[Angus Kerr]

I thought I'd start a new thread since this has gone wildly off-course.

I've just completed the mod, and I must say, it not only looks fantastic, but sounds much better too! A lot of the things I didn't like are, I won't say gone, but significantly reduced.

One thing though, because the original grille was integrated and one with the body, it was earthed. In my attempt to make the mic look good, I inadvertently isolated the grille from the body of the mic with paint!. So I have a ground hum as the capsule is not faraday caged properly. I'll have to scrape off some paint and try to get a good earth on those grilles.

I'll trawl the mic modder forums regarding the resonator - baffle, but here I trust the guy who actually did the granddaddy of mods. The internet is full of weird strange people who say lots of things.

On Monday, September 28, 2015 at 10:38:10 PM UTC+2, Peter Larsen wrote:
On 28-09-2015 21:08, Scott Dorsey wrote:

Angus Kerr wrote:

@Scott: I wanted to ask you what you recommend. I see there are two grilles=
, inner and outer. Should I put both back, or is it better with just one? I=
see the Michael Joly mod uses single grille. What's your take?

I tend to leave both in because it's more rugged. It might sound better
with one, but it won't take a drumstick.

@Scott - I'm hoping the mods take some of that away. I'm also hearing over-=
hyped top end. Hard to believe from what I'm hearing that the U87 have a pr=
esence peak and the 219 does not. It sounds the other way round to me. Will=
taking the baffles out smooth it out? It's a lot more aggressive in the to=
p end than the U87, which brings out none of that sibilance.=20

Both the U87 and the 219 have exaggerated top end, but they are different
kinds of exaggerated top ends.

As I recall this from some mic-modder site the 219 has a hf resonator in
front of the membrane that causes its peak. Remove it and - still from
memory - it will roll off from 10 kHz instead with a much more benign
and easili compensated response. My recollection may be incomplete, do
not read as a how2 manual, but as a search4.

--scott

Kind regards

Peter Larsen

[Angus Kerr]

On Monday, September 28, 2015 at 11:23:46 PM UTC+2, Angus Kerr wrote:
I thought I'd start a new thread since this has gone wildly off-course.

I've just completed the mod, and I must say, it not only looks fantastic, but sounds much better too! A lot of the things I didn't like are, I won't say gone, but significantly reduced.

One thing though, because the original grille was integrated and one with the body, it was earthed. In my attempt to make the mic look good, I inadvertently isolated the grille from the body of the mic with paint!. So I have a ground hum as the capsule is not faraday caged properly. I'll have to scrape off some paint and try to get a good earth on those grilles.

I'll trawl the mic modder forums regarding the resonator - baffle, but here I trust the guy who actually did the granddaddy of mods. The internet is full of weird strange people who say lots of things.

On Monday, September 28, 2015 at 10:38:10 PM UTC+2, Peter Larsen wrote:
On 28-09-2015 21:08, Scott Dorsey wrote:

Angus Kerr wrote:

@Scott: I wanted to ask you what you recommend. I see there are two grilles=
, inner and outer. Should I put both back, or is it better with just one? I=
see the Michael Joly mod uses single grille. What's your take?

I tend to leave both in because it's more rugged. It might sound better
with one, but it won't take a drumstick.

@Scott - I'm hoping the mods take some of that away. I'm also hearing over-=
hyped top end. Hard to believe from what I'm hearing that the U87 have a pr=
esence peak and the 219 does not. It sounds the other way round to me. Will=
taking the baffles out smooth it out? It's a lot more aggressive in the to=
p end than the U87, which brings out none of that sibilance.=20

Both the U87 and the 219 have exaggerated top end, but they are different
kinds of exaggerated top ends.

As I recall this from some mic-modder site the 219 has a hf resonator in
front of the membrane that causes its peak. Remove it and - still from
memory - it will roll off from 10 kHz instead with a much more benign
and easili compensated response. My recollection may be incomplete, do
not read as a how2 manual, but as a search4.

--scott

Kind regards

Peter Larsen

On Monday, September 28, 2015 at 11:23:54 PM UTC+2, Scott Dorsey wrote:
Angus Kerr wrote:
Both the U87 and the 219 have exaggerated top end, but they are different
kinds of exaggerated top ends.

