Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Russell Ault]

Anyway, as it occurs to me, with my limited knowledge on the subject, if a microphone's frequency range reasonably covers the original intended content, I'm not too worried about the response curve.  Should I be?  Seems that durability/longevity and noise rejection in the null zones (I think I made that term up, but you get the point) in the polar pattern would be the prime reasons to choose one mic over another.

Let me try a slightly different angle on what others have said: how many performers have you worked with change microphone placement as part of their performance (or, perhaps, how many have you worked with that haven't)?

A microphone's frequency response is always a function of two things: distance, and angle.

Let's pretend to measure the frequency response of a microphone. You setup a pink noise signal generator and split the output. One half goes into a theoretically-perfect full-range point-source transducer (if you're Dave Rat you'd use an earbud of some kind), the other into your measurement rig's "reference" channel. The microphone we're going to measure then provides the "measurement" input to the test rig. Setup the microphone on a stand 1 meter away from the speaker (and as far away as possible from any possible sources of reflection) and take a transfer function measurement. This is your "frequency response curve". As we vary the distance between the microphone and the speaker, and the angle of the microphone relative to the source of the sound, the transfer function graph will start to change.

In the case of an omnidirectional mic, these changes tend to be fairly predictable: as the microphone gets closer to the speaker, the graph looks basically the same but there's an overall level increase. If you change the angle of the microphone the graph will start to get quieter, and you might notice some additional high-frequency attenuation (with the frequency of the additional attenuation dependant on the diameter of the microphone capsule), but other than that the graph will continue to look pretty similar.

Let's replace the omni with, say, an SM58. As the microphone gets closer to the speaker the graph starts to get evenly louder, but eventually you'll start to notice that the low end is increasing in level faster than the mids and the highs. This is often referred to as "proximity effect", and all directional microphones exhibit some (although some, like the KSM8, exhibit much less than others). Adjusting the angle of the microphone reveals another startling revelation: the graph starts to get quieter, but not at all frequencies at the same time. As you approach 90 degrees you'll notice that anything above 4k has been attenuated by almost 9dB, while the low end is only down about 4dB. If you point the back of the SM58 at the speaker, the low end will be down by about 10dB, the mids by 20dB or more, but the high end will have started to increase again, with 8k being only about 10dB down. In other words, an SM58 pointed directly at a singer's mouth with sound different than if it's pointed at a singer's nose or chin.

Now let's replace the SM58 with something very expensive and, say, classical-music-broadcast oriented. More than likely, the graph with the microphone angled at 90 degrees looks almost the same as the 0-degree graph, just 6dB lower. In other words, this microphone when pointed directly at a violin will sound about the same as if the same violin were off to the side, only that the side-violin will be half as loud by dBm.

The problem with "frequency response curves" is that at best they tell you something about how a microphone performs in a straight line, but most of the time that's not how they're actually being used. Even for omnidirectional microphones the frequency response at a 30 degree angle will be meaningfully different than at 0 degrees (insert conversation about "free-field vs. diffuse-field" here), and that's for microphones that aren't supposed to change with angle.

-Russ

[Scott Bolt]

Years ago, I did a detailed microphone shoot-out with my full band PA using Sennheiser e835s, EV ND767a's, SM58's, Beta 58'S and OM7's.

Several BE's worked the board where the mics were eq'ed to taste and the band played in a full sized venue with the PA (no audience).

We all walked away with slightly different views of the microphones, but mostly very similar views:

1)  No one cared for the Beta 58's.  No amount of eq could make them sound "good" compared to the other microphones.
2)  The SM58's were "OK" to everyone, and all present had no problem dialing in "good"; however, the SM58's dialed in for a "good" sound were the most feedback prone compared to the other microphones in the list (no one wanted to turn up the Beta 58's ).  The SM58 was also the most forgiving mic for bad singer technique (we were fortunate to have one lead singer that truly had horrible technique to demonstrate this fact ).
3)  Most people felt that the e835's provided better clarity than the SM58 and allowed the volume to be put up just a bit higher too.
4)  The ND767a's were the most favored most often, and had really good GBF characteristics.  The only downfall with this mic was that it requires singers that stay on the mic to be consistent .... or more compression than is wise in a live system (which just goes to show you that not every mic is good for every situation).
5)  The OM7 did in fact live up to its hype on getting the most gain before feedback; however, it also required the most tweaking in both the input gain trim (much higher than other mics), and the channel eq to get it dialed in.  Not everyone agreed with me on this mic, but I thought it sounded worse than the Beta 58 (although a different kind of bad).  Others thought it was OK.

