D
DJL
Self Banned
PsyCoNo said:
That is interesting and sometimes I wonder if mic graft/charts are of any real help?
That is interesting and sometimes I wonder if mic graft/charts are of any real help?
F
Flatpicker
...
You're very welcome - just glad I'm good for something.PsyCoNo said:
F
Flatpicker
...
Ok, I didn’t really have time to type all this, but I just can't help myself…DJL said:
I’ve come to the conclusion that the frequency charts are mostly useless. You have to “hear” the mic WITH the source you’re recording it with – period. No freq chart is going to tell you what it will sound like. You can take a signal generator, feed a square wave into an amp, and sweep the frequency from 20-20khz, then repeat the process with a sine wave. Of course, you can see where I’m going with this - they will have the same frequency response, but sound completely different!
Another problem with the freq charts is that they are compiled and averaged from several mics, not just one. Also, there’s no law keeping the manufacturer from fudging the data just a little. And trust me, they do!
The differences in mics is how they “distort”. Yes, you heard me right, but remember, distortion isn’t always a bad thing. For example, why do we like the sound of tubes? Because they distort in a good way (well, some of them do, anyway - heh, heh). So when we say that we like the way a mic “colors” the sound, we are actually saying that we like the way it “distorts” the sound. In most cases, we wouldn't even like a mic that exactly reproduced the source with no distortion. We’d call it “sterile”.
Now, my gripe is this: Why doesn’t the industry standardize a method for measuring microphone distortion and publish a chat showing the harmonic frequency peaks? To me, that would be WAY more useful for mic comparisons.
Again, my $.02 +/- a few dB…
BTW, I'm very intrested in hearing opinions/samples of the NT1A. (thanks for yours, C9). I definitely like the nickel plated body better than the old painted one.
P
PsyCoNo
New member
Don't they already give distortion spec's? Like for the C1, @ 142dB SPL the THD is only 1% of the output. . . is this what you meant or did you mean the so called 'good' distortion. Also, it would seem to me that every persons definition of 'good' will be slightly different. How would they account for this?
F
Flatpicker
...
Well, THD, “total harmonic distortion” is just that - a “total” measurement. It doesn’t tell us which harmonics are distorting. When I speak of distortion, I’m speaking of the (in)accurate way each mic reproduces the sounds applied to it. It will emphasize or de-emphasize certain harmonics and that’s what gives it its sound or color. Some mics appear to sound clean and clear and some sound warm and thick. This is the mic’s electro/mechanical reaction to the sound applied to it. When it emphasizes/de-emphasizes certain harmonics, it changes, or “distorts” the original sound. Also, when you apply two or more sounds (notes, instruments, voices, whatever) simultaneously, things get WAY more complicated.PsyCoNo said:
There would have to be some standardized test whereby various tones, sounds, etc… are fed into the mic and a spectrum analysis made of the output. I have no idea how that could be done correctly and justifiably, but it would have to be standardized.
The same goes for preamps. For example, years ago Rupert Neve and other great engineering gurus set out to design a clean and accurate mic pre. They instead created something that distorted in all the right places (whether accidentally or on purpose) to give us that great “Neve” sound.
BTW, I don’t claim to be an “expert” at any of this stuff, and in some cases I might just be plain wrong, but this is the way I see it. And of course, YMMV
alanhyatt
PMI Audio Group
Flatpicker said:Ok, I didn’t really have time to type all this, but I just can't help myself…
I’ve come to the conclusion that the frequency charts are mostly useless. You have to “hear” the mic WITH the source you’re recording it with – period.
Also, there’s no law keeping the manufacturer from fudging the data just a little. And trust me, they do!
Why doesn’t the industry standardize a method for measuring microphone distortion and publish a chat showing the harmonic frequency peaks?
Yes, you are correct. Graphs mean squat, and there is no Ad Hoc agency to regulate any testing proceedures like there is on other products. Any manufacturer can basically test and print what they want to obtain great graphs and specs.
This has been the way it is for a long time, and many companies take advantage of it. We don't!
P
PsyCoNo
New member
It is hard to measure which harmonics are distorting though (or so it seems to me). Every sound that is produced emits different harmonics (unless it is a perfect sine wave of infinite duration, hehe). Harmonics are always whole number multiples of the fundamental frequency. The second harmonic is exactly twice the frequency of the fundamental. With all this being true, in order to get an accurate reading on the microphone you would need to test tones at all freqencies individually. Once you did this, how would you even write it? Would you just show the highest levels of harmonic distortion? And, if a mic distorts the harmonics why wouldnt it also distort the fundamental frequency if it was at the same frequency (not at the same time of course). Im not sure if you are following me at all. . . hopefully. . . lol
Sorry about being hard about this, I am finding it very interesting. Im not quite sure what Im talking about (I think I know just enough to get me in trouble, haha). But, if there is a way to do this, it would yield very useful indeed.