Taking off the baffle will do ....? Or should I just take one off and take a listen...

Do you mean the resonator in front of the capsule? Don't remove that, it
does add a high frequency pole but that's not a bad thing.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Angus Kerr]

On Monday, September 28, 2015 at 11:35:25 PM UTC+2, Angus Kerr wrote:
On Monday, September 28, 2015 at 11:23:46 PM UTC+2, Angus Kerr wrote:
I thought I'd start a new thread since this has gone wildly off-course.

I've just completed the mod, and I must say, it not only looks fantastic, but sounds much better too! A lot of the things I didn't like are, I won't say gone, but significantly reduced.

One thing though, because the original grille was integrated and one with the body, it was earthed. In my attempt to make the mic look good, I inadvertently isolated the grille from the body of the mic with paint!. So I have a ground hum as the capsule is not faraday caged properly. I'll have to scrape off some paint and try to get a good earth on those grilles.

I'll trawl the mic modder forums regarding the resonator - baffle, but here I trust the guy who actually did the granddaddy of mods. The internet is full of weird strange people who say lots of things.

On Monday, September 28, 2015 at 10:38:10 PM UTC+2, Peter Larsen wrote:
On 28-09-2015 21:08, Scott Dorsey wrote:

Angus Kerr wrote:

@Scott: I wanted to ask you what you recommend. I see there are two grilles=
, inner and outer. Should I put both back, or is it better with just one? I=
see the Michael Joly mod uses single grille. What's your take?

I tend to leave both in because it's more rugged. It might sound better
with one, but it won't take a drumstick.

@Scott - I'm hoping the mods take some of that away. I'm also hearing over-=
hyped top end. Hard to believe from what I'm hearing that the U87 have a pr=
esence peak and the 219 does not. It sounds the other way round to me. Will=
taking the baffles out smooth it out? It's a lot more aggressive in the to=
p end than the U87, which brings out none of that sibilance.=20

Both the U87 and the 219 have exaggerated top end, but they are different
kinds of exaggerated top ends.

As I recall this from some mic-modder site the 219 has a hf resonator in
front of the membrane that causes its peak. Remove it and - still from
memory - it will roll off from 10 kHz instead with a much more benign
and easili compensated response. My recollection may be incomplete, do
not read as a how2 manual, but as a search4.

--scott

Kind regards

Peter Larsen

On Monday, September 28, 2015 at 11:23:54 PM UTC+2, Scott Dorsey wrote:
Angus Kerr wrote:
Both the U87 and the 219 have exaggerated top end, but they are different
kinds of exaggerated top ends.

Taking off the baffle will do ....? Or should I just take one off and take a listen...

Do you mean the resonator in front of the capsule? Don't remove that, it
does add a high frequency pole but that's not a bad thing.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

OK, if you insist. You did mention in your article that it would be good to leave it in for vocals, take it out for horns.

What I'd like to know is exactly how does it change the response? The mic is very bright (I think) as it is, and could do with a little less aggressive upper mid-range (3-6kHz). I see the plot on recording hacks there are quite a few bumps between 1 and 10kHz

-A.

[Peter Larsen[_3_]]

On 28-09-2015 22:40, Angus Kerr wrote:

Do you mean the resonator in front of the capsule? Don't remove that, it
does add a high frequency pole but that's not a bad thing.

It is a high Q thing, I'd rather have a roll off that is easy to
compensate than a source of sibilance.

--scott

Hey, I got to disagree with Scott. Of course easier not having heard the
mic in modded state. I read a lot up on it a couple of years ago because
a friend asked whether to buy. He ended up getting a pair in ex works
state and is happy with them for vox.

OK, if you insist. You did mention in your article that it would be
good to leave it in for vocals, take it out for horns.

Can't comment.

What I'd like to know is exactly how does it change the response?

As I remember this - I may or may not have a pdf somewhere, but a couple
of desktops have died on me since - it is a sharp peak centered at 14 kHz.

The mic is very bright (I think) as it is, and could do with a little
less aggressive upper mid-range (3-6kHz).

