Pickup Frequency Response Curves

Bob Kagy · Post by **Bob Kagy** » 20 May 2005 12:26 pm

Remember the frequency response curves hi-fi speaker manufacturers used to put out? I've often wondered whether there are comparable graphs for steel guitar pickups.

There's a good thread over in Pedal Steel on measuring pickup impedances and it led me to wonder about this again. In spite of all the variables involved, it would seem like one way to get a handle on a which pickups might look like good choices, rather than having to guess.

Maybe this is a pretty dumb question, but with the occasional discussion on humbucking pickups' unwanted characteristics that need to be EQ'ed out, and individual preferences for sound, I wonder if there's any info on this out there.

BK

Jack Stoner · Post by **Jack Stoner** » 20 May 2005 1:34 pm

The only one that MIGHT have that is Bill Lawrence. Check Bill's site and read some of his (deep) dissertations on pickup design (or at least he used to have some of that on his site). Or call him but be prepared to listen to him for the "duration"..
http://www.billlawrence.com/

Jennings Ward · Post by **Jennings Ward** » 20 May 2005 5:12 pm

THAT IS A VERY GOOD IDEA: BUD CARTER OF CARTER GUITARS MAY BE ABLE TO SHED SOME LITE ON THE SUBJECT...FROM MY EXPERIENCE, THERE IS NO STANDARD REFFERENC AS SUCH,,,,,EACH PICK UP IS AN ENITY UNTO ITSELF, DUE TO SO MANY VARABLES AND THE FREQUENCY RANGE IT MUST BE CAPABLE OF REPRODUCING, AND WATTAGE OUTPUT.....I WISH I KNEW OF A " STANDARD" JUST FOR COMPAIRSON, IS WHAT THE THREAD IS ALL ABOUT...ALMOST GOT LOST.....IF ANYONE HAS ANY ELECTRONIC INFORMATION ON THIS SUBJECT , PLEAS INFORM ME... USUALLY THIS TYPE OF TRANSDUCER IS DESIGNED TO DO A SPECIFIC JOB WITHAN CERTAIN PARAMATERS, WHICH BOILS DOWN TO COST, BEING THE BOTTOM LINE.........JENNINGS

------------------
EMMONS D10 10-10 profex 2 deltafex ne1000 pv1000, pv 31 bd eq, +

Dave Boothroyd · Post by **Dave Boothroyd** » 22 May 2005 12:22 am

I read this and started to think how you would do the test.
It's easy with a speaker. The input source is a simple signal generator that sweeps through the audio frequencies and the test simply measures the power of the output.
The input to a pickup is the mechanical movement of the string, and a string never produces a single frequency (except maybe just before it stops sounding).
You would need a test rig that vibrated a steel wire above the pole piece in a typical elliptical path at a range of frequencies from say 20Hz to 25,000Hz.
Because strings are in different thicknesses and different distances from the magnets, you would need to be able to control for that too.
I suppose you could try energising the pickup with a coil supplied with pink noise and reading the pickup output with a spectrum analyser, but you would have to do it knowing what the frequency response curve of the eneergising coil was, and correct the readings for that.
All in all, it would be a bit of a nightmare.
I can see why it is not done.

------------------
Cheers!
Dave

Bob Kagy · Post by **Bob Kagy** » 22 May 2005 5:58 am

Jack, I always appreciate your comments. I checked out Bill L's website. Lots of good info.

Jennings thanks. The Carter website goes a long way with their recordings of the same song with different pickups, but I've always thought my PC speakers weren't allowing the full story to come through.

Dave, I was afraid somebody was going to point that out. I tried to imagine a test setup to get a frequency graph and got into a similar mental block of obstacles and complications.

T. C. Furlong · Post by **T. C. Furlong** » 22 May 2005 6:49 am

Let me start by first admitting that I am a total measurement nerd. Like Dave, I too got to thinking about how would generate a source for measuring a pick up. With loudspeakers, the best way to measure frequency response is by using pink noise and looking at something called the transfer function. A transfer function measurement is the difference between the original (or reference) electrical source and the measured signal (microphone's output). In audio, we look at the transfer function at many points in the signal chain. So then I was thinking about Bob's original question... "What unwanted characteristics with a humbucker need to be EQ'd out?" I assume that this question could be completed "compared to a single coil" If that is what we'd like to know, we can just make a transfer function measurement between a single coil pick up and a humbucker. That should reveal the bumps and lumps (or differences)in the frequency response of a humbucker compared to a single coil. It is important to note that frequency response is not the only thing to consider. EQ cannot fix phase response issues (sound arriving slightly later at different frequencies). In addition, there is a relatively new consideration that I have been researching. I call it "circuit speed" I believe this may have a significant effect on the listener's perception of audio quality.

It might be difficult to mount a humbucker and a single coil both at the best position relative to the vibrating string, but I would guess that a swap of positions could produce data that could be averaged or inverted or something.

TC

db · Post by db » 22 May 2005 7:22 am

I would think that there would have to multiple curves.
"Base line"-"nominal-input" curve.
"Transient response"-"attack-impulse" curve.
"Distance"-"drop-off" curve.
And redundant curves for different "load impedances".
But, the big question is the source of the "simulated input".
There would have to be a universally accepted standard here.