Sorry about being hard about this, I am finding it very interesting. Im not quite sure what Im talking about (I think I know just enough to get me in trouble, haha). But, if there is a way to do this, it would yield very useful indeed.
F
Flatpicker
...
Heh heh. Join the club!PsyCoNo said:
It could all be displayed very easily using a spectrum analyzer. The difficult part is deciding what sounds to put into the mic. Do you use more than one sound at a time? Do you play notes that make a chord? Do you play notes that don't make a chord? Do you use white noise? etc, etc...
All I know is, someone way smarter than me would have to figure it out.
P
PsyCoNo
New member
Right, but the spectrum analysis can only show one set of harmonics at a time. . . otherwise you wont be able to see individual harmonics, right? lol. . .i have no idea.
C
c9-2001
New member
i just left my friends studio.. we were tracking vocals with a u87ai going into his soundcraft ghost.. sounded great.. but we hooked up a c1 with a neve pre.. o my.. sounded better than the u87ai..
(when we plugged the u87ai into the neve, you know what mic came out on top)
i've been under-rating the C1.. now i know what it can do with a good pre in the signal chain..i might just have to order one
(when we plugged the u87ai into the neve, you know what mic came out on top)
i've been under-rating the C1.. now i know what it can do with a good pre in the signal chain..i might just have to order one
O
ozraves
New member
c9-2001 said:
i was kind of shocked when i plugged my c1 into a grace 101. it took me about two seconds to realize that a quality pre drastically improved the sound of this mic.
D
DJL
Self Banned
Flatpicker said:Heh heh. Join the club!
It could all be displayed very easily using a spectrum analyzer. The difficult part is deciding what sounds to put into the mic. Do you use more than one sound at a time? Do you play notes that make a chord? Do you play notes that don't make a chord? Do you use white noise? etc, etc...
All I know is, someone way smarter than me would have to figure it out.
You'd use a pink noise generator.
B
billisa
New member
This could be wrong, but I recall back in the 70's the FTC came in and enforced specs standards on the home audio market, regarding things like watts per channel, etc. because some manufacturers were putting out numbers for their receivers and amps that were complete balderdash. "500watts total power!!!"
Personally, while I totally believe numbers don't tell the whole story, on things like s/n ratio, or on self-noise, they can be very helpful, if properly stated... For instance, the Studio Projects B1's stated self-noise is like 12db. On the Samson CO1 it's 23db. When you hear these two you realise the specs give a very accurate picture of what you can expect. If manufacturers were obliged (forced) to state these figures in meaningful ways, as determined by experts who can determine the parameters, the consumer would greatly benefit. If I want to record a pipe organ at 20hz, then if the mic I'm using has a stated response of 20hz (-1db), then I ought to be able to expect that the mic has picked up what I'm recording with some accuracy, not at - 45db...
This is particularly true in an increasingly internet oriented shopping economy. Objective standards could dramatically assist people who simply can't try out potential purchases online.
I know that specs can't be totally relied on, but if I buy a car that promises 35mpg highway, it ought to be in the ballpark and not dependent on only driving downhill on a 42 degree slope being towed by a tractor trailer, with teflon bushings and graphite bearing grease, during a 65mph tailwind...
Personally, while I totally believe numbers don't tell the whole story, on things like s/n ratio, or on self-noise, they can be very helpful, if properly stated... For instance, the Studio Projects B1's stated self-noise is like 12db. On the Samson CO1 it's 23db. When you hear these two you realise the specs give a very accurate picture of what you can expect. If manufacturers were obliged (forced) to state these figures in meaningful ways, as determined by experts who can determine the parameters, the consumer would greatly benefit. If I want to record a pipe organ at 20hz, then if the mic I'm using has a stated response of 20hz (-1db), then I ought to be able to expect that the mic has picked up what I'm recording with some accuracy, not at - 45db...
This is particularly true in an increasingly internet oriented shopping economy. Objective standards could dramatically assist people who simply can't try out potential purchases online.
I know that specs can't be totally relied on, but if I buy a car that promises 35mpg highway, it ought to be in the ballpark and not dependent on only driving downhill on a 42 degree slope being towed by a tractor trailer, with teflon bushings and graphite bearing grease, during a 65mph tailwind...
L
Lanstar0
New member
Personally, I like the AT3035... 