You will find very few singers who wont sing on an SM58, but as you can see, there are better mics out there under different singers and different situations. 

I still have a complement of the now unavailable EV ND767a's.

I have also done a much less "real" comparison to the Shure SM86 (condenser vs dynamic mic).  I thought it was a very good sounding microphone, but did seem to get more stage noise pickup than the ND767a (but most mics do).

My advice is to try a few mics (borrow from other bands .... or do a trade for a night) and see what works best for you.  Forget the specs.  It is just going to make you pull your hair out

[John L Nobile]

Thanks John, I appreciate your input and conversation.

I'm not necessarily talking about an actual plotted graph that we can look at.  More so (attempting) to use term "response curve" as a way to quantify and/or communicate the differences in the way 2 mic's may sound.  Really, just for lack of a better term.  Seems logical to me that response curve, whether it be on paper or perceived through listening is a fair term to use.  Maybe there's a better word for it?

But, that brings up another question, if we had an accurate graph for a mic you liked, and an accurate graph for another mic, couldn't the 2nd mic be EQ'd to sound reasonably close to the one you like, using the graphs alone?

The bottom line for a mic is how it cuts in the mix. Can you hear the vocal clearly without fighting the channel all night. Beta 58A and Sm86 work for me. Senny is extremely popular but I'm in the Shure camp. Kinda like Ford vs Chevy mentality I guess.

[Russell Ault]

[...]
But, that brings up another question, if we had an accurate graph for a mic you liked, and an accurate graph for another mic, couldn't the 2nd mic be EQ'd to sound reasonably close to the one you like, using the graphs alone?

I missed reading this while I was writing my earlier response. For the reasons I mentioned, unless the microphone in question is only picking sound from exactly one direction at exactly one distance from the mic, then the practical answer is "no".

-Russ

[Chris Grimshaw]

On another forum, someone asked a similar question. The answer is this: there's more to a microphone than its on-axis response.


Proximity effect. Shape and severity. The EV RE20 is one of the best mics out there for not-having-a-proximity-effect. However, I'm not convinced that proximity effect is a bad thing. A good singer will use it as part of the performance - if they need to sing very low notes (which humans aren't particularly efficient at producing), then getting in close to the mic will let them EQ their own voice to give it the extra LF heft that's needed.
There's also the fact that, if you stay fairly close to the mic, the engineer can use some EQ to bring down the low end a bit, directly reducing the low-end spill from the PA system.
The peak frequency of the proximity effect will change mic-to-mic. Generally, condensers will show a rising response towards the very-low-frequencies, while dynamic mics will have their peak at the same frequency as the LF rolloff knee when used at a distance. Typically around 100-160Hz.
The LF difference between 12", 2" and 1" will also vary mic to mic. The RE20 might be +0.5dB at 2" and +1dB at 1" (yes, it doesn't perfectly eliminate proximity effect, but it does get very close) while the Sennheiser e935 might be +6dB at 2" and +12dB at 1".