Check your actual mics. I reckon that most DAW software nowadays have
all the analysis that CoolEdit96 was capable of. Record a violin and
make a FFT analysis with longest possible wordlength, you can make a
parallel recording with a mic with known response, but string quartets
are excellent measurement noise sources, at least using quality catgut.
Compare.

You need at least 15 minutes of good audio to analyze, there is probably
some suitable partita by Bach's wife attributed to him. Or just record
room noise, it can be surprisingly smooth, in fact a characteristic of a
good room is that it is smooth and it gives you a lot better correlation
to perceived sound of mic than on axis response.

I see the plot on recording hacks there are quite a few bumps
between 1 and 10kHz

Not gonna look - other things to do today. Also this is about your
actual mics, not about those they tested.

You can also measure the playback of white noise via your monitors and
use a known linear mic as input in a compressor sidechain to remove the
loudspeaker response. That too will give you the a weighted "on axis +
random" response. To test the setup measure a known linear mic, it
should come out as linear. I have not tested actually doing this, but it
is what KinoVox told me they did in the old days when they also did
engineering and design and had a dead room and verified peoples
microphones for them.

-A.

Kind regards

Peter Larsen

[Phil Allison[_4_]]

Peter Larsen wrote:



You can also measure the playback of white noise via your monitors and
use a known linear mic as input in a compressor sidechain to remove the
loudspeaker response. That too will give you the a weighted "on axis +
random" response. To test the setup measure a known linear mic, it
should come out as linear. I have not tested actually doing this, but it
is what KinoVox told me they did in the old days when they also did
engineering and design and had a dead room and verified peoples
microphones for them.



** Think you have mixed up the procedure a bit, cos that is not gonna work.

The standard way of calibrating mics in an anechoic room uses a test speaker and a known flat response mic ( normally a small diaphragm omni ) placed on axis and connected in a gain control feedback loop with the amplifier as you described, BUT the test signal is a slow swept sine tone. When set up and adjusted correctly, the signal level from the reference mic is the same at all frequencies - so it can be assumed the SPL at that spot remains the same.

The mic to be calibrated is simply placed close to the reference mic and its signal level recorded during a sweep, which gives its response.

If time is no object, one could do a large number of spot frequency tests instead, noting the difference in level between the reference and
mic under calibration at each spot and plot a curve manually.

If 1/3 octave pink noise signals were available, they could be used instead of sine tones, with the disadvantages of less precision and unsteady levels at low and mid frequencies.

I have tried these two myself on occasion, using a AKG CK2 as a reference, and can assure readers that the entertainment value rarely exceeds the tedium.


.... Phil

[Mike Rivers[_2_]]

On 9/29/2015 3:28 AM, Peter Larsen wrote:
You can also measure the playback of white noise via your monitors and
use a known linear mic as input in a compressor sidechain to remove the
loudspeaker response. That too will give you the a weighted "on axis +
random" response. To test the setup measure a known linear mic, it
should come out as linear. I have not tested actually doing this, but it
is what KinoVox told me they did in the old days when they also did
engineering and design and had a dead room and verified peoples
microphones for them.

What an interesting technique. I should try it one of these days. It
doesn't sound like it would work, but it probably does something
interesting.

--
For a good time, visit http://mikeriversaudio.wordpress.com

[Scott Dorsey]

Angus Kerr wrote:
OK, if you insist. You did mention in your article that it would be good to=
leave it in for vocals, take it out for horns.

What I'd like to know is exactly how does it change the response? The mic i=
s very bright (I think) as it is, and could do with a little less aggressiv=
e upper mid-range (3-6kHz). I see the plot on recording hacks there are qui=
te a few bumps between 1 and 10kHz

The baffle does three things:

1. It forms a helmholtz resonator, with the entrance around the edge and an
otherwise sealed chamber in front of the diaphragm. Because the entrance
is so big, the Q of the resonance is very wide, and because the space is
so small, the center frequency is very high. This increases the response
in the top octave.