L
lumbago
New member
I have a Rode NT-1 and a NT-1A and a C1.
The NT-1A is a way better mic. than the NT-1. The NT-1 gets assigned to the ambience mic tasks.
Studio Projects C1 suits more voices than the NT-1A, although I use a Rode NTK in preference to both.
The NT-1A has presence humps which suit instruments like acoustic piano and nylon string guitar, and some male voices.
You need to judge them with your own source material.
I'm using a Focusrite Trakmaster with all these mics (no EQ or compression - just the A-D card) for the purpose of comparison.
Clean enough for any home studio.
The NT-1A is a way better mic. than the NT-1. The NT-1 gets assigned to the ambience mic tasks.
Studio Projects C1 suits more voices than the NT-1A, although I use a Rode NTK in preference to both.
The NT-1A has presence humps which suit instruments like acoustic piano and nylon string guitar, and some male voices.
You need to judge them with your own source material.
I'm using a Focusrite Trakmaster with all these mics (no EQ or compression - just the A-D card) for the purpose of comparison.
Clean enough for any home studio.
B
bleyrad
New member
careful though, there are actually three different "versions" of the NT1:
1) NT1 with original capsule
2) NT1 with "newer" capsule
3) NT1A
So which are we talking about? I have the "old capsule" version of the NT1, which is really a great mic. After they changed the capsule, apparently everyone started hating the new version. And there are mixed feelings on the NT1A.
1) NT1 with original capsule
2) NT1 with "newer" capsule
3) NT1A
So which are we talking about? I have the "old capsule" version of the NT1, which is really a great mic. After they changed the capsule, apparently everyone started hating the new version. And there are mixed feelings on the NT1A.
B
baekgaard
New member
"Objective facts..."
First, I don't have an NT1-A, so I cannot comment on it. I do have a B1 and a C1, though... and have actually made some measurements of the actual frequency response of those samples that I happen to own. Look for my thread on the Oktava mics from GC here: http://www.homerec.com/bbs/showthread.php?s=&threadid=101544
Second, on the issue of putting numbers to why mics sound different, there are many factors, as has already been mentioned. None of these fully describe what we hear, but many can be understood scientifically. I'll make an attempt at explaining some of the most important ones in “layman's” terms (as far as I understand them
-- but now re-reading what I put below, I may have failed utterly in doing so :-(
*** The frequency response of a mic shows how it amplifies/attenuates different frequencies. Usually, only omni mics can have fully flat responses. All other mics (and most real-world omni mics) have peaks and valleys, and how the peaks and valleys are in the graph has a big influence on the sound: It can be present, muddy, warm, having air, sizzle, etc. Some of these terms implies certain frequency areas that are being attenuated.
You can achieve a bit similar effect by playing with a good EQ, but all EQ's have side effects as well, so you cannot really make one mic sound just like another by EQ'ing... although in some cases, you can come close.
*** Another factor is the transient response (impulse response). This is how the mic “behaves” when it sees a transient in the sound, such as a click with a lot of high frequency content. If you measure the transient response of a mic (or any other system), it actually means you can derive the frequency response also! Why is this? Well, a “perfect” click (called a delta function) is a signal that contains all frequencies. If you could record such a click, you would get the “transfer function” of the system being measured, here roughly equivalent to the transient response. By doing a FFT of this, you will get the frequency response.
What about the other way round? Well, in order to get to the transient response of a system, you need to know not only how the frequency amplitude response is, but also the frequency phase response, which is how the phase is distorted for different frequencies. And that information is not usually shown by any vendor...
*** A third factor is the distortion of a microphone. This comes in at least two flavors: harmonic distortion, and what is not harmonic. If you send a pure sine wave (which has no overtones) through a mic, you would expect to hear only that. But due to many factors, the sine wave will not look fully like a perfect sine wave after wards. One commonly seen phenomena is that the top of the signal gets slightly compressed. Physically, what happens is that different over-tones (harmonics) is added to the original signal. In the case of a very distorted signal, you may even see a square wave – which is a signal with a lot of overtones added to it.
[Incidently, this is why square, triangle and saw waveforms are popular on subtractive synthezisers. There are a lot of overtones that can be filtered by your nice analog filters, and that gives the “character” of the sound. If you used sine waves as the fundamental waveform here, the usage of filters would be more or less in vain, and you would get a boring sound.]
Back to mics: Some harmonics sound pleasant to us, some don't. The even-ordered harmonics (i.e base frequency multiplied by 2, 4, 6, etc) are usually nice. These are generated by tube equipment. So, when you record through a tube, you add some nice sounding harmonics, due to the harmonic distortion.
