There's not enough juice in NFB to counter that. NFB is not a multi band compressor IMHO.

After all this thug of war thing between us i don’t think anyone is managing to persuade anyone. Maybe we just move on until someone comes up with a legit test to put this to bed.

I see the logic in your proposals, i just can't see it in my tests, and TBH in my world, and how i use mics i don’t care. In best case scenario, and how I view this it could work as low band compressor + what @soliloqueen said happened in her test.

I really don't have anything else to add.

 

[OPR]

My statement there comes straigt from testing result where i record same exact source with NFB engaged, and disengaged on a u87. I then apply Eq curve that is measured by REW by injecting the signal and exporting that curve into Voxengo Curve EQ on disengaged NFB file. The two files cancel out in null test. I admit i didn't test this with a u67, but i don't see why i would get different result.

Interesting that you got full cancellation of the signal recording two separate passes of the same material. I cant even get full cancellation doing that with the same mic let alone one with circuit alteration.
maybe I'm missing something in your method?

 

[kingkorg]

Interesting that you got full cancellation of the signal recording two separate passes of the same material. I cant even get full cancellation doing that with the same mic let alone one with circuit alteration.
maybe I'm missing something in your method?

It's my mic measurement jig i use for measurements. I have to swap between reference and DUT mics and they have to be perfectly aligned for sub 1db measurements.

 

[OPR]

Thanks for posting that George. For some reason I couldn't quote your post. Yes would seem imply a reduction in excessive diaphragm excursions also.

 

[OPR]

It's my mic measurement jig i use for measurements. I have to swap between reference and DUT mics and they have to be perfectly aligned for sub 1db measurements.

Anechoic chamber?
So you're able to record sound picked up by a microphones capsule on two separate passes that null to complete silence? like a digital copy of a wav file?

Was the source file music? a sweep?

 

[soliloqueen]

Changing voltage to the capsule changes the amount of distortion the capsule generates.

As a side note to the discussion of stiffness/tension, in the world of capsules, this is what makes the pvc M7 different from mylar capsules. PVC has self rigidity. I believe the earliest styroflex CK12s (which apparently sucked) would have also had self rigidity, given styroflex that I have encountered.

They do suck. So do tons of early m7s. I was for a while in possession of like m7 number 12 or something. From a collection. That capsule blew chunks. The holes were like 20% too shallow. There was no saving it. Someone here encountered a similar one

 

[tomas.borgstrom]

They do suck. So do tons of early m7s. I was for a while in possession of like m7 number 12 or something. From a collection. That capsule blew chunks. The holes were like 20% too shallow. There was no saving it. Someone here encountered a similar one

I have also had a very early M7 that was very bass shy even after restoration.

By the way. I like U67 an m269 better with S2 lifted. It sounds more choked in the low end with that circuit.

 

[GeorgeToledo]

Thanks for posting that George. For some reason I couldn't quote your post. Yes would seem imply a reduction in excessive diaphragm excursions also.

Oh, this is from Gotham/Neumann’s original catalog literature, by the way.

 

[GeorgeToledo]

They do suck. So do tons of early m7s. I was for a while in possession of like m7 number 12 or something. From a collection. That capsule blew chunks. The holes were like 20% too shallow. There was no saving it. Someone here encountered a similar one

Neumann, the company, was still trying to figure out how to properly implement what was a Braunmuhl-Weber invention and design.

George Neumann had more to do with the early stuff, like the M1 capsule, and I think that one had most prior work already done as well.

It’s a bit of a sidetrack to the topic of this thread, but worth mentioning in that Neumann was much like DIY-ers here.

 

[GeorgeToledo]

Anechoic chamber?
So you're able to record sound picked up by a microphones capsule on two separate passes that null to complete silence? like a digital copy of a wav file?

Was the source file music? a sweep?

I was going to let that one pass because it seemed so dubious, but also, are we talking full null? Or, a 6dB null or something like that.