2. It forms a blast filter so that the plosives do not cause the diaphragm to
bottom out.

3. Because now the entrance is around the edge instead of in the center it
alters the microphone pattern.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Angus Kerr]

On Tuesday, September 29, 2015 at 2:55:34 PM UTC+2, Scott Dorsey wrote:
Angus Kerr wrote:
-snip-

What I'd like to know is exactly how does it change the response? The mic i=
s very bright (I think) as it is, and could do with a little less aggressiv=
e upper mid-range (3-6kHz). I see the plot on recording hacks there are qui=
te a few bumps between 1 and 10kHz

The baffle does three things:

1. It forms a helmholtz resonator, with the entrance around the edge and an
otherwise sealed chamber in front of the diaphragm. Because the entrance
is so big, the Q of the resonance is very wide, and because the space is
so small, the center frequency is very high. This increases the response
in the top octave.

2. It forms a blast filter so that the plosives do not cause the diaphragm to
bottom out.

3. Because now the entrance is around the edge instead of in the center it
alters the microphone pattern.

Thanks for the info: I'll leave it for now. I presume some electronic compensation is required if the baffle is removed?

I'll give this mic its due: it is tough as nails

Mine were manufactured in 1995, spent most of their time in their musty box in a climate where in summer humidity goes up to almost 90% and temperatures in the '90's and beyond. When I pulled them out the nasty foam had disintegrated and deposited itself all over the mic and capsule.

I plugged it in and it is as quiet as anything - breathe on it, nothing. Clean as a whistle. I was pleasantly surprised. When I pulled the first one apart, corrosion is happening to some parts, and the diaphragm has bits of nasty foam residue on it, and it just works. Can understand that this would have been a great mic for broadcast and hard knocks.

The MK219-SDM (Scott Dorsey Mechanical) mod has made it a little less harsh and more transparent in the upper mids, but it is still quite forward - it'll probably be better in a dense mix where maybe more 'cut' is required. I'll probably do the electronic mods as well - mine is I think a rev 1 board.. Then I'll have the MK219-SDME!

All I have to do is find some way of electrically bonding the basket grill to the body for a vox session tomorrow - I'm thinking of mixing in graphite and aluminium filings into some epoxy to make a conductive glue.

Anybody done this?

@Scott, how did you ensure that the grille was electrically connected to the microphone body?

-Angus.

[Peter Larsen[_3_]]

On 29-09-2015 10:32, Phil Allison wrote:
Peter Larsen wrote:



You can also measure the playback of white noise via your monitors and
use a known linear mic as input in a compressor sidechain to remove the
loudspeaker response. That too will give you the a weighted "on axis +
random" response. To test the setup measure a known linear mic, it
should come out as linear. I have not tested actually doing this, but it
is what KinoVox told me they did in the old days when they also did
engineering and design and had a dead room and verified peoples
microphones for them.



** Think you have mixed up the procedure a bit, cos that is not gonna work.

The standard way of calibrating mics in an anechoic room uses a test speaker and a known flat response mic ( normally a small diaphragm omni ) placed on axis and connected in a gain control feedback loop with the amplifier as you described, BUT the test signal is a slow swept sine tone. When set up and adjusted correctly, the signal level from the reference mic is the same at all frequencies - so it can be assumed the SPL at that spot remains the same.

The mic to be calibrated is simply placed close to the reference mic and its signal level recorded during a sweep, which gives its response.

Thank you for fixing my suggestion!

Kind regards

Peter Larsen

If time is no object, one could do a large number of spot frequency tests instead, noting the difference in level between the reference and
mic under calibration at each spot and plot a curve manually.

If 1/3 octave pink noise signals were available, they could be used instead of sine tones, with the disadvantages of less precision and unsteady levels at low and mid frequencies.

I have tried these two myself on occasion, using a AKG CK2 as a reference, and can assure readers that the entertainment value rarely exceeds the tedium.


... Phil

[Scott Dorsey]

Angus Kerr wrote:

Thanks for the info: I'll leave it for now. I presume some electronic compe=
nsation is required if the baffle is removed?

You can't compensate electronically for what the baffle does, because what
the baffle does varies with direction.

But... you can remove the baffle, put it in a box, and decide for yourself
if you like it, then put it back on again. The problem with doing this is
that by the time you realize you were better off with it, you might have
lost the box. I spent an hour looking for a box of stepped attenuators in
the garage yesterday for a similar reason.