The polar pattern is of great importance, in a couple of different ways. First up, there's the area which I'd call the "intended" pickup area, a cone of +/-45 degrees from on-axis. Ish. Maybe +/-60 degrees. Singers that play instruments (and those that are just clumsy with their mic technique) often address the mic at different angles, and I'd want the response to stay consistent within that intended pickup area. The EV N/D767a wasn't great for this - there's a layer of very dense foam in the headbasket, arranged so that the on-axis response is unaffected, but moving a little off-axis reduces the level quite quickly. I suspect that altering the layout of the foam might change that, but I haven't tested that yet. By contrast, the Sennheiser e935 mics that I moved to are very good in that regard. The response stays similar over a wide range of angles, which means guitarists etc can enjoy decent sound as they move around. My bread-and-butter events are (were, before the virus) smallish festivals, where I might have up to 10x acts in a day. 45 mins on, 15 min turnaround. A mic that's easy to get on with keeps the musicians happy.


The rest of the polar pattern (the rejection zone) is also important. Over those angles, I'd want the sound levels to drop away quickly and evenly (with regards to the frequency response). Aiming for outright "maximum rejection" can come at the cost of "even" rejection. ie, you might end up with a mic where unintended sources sound very coloured and un-natural.
Let's say you've got a nice gig where the singer is playing the piano. You might pick some nice mics (maybe a couple of 414s) and dial in a gorgeous piano sound. Then, you bring up the vocal fader and the piano sound turns to mush because the vocal mic's rear pickup has a crazy frequency response, and that directly screws up the piano sound. There's not a whole lot that can be done about that (although you might get lucky time-aligning the mics to one place or another) except changing the vocal mic for something better-behaved.


With apologies to Tim, the phase response of a microphone is directly tied to its frequency response. Microphones are minimum phase devices, so by definition a frequency response wriggle will show up in the phase curve. Similarly, a flat frequency response will result in a flat phase response. The only exception is when circuitry adds phase shifts. For instance, the AKG D224E has a passive crossover built-in. As we know from speaker design, passive crossovers = phase shifts, so the D224E will have a large shift in the 800Hz region, where the two capsules meet.


Distortion profile. Not a huge problem in most cases, but does show up occasionally. Some mics have components which are designed to add some harmonic distortion - valve-based preamps typically do (mostly by design - valves can have very low distortion, too, but then they'd just use transistors), and transformers can be designed to saturate when subject to strong LF signals.
You should also consider the clipping point of the microphone, but that certainly applies to condensers more than dynamic mics. A cheap Amazon mic might tap out at 110dB, which is fine for talking etc, but won't hold up if a powerful singer gets going. I typically make sure any mics to be used live can handle 140dB to make sure there isn't any unexpected distortion. After all, it won't show up on the meters at the desk.


The handling noise of a microphone should not be discounted. I've known some truly awful mics for that, but fortunately they weren't intended for hand-held vocal use. Even so, one of the differences between a cheap and expensive mic will be how much noise is made as it's handled.



There's probably more, but I think that's enough to be getting on with.

Chris

PS - From the above, it should be easy to see why EQing one mic won't always make it sound like another.

[frank kayser]

Hi Frank,

What qualities do the other mics have that has you like them more?

-Seth-

Hi Seth,
As I said, others can (and did!) address your questions much better than I...

I work with many female singer-songwriters where clarity and a smooth "response curve" is needed.  My go-to mic for these people is usually a Sennheiser 935.  Comparing the Senn 935 to the 835, I find the 935 a bit less harsh.  Both the 835 and the 935 seem to have less proximity effect than a SM58.

The Heil PR22 is my other go-to mic.  I find it similar to the Senn 935, but different.  I have not sorted that difference out in my mind yet (after over a year) but there is something.  In my opinion, they are close enough to use interchangeably in most ensemble performances.

I also bought a Heil PR35, and was expecting nothing but wonderfulness with this mic.  One woman, with whom I worked for over 10 years, who's voice is alto, rich, clear, clean, precise, borderline sultry, just did not sound right on the PR35.  I know her voice. I chalked it up to me being off my game.  I tried it with another woman, with whom I've worked nearly as long, who's voice is this high, clean, clear "Celtic" voice, and she also sounded less than optimum. Fought it all night.  I came to the realization that on the voices that were already clean and clear on the Senn935, I found the PR35 mic to be lacking.  I don't know what, but it was an immediate oops, never to be repeated.  Not for her, or her, either!