It is worth remembering that these methods of analysis are crude considering phenomena that happens in actual usage. To fully observe the plosive issue being discussed the diaphragm excursion has to be very close to hitting the backplate if not actually hitting it. Then, does S2 keep the membrane from hitting the backplate and/or reduce the time the diaphragm spends towards the backplate direction before reaching equilibrium.

I don’t think a microphone frequency response test is testing the issue at hand.

ALSO, the testing approach has to be capable of accurately analyzing low frequencies in the first place.

 

[ZHOUJUN]

I have a pair of Neummann PVC M7s and they both have some cracking, reduced output gain, cracking on the back, and I have seen very nice looking PVC M7s but no loss of low end at all.

 

[soliloqueen]

Thanks for posting that George. For some reason I couldn't quote your post. Yes would seem imply a reduction in excessive diaphragm excursions also.

to me this suggests more simply that it makes sure that excessive diaphragm excursions can't reach the preamplifier, which is true.

 

[Paul W]

Do you "All"(Korg, Khron, Paul) also believe that this little electrostatic speaker effect that you agree causes mechanical movement of the diaphragm opposing the input signal, has zero measurable or audible effect on the sound besides purely passing though electrically?

There's an ambiguity here that may be causing confusion.

When you ask whether "the difference" would be audible, you need to make it clear what you're holding constant.

For example, if you could magically make the capsule act as a transducer in the normal direction but not the other (like a speaker), but hold the amount of electrical feedback constant, the difference might be audible, because the speaker-like driving of the diaphragm in opposition to the LF signal is part of the high-pass effect you normally hear.

On the other hand, if you instead hold the degree of LF rolloff constant, by increasing the feedback enough to get the same total amount of rolloff just by electrical cancellation, you might not hear a difference. (Or similarly, if you make up for the lower degree of rolloff by rolling off a bit more LF in post, it may make little or no difference how you accomplish that same degree of rolloff, as I think is kingkorg's point, at least as long as you're in the regime where the diaphragm and the amp are behaving pretty linearly.)

To approximate the first case, I guess you could put a plain ceramic capacitor in place of the capsule and inject a signal there, and measure how much of the LF rolloff is just due to normal electrical cancellation, without the diaphragm wiggling.

You could also measure the amount that the diaphragm wiggles, when the capsule IS in the circuit, by injecting a signal there and carefully measuring the volume of the resulting LF emitted like a speaker. (Not just noticing that it does in fact emit sound.)

 

[Paul W]

I think that assertion is frequency-dependent. At 25kHz, the condenser is likely more efficient. I don't think this is a simple mechanical equation. Frequency, impulse response, and other factors (i.e. polarizing voltage impacts output level, and therefore "efficiency") all change how efficient the capsule is. I think that's part of what OPR is getting at - the capsule itself is not a linear device, and therefore, the effects of negative feedback *through* the capsule will not be the same as the effect of purely electrical NFB.

My reasoning (which I'm not sure is right) goes like this:

For signals in the middle range, between low hundreds and low thousands of Hz, an SM-57 puts out a much stronger signal than a similar-diameter condenser. At least in that middle band, then, the condenser appears to be a very inefficient transducer.

The condenser may have a pretty flat frequency response down to 25 Hz or further, but unless it has a huge bass bump, it can't be a dramatically more efficient transducer in the bass than it is in the midrange. So presumably, it's comparably inefficient in the bass to how inefficient it is in the midrange.

 

[OPR]

Let's all move on and put our energy into devising a test that we can all agree on as this back and forth BS is a waste of time and energy...
The push and pull of this thread has had good some food for thought so let's pull together and get to the bottom of this.

I think a null test is a good idea as long as the test method is sound... Boom tshhh! Pardon the pun cough cough!
In this case a sweep won't do it because it doesn't reflect the real world application of the filter and isolating one frequency at time wont show what happens when lower frequencies are present at the same time as higher frequencies.