Now, no real sound sources generates pure sine waves. There are always overtones. The relationship between the overtones (i.e. how much there is of each multiple of the base frequency) is an important part of helping your ear distinguish between a pan flute and a distorted guitar. Drums are different, since they have less pure frequencies, but I won't elaborate on that now.
What happens then if a real instrument is recorded? Well, for each frequency existing in the original signal (including the original overtones) new harmonics will be added. All will be a multiple of the fundamental (or root) frequency. There is no real difference between the ones from the sound source and the ones you add... so, by e.g. recording through a tube, you change the sound character. This is actually distortion, and in some cases, that is what you look for.
I have tested a B1 and a TB1, for instance. The TB1 sounded a slight bit nicer on my acoustic guitar, since it added some pleasantly sounding harmonics – just a tiny bit, but enough for the ear to detect it. Same applied on my voice. Another extreme of this is a distorted guitar, where the sole purpose is to get a pleasantly sounding distortion from your Marshall stack
So harmonic distortion is often good in very small quantities – and in some cases even in large quantities But the more you add, the more it will change the character of your sound. If you have a really nice acoustic guitar, you might want to record it as faithfully as possible!
*** One other kind of distortion is adding “noise” of different kinds. It if happens at frequencies that were not there originally (or even at certain harmonics, usually un-even harmonics), it does not sound nice.
*** A third kind of distortion is also important to understand. If you have to frequencies at the same time that are not related, you might get inter-modulation or cross-modulation. TO make a long story short, it means that you suddenly add a frequency that was never there originally, and is not in some “nice” relation to the original frequencies. This is almost always bad! If you have, say, a 2 kHz tone and a 2.3 kHz tone, they may cross-modulate, and you then suddenly start getting a 300 Hz tone as well.
You ear usually does this for you, btw... and is one part of how the ear detects low frequencies. If a deep bass at, say 30 Hz, is present, there will be also 60 Hz, 90 Hz, 120 Hz, 150 Hz, etc present. Now, if even if you remove the 30 Hz completely, the ear will think it heard the 30 Hz signal, if it detects 60 Hz, 90 Hz, 120 Hz, 150 Hz, etc. This can be used to “cheat” the aural system to hear deeper sounds than what the speakers really can produce. And that is why a “straight” bass tone in some cases may be less interesting than one with a lot of overtones (which you may get by recording through a tube amp).
*** Still another factor that is not fully described by the frequency response is resonance. Resonance will certainly show up as a peak in the frequency response, but it will also make its presence known by adding certain frequencies even when none of those were present in the original sound. Think about a drum, for one second. When you hit it, you hear its resonance. But a drum also sounds when you hit other drums nearby it, or if the bass player hits an annoying note that makes your entire kit rattle... Same can be heard with a microphone. The MK-219 is one famous example of an annoying resonance, I've been told.
Hope some of you found this interesting...
-- Per.
First, I don't have an NT1-A, so I cannot comment on it. I do have a B1 and a C1, though... and have actually made some measurements of the actual frequency response of those samples that I happen to own. Look for my thread on the Oktava mics from GC here: http://www.homerec.com/bbs/showthread.php?s=&threadid=101544
Second, on the issue of putting numbers to why mics sound different, there are many factors, as has already been mentioned. None of these fully describe what we hear, but many can be understood scientifically. I'll make an attempt at explaining some of the most important ones in “layman's” terms (as far as I understand them
-- but now re-reading what I put below, I may have failed utterly in doing so :-(*** The frequency response of a mic shows how it amplifies/attenuates different frequencies. Usually, only omni mics can have fully flat responses. All other mics (and most real-world omni mics) have peaks and valleys, and how the peaks and valleys are in the graph has a big influence on the sound: It can be present, muddy, warm, having air, sizzle, etc. Some of these terms implies certain frequency areas that are being attenuated.
You can achieve a bit similar effect by playing with a good EQ, but all EQ's have side effects as well, so you cannot really make one mic sound just like another by EQ'ing... although in some cases, you can come close.
*** Another factor is the transient response (impulse response). This is how the mic “behaves” when it sees a transient in the sound, such as a click with a lot of high frequency content. If you measure the transient response of a mic (or any other system), it actually means you can derive the frequency response also! Why is this? Well, a “perfect” click (called a delta function) is a signal that contains all frequencies. If you could record such a click, you would get the “transfer function” of the system being measured, here roughly equivalent to the transient response. By doing a FFT of this, you will get the frequency response.
What about the other way round? Well, in order to get to the transient response of a system, you need to know not only how the frequency amplitude response is, but also the frequency phase response, which is how the phase is distorted for different frequencies. And that information is not usually shown by any vendor...