One specific singer who has a wonderful, but somewhat problematic voice to get right, normally travels with her own Neumann KMS104, which is a very good match for her voice. She has a voice which is slightly nasal, a bit thick in the middle, and doesn't give a good pronunciation at the end of her words. (wow. That was brutal!) When her voice is reproduced "right", her voice is absolutely mesmerizing!

Oh.  Back to mics...

I had worked with her a couple times by this point, and this performance, she forgot her mic, so I put the Heil PR35 on her for soundcheck expecting to swap it out for the Senn 935. I was encouraged by how it worked with her, and kept it on her for the performance.   Afterwards, she and her entourage (a couple friends and her recording engineer) said they loved the performance and her voice never sounded better, asking what mic I used.  For her voice, the PR35 was magic, equally as good as her Neumann.  A different singer liked her own voice with the PR35 so much, went out and bought her own.

I've worked with these four women using an SM58 too.  Their voices carried the day, but lacked just a bit of their normal sparkle and character.  An SM58 always works.

And for those singers and sound folks who can get a Beta58 to sound great, more power to them.  "nuff said.

John's quote below says it all...

The response curve will not tell you what it will sound like with different voices or instruments. If it comes down to graphs vs ears, the latter always wins.

[Erik Jerde]

Back in the day, when you were lucky to have 3 fixed bands of EQ on a channel, I get that a microphone could really make a difference.  But, in this day and age, with most digital mixers having 4 or more fully variable parametric bands of EQ available on each channel, is the microphones native response really that big of a game changer?

I'd much rather have a mic that takes no work on the console than one that I need piles of EQ on to make it sound right.  It makes the job immensely easier to not have to fight with a source to get it sounding right.  The less EQ the better.  Fix things at the source rather than trying to make it up later.  It's kinda like guitar players who are always working on "tone."  If the guitar to amp sounds like shit there's no way you're going to fix it with a pile of pedals.  Fix that simple setup, everything else is just band-aids.

[Kevin Maxwell]

The absolute minimum quality mic I would use on a singer would be an SM58. I used to use cardioid condenser mics on singer more often until a time I worked with a group that I realized I really didn’t want to hear the vocals that clearly. So I have switched back to using the SM58 more lately.

The one thing I really don’t like about the SM58 is if you have a singer that then talks (a Lot) and swallows it like they did when they sing the proximity effect that works so great for them when singing now is working against them for speaking clarity. It gets muddy sounding. And most people sing a lot louder then they talk so that compounds the issue. I never use an SM58 as a podium or announce mic if I can help it. If I am using a dynamic mic for those purposes I usually use a Shure Beta 57. Because it doesn’t have the extreme proximity effect like the SM58.

[Dave Pluke]

So, similar to performers being picky about "their tone", I get that we can get geeked out on microphones.

Having spent significant amounts of time at both ends of a microphone, I'll just add that, to a Singer, the most important consideration is how the mic sounds in their mix.  Sure, you can 4 band parametric the cr@p outta a mic at FOH but, if the monitor send is pre-EQ and the mic doesn't sound or respond the way the Singer expects, it'll be an issue.  Better to rely on a known commodity that doesn't require a lot of tweaking.

In general, life for FOH and Monitor Mixers is easier if all vocal mics are the same model.  Some Performers will have strong feelings that ruin that plan.  Unless Rider'd otherwise, I'll provide a set of SM58 (wired or wireless) or sE V7 (ditto) vocal mics.  I also keep a few "upscale" mics on hand for the second scenario.

 Dave

[Seth Udoll]

I'm learning a lot more from you guys than I thought I would.  Thanks, so much, for all the shared experience and knowledge. 

All the blocks in my mind are starting to fall in place.  I had a feeling I was missing part of the whole picture and the it's becoming more and more clear. 

Gonna let it sink in another day or so and kinda repeat back to you guys my understanding of all the various input.

All of you are so awesome!  What a great resource this forum is!  Really... your time and input is much appreciated.

Thank you.

-Seth-

[ProSoundWeb Community]