Here's a test I just tried to confirm possibly similar to what Korgs tried? but I don't know as no real detail or wav files wee offered...

I've devised a possible test details below.

So I tried two method of nulling the files.

Eq generated from pink nose with S2 on, captured by Fab filter applied to recorded music with s2 off
Eq generated from music, captured by Fab filter applied to recorded music with s2 off
As a control, I also recorded a 2 passes of the 67 with S2 on and nulled as a reference, as this would represent the best case scenario. ie So the theory that is ,if the EQ'd files match the cancellation of the unprocessed Audio recorded in separate 2 passes as that would indicate that the only difference was in the frequency domain if not, then other forces must be at play that cannot be replicated with EQ.

So far in my test fab filter pro Q was unable compensate for the difference enough to equal the null of the unprocessed files.

You can download them here.

THE S2 FILES

This is just an Idea so if you have any improvements or if the method is flawed in some way or you have a better method please post up.

Peace!

 

[kingkorg]

Eq curve has to be exact as in the circuit for this to work. I used REW and CurveEQ to get the curve.

I inject signal from REW into messeingang on u87. I get the de-emphasis curve.

Then i import this measurement into Voxengo Curve EQ to get this exact curve and use it on disabled NFB file.



Also even if i am wrong, if we have perfectly positioned microphones the midrange should still cancel out as the NFB isn't affecting those frequencies.


If we focus strictly on the low end cancelation the positioning of the mic doesn't have to be that critical because of the wavelength.

If we go totally crazy about this, and want to avoid variables, the S1 switch could be disconnected remotely.
For this test to work, nothing in the testing room should change, so no humans or other moving objects in the room.

 

[OPR]

Eq curve has to be exact as in the circuit for this to work. I used REW and CurveEQ to get the curve.

I inject signal from REW into messeingang on u87. I get the de-emphasis curve. View attachment 141855

Then i import this measurement into Voxengo Curve EQ to get this exact curve and use it on disabled NFB file.

View attachment 141856

Also even if i am wrong, if we have perfectly positioned microphones the midrange should still cancel out as the NFB isn't affecting those frequencies.


If we focus strictly on the low end cancelation the positioning of the mic doesn't have to be that critical because of the wavelength.

If we go totally crazy about this, and want to avoid variables, the S1 switch could be disconnected remotely.
For this test to work, nothing in the testing room should change, so no humans or other moving objects in the room.

Thanks for the extra detail Korg.

Firstly Is the test signal a sweep?

If so.
Did you read my previous post regarding sweeps and this particular test?
Here it is again.

"In this case a sweep won't do it because it doesn't reflect the real world application of the filter and isolating one frequency at time wont show what happens when lower frequencies are present at the same time as higher frequencies."

In other words your method wont reveal what we are looking for as the NFB, being input dependent, isn't active while you're recording any frequency above it's cutoff point so none of the potential damping effects that I'm hypothesizing will be present.

Does that make sense to you?

You could probably prove this to your self by recording full band width music music through your test setup and recording a version with the feedback enabled and disabled as well a copy of the disabled track as a control and applying you curve so you can see if the null is equal.

 

[kingkorg]

Thanks for the extra detail Korg.

Firstly Is the test signal a sweep?

If so.
Did you read my previous post regarding sweeps and this particular test?
Here it is again.

"In this case a sweep won't do it because it doesn't reflect the real world application of the filter and isolating one frequency at time wont show what happens when lower frequencies are present at the same time as higher frequencies."

In other words your method wont reveal what we are looking for as the NFB, being input dependent, isn't active while you're recording any frequency above it's cutoff point so none of the potential damping effects that I'm hypothesizing will be present.

Does that make sense to you?

You could probably prove this to your self by recording full band width music music through your test setup and recording a version with the feedback enabled and disabled as well a copy of the disabled track as a control and applying you curve so you can see if the null is equal.

I did the original test with pink noise and a drum beat. Makes sense what you wrote
Also this is with u87. The NFB circuit is way different.