*** A third factor is the distortion of a microphone. This comes in at least two flavors: harmonic distortion, and what is not harmonic. If you send a pure sine wave (which has no overtones) through a mic, you would expect to hear only that. But due to many factors, the sine wave will not look fully like a perfect sine wave after wards. One commonly seen phenomena is that the top of the signal gets slightly compressed. Physically, what happens is that different over-tones (harmonics) is added to the original signal. In the case of a very distorted signal, you may even see a square wave – which is a signal with a lot of overtones added to it.
[Incidently, this is why square, triangle and saw waveforms are popular on subtractive synthezisers. There are a lot of overtones that can be filtered by your nice analog filters, and that gives the “character” of the sound. If you used sine waves as the fundamental waveform here, the usage of filters would be more or less in vain, and you would get a boring sound.]
Back to mics: Some harmonics sound pleasant to us, some don't. The even-ordered harmonics (i.e base frequency multiplied by 2, 4, 6, etc) are usually nice. These are generated by tube equipment. So, when you record through a tube, you add some nice sounding harmonics, due to the harmonic distortion.
Now, no real sound sources generates pure sine waves. There are always overtones. The relationship between the overtones (i.e. how much there is of each multiple of the base frequency) is an important part of helping your ear distinguish between a pan flute and a distorted guitar. Drums are different, since they have less pure frequencies, but I won't elaborate on that now.
What happens then if a real instrument is recorded? Well, for each frequency existing in the original signal (including the original overtones) new harmonics will be added. All will be a multiple of the fundamental (or root) frequency. There is no real difference between the ones from the sound source and the ones you add... so, by e.g. recording through a tube, you change the sound character. This is actually distortion, and in some cases, that is what you look for.
I have tested a B1 and a TB1, for instance. The TB1 sounded a slight bit nicer on my acoustic guitar, since it added some pleasantly sounding harmonics – just a tiny bit, but enough for the ear to detect it. Same applied on my voice. Another extreme of this is a distorted guitar, where the sole purpose is to get a pleasantly sounding distortion from your Marshall stack
So harmonic distortion is often good in very small quantities – and in some cases even in large quantities
But the more you add, the more it will change the character of your sound. If you have a really nice acoustic guitar, you might want to record it as faithfully as possible!*** One other kind of distortion is adding “noise” of different kinds. It if happens at frequencies that were not there originally (or even at certain harmonics, usually un-even harmonics), it does not sound nice.
*** A third kind of distortion is also important to understand. If you have to frequencies at the same time that are not related, you might get inter-modulation or cross-modulation. TO make a long story short, it means that you suddenly add a frequency that was never there originally, and is not in some “nice” relation to the original frequencies. This is almost always bad! If you have, say, a 2 kHz tone and a 2.3 kHz tone, they may cross-modulate, and you then suddenly start getting a 300 Hz tone as well.
You ear usually does this for you, btw... and is one part of how the ear detects low frequencies. If a deep bass at, say 30 Hz, is present, there will be also 60 Hz, 90 Hz, 120 Hz, 150 Hz, etc present. Now, if even if you remove the 30 Hz completely, the ear will think it heard the 30 Hz signal, if it detects 60 Hz, 90 Hz, 120 Hz, 150 Hz, etc. This can be used to “cheat” the aural system to hear deeper sounds than what the speakers really can produce. And that is why a “straight” bass tone in some cases may be less interesting than one with a lot of overtones (which you may get by recording through a tube amp).
*** Still another factor that is not fully described by the frequency response is resonance. Resonance will certainly show up as a peak in the frequency response, but it will also make its presence known by adding certain frequencies even when none of those were present in the original sound. Think about a drum, for one second. When you hit it, you hear its resonance. But a drum also sounds when you hit other drums nearby it, or if the bass player hits an annoying note that makes your entire kit rattle... Same can be heard with a microphone. The MK-219 is one famous example of an annoying resonance, I've been told.
Hope some of you found this interesting...
-- Per.
D
DJL
Self Banned
baekgaard, besides the cheap budget 797 Audio/Studio Projects mic's... what other mic's do you have that you tested.
D
DJL
Self Banned
When I made my comments I was talking about #1... the NT1 with original capsule.bleyrad said:
B
baekgaard
New member
DJL said:
Recently, I've only been testing the 797/SP mics mentioned, the Oktava's and some Behringer ECM-8000 (for reference).
In my younger days, I worked for B&K and was at some point involved in simulations and measurements when designing the 40xx series mics for studio purposes -- but that was many years ago
-- Per.