All I have to do is find some way of electrically bonding the basket grill =
to the body for a vox session tomorrow - I'm thinking of mixing in graphite=
and aluminium filings into some epoxy to make a conductive glue.

Anybody done this?

I never had a problem with that, but aluminum won't work. Graphite probably
will. The commercial stuff they sell is silver-powder-based.

@Scott, how did you ensure that the grille was electrically connected to th=
e microphone body?

I just crimped it in with an arbor press and didn't do anything else. I did
scrape the paint off the inside rim after repainting the case though. But
conductive epoxy is certainly not a bad plan.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Angus Kerr]

On Friday, October 2, 2015 at 4:06:35 PM UTC+2, Scott Dorsey wrote:
Angus Kerr wrote:

Thanks for the info: I'll leave it for now. I presume some electronic compe=
nsation is required if the baffle is removed?

You can't compensate electronically for what the baffle does, because what
the baffle does varies with direction.

I meant to cope with the rolled off top end..
But... you can remove the baffle, put it in a box, and decide for yourself
if you like it, then put it back on again. The problem with doing this is
that by the time you realize you were better off with it, you might have
lost the box. I spent an hour looking for a box of stepped attenuators in
the garage yesterday for a similar reason.

Luckily I have a spare drawer in my component trays. But I'm so happy, I'll keep it as is.


All I have to do is find some way of electrically bonding the basket grill =
to the body for a vox session tomorrow - I'm thinking of mixing in graphite=
and aluminium filings into some epoxy to make a conductive glue.

Anybody done this?

I never had a problem with that, but aluminum won't work. Graphite probably
will. The commercial stuff they sell is silver-powder-based.


Update, I tried a combination of aluminium powder and graphite mixed up with epoxy.. it certainly was a great insulator, I'm talking hundreds of megohms with the probes of my multimeter about 2mm apart. What I did was remove the epoxy, scrape all of the paint off the inside where the grille made contact, and viola! instant silence. I then epoxied back.

I am pleased with this mike, the SDM mods have opened it up, removed the upper mid harshness, and generally taken it from a 'well, I've got nothing else so...' to a 'hmmm, this actually is working for this voice better than the U87'. Maybe not quite. But is is sounding like a decent mike, with a different and pleasant sounding colour palette to the U87. Almost C414 like, but maybe not as neutral.

I have two of these, and deliberately modified only one, so I could A/B the unmodified one against the modified one. Bearing in mind that they probably sound different anyway since they are 1995 vintage, rev 1 board and metal can FET, but the difference is NOT subtle. The unmodified one has some nasty upper midrange peaks, is sibilant, and doesn't have the extended lower range either, and had a veiled top end. The modded mike is sounding great. Now to do a proper 'pre' and 'post' recording of the other to really try to get a benchmark comparison. I might do violin, vocal and acoustic guitar. Really try to get the same placement before and after.

-Angus.

[Mat Nieuwenhoven]

On Tue, 29 Sep 2015 02:32:16 -0700 (PDT), Phil Allison wrote:

snip

The standard way of calibrating mics in an anechoic room uses a
test speaker and a known flat response mic ( normally a small
diaphragm omni ) placed on axis and connected in a gain control
feedback loop with the amplifier as you described, BUT the test
signal is a slow swept sine tone. When set up and adjusted
correctly, the signal level from the reference mic is the same at all
frequencies - so it can be assumed the SPL at that spot remains the
same.

The mic to be calibrated is simply placed close to the reference mic
and its signal level recorded during a sweep, which gives its response.

If time is no object, one could do a large number of spot frequency
tests instead, noting the difference in level between the reference and
mic under calibration at each spot and plot a curve manually.

If 1/3 octave pink noise signals were available, they could be used
instead of sine tones, with the disadvantages of less precision and
unsteady levels at low and mid frequencies.

The Windows software from http://www.artalabs.hr/ might help in such
a
case (it's originally for loudspeaker measurements). It can make
measurement based on a known response of a measurement mic, which
doesn't have to be flat if the deviations are documented. It uses a
variety of noise and sine signals. You don't need necessarily a dead
room, just a large one: it can use only the sound up to the first
reflection, at least in some cases.

Mat Nieuwenhoven