------------------
Dan Balde
U-12/8&5, S-7/D 3&1, S-6/E,A & G3

Paul Arntson · Post by **Paul Arntson** » 22 May 2005 8:24 am

.[This message was edited by Paul Arntson on 24 May 2005 at 08:05 PM.]

db · Post by db » 22 May 2005 12:45 pm

That's what "Hall-Effect" devices are designed to do.
They are used for gear/shaft rotation counting.
The "square wave" idea may help with the "Transient-Attack"-"Impulse" analysis.
There may be "Standard" test-equipment and procedures for the evaluation
of these types of devices, that could also work here.
The other thing is that all manufacturers would have to adopt this same procedure.
Or someone would have to go out and purchase every pick-up to perform this same test on!
And what would be the incentive to do this!
------------------
Dan Balde
U-12/8&5, S-7/D 3&1, S-6/E,A & G3[This message was edited by db on 22 May 2005 at 01:50 PM.]

Paul Arntson · Post by **Paul Arntson** » 22 May 2005 4:39 pm

.[This message was edited by Paul Arntson on 24 May 2005 at 08:05 PM.]

jerry wallace · Post by **jerry wallace** » 23 May 2005 7:42 am

Gentlemen, I agree its not very practical to do frequency response curves on pickups for all of the reasons pointed out..

The bottom line is "how does it sound when I play it"..This is partly why I have a policy of try it, if your not satisfied return it for a refund less shipping.

------------------
Jerry Wallace/TrueTone pickups-2001 Zum: D-10,8+6, "98 Zum: D-10,8+8,Nashville 1000,Session 500 ,Session 400 head only amp,Tubefex,ProfexII, Artesia, New Mexico
http://www.jerrywallacemusic.com

Donny Hinson · Post by **Donny Hinson** » 23 May 2005 9:48 am

We can adjust for frequency response in any number of ways, but the best way, and the easiest way, is to turn the knobs on the amp. Equalization can be done with the pickup, the guitar tone control, the volume pedal, the amp tone controls, the speaker, the cabinet, the cords, and lastly...the hands. I find I get the easiest changes, and the most variance using the hands and the amp tone controls.

Your mileage may vary, especially if you have a crappy amp.

Vern Wall · Post by **Vern Wall** » 23 May 2005 7:38 pm

I have been studying this topic and found a number of interesting facts. One is that guitars use magnetic pickups almost exclusively. Violins use magnetic pickups only rarely, piezo being preferred. Guitar pickups are mounted under the strings, but violin pickups are mounted in the bridge, on the bridge, under the bridge, in the sound post, and a few other places.

The reason for this is also interesting. Violin players like the natural sound of their instrument, and guitar players don't. That explains all the whizbangers that guitar players use. They don't want guitar sound, they want to synthesize something else. Another important fact is that technical specifications have very little to do with the actual sound as percieved by the human ear. Everyone acknowledges that there is nothing quite as subjective as percieved sound.

So response curves are only a vague guide to whether you will like a product. Speakers, for instance, have lousy response curves. People buy them because they like the sound. And in the end, the sound you hear is what it's all about.

db · Post by db » 23 May 2005 8:23 pm

With a "data base" of "response curves" . . .
One could spot pick-ups that they "already have" or have heard . . .
And could see how they "fall" in the over-all "curve".
And would be able to make an educated selection towards their desired goal!
"Brighter, Darker, thicker . . . Whatever!"
One would never be selecting "dumb"!
They would be able to see their "favorites" and where they "show-up"
and see which way to go to get closer to their desired tone characteristics.
------------------
Dan Balde
U-12/8&5, S-7/D 3&1, S-6/E,A & G3[This message was edited by db on 23 May 2005 at 09:36 PM.]

Bob Kagy · Post by **Bob Kagy** » 24 May 2005 9:45 am

Dan B., that's right where I was coming from in starting the thread - you expressed it much better than I did.

I remember when I had a Sho-Bud Pro II and switched my stock single coil tapped pickups out for some that were the latest local buzz, I was disappointed and wondered what was wrong with me & stuck with them for a long time anyway trying to figure out how to get a good sound - same hands, guitar, amp, cables, strings, picks bar, etc. Never did.

In spite of that I really agree that the hands amp and eq are all major factors, but having said that, in my case, not a particularly good player, for me they're not the whole story.

I've learned a lot from all the answers and pretty much understand that there won't be a pickup response data base anytime soon, if ever. Thanks everybody that contributed.

Violins - that's a whole nother thread, wow.

Jerry W., probably within a year . . .

BK

Vernon Hester · Post by **Vernon Hester** » 24 May 2005 11:00 am

Fellas go to this site and it explains all you need to know about pickup frequency response and how to measure.
Vern http://www.buildyourguitar.com/resources/lemme/

db · Post by db » 24 May 2005 6:19 pm

YES! Vernon,
That's the ticket!
So, now we need somebody to buy this thing and get a deal with every pick-up manufacturer to be loaned a pick-up for evaluation.
Is this a "pipe-dream" or do-able?
I would think that MFGS would like the oportunity to have their PU evaluated
and the specs' available to potential buyers.
------------------
Dan Balde
U-12/8&5, S-7/D 3&1, S-6/E,A & G3[This message was edited by db on 24 May 2005 at 07:22 PM.]

Paul Arntson · Post by **Paul Arntson** » 24 May 2005 6:57 pm

Good find, Vernon!

Vernon Hester · Post by **Vernon Hester** » 25 May 2005 2:17 am

There is another site where some university Students did about the same test, if I find it again I will post the url.
We did test on Mikes,speakers and pickups when I was working. We had all the goodies(test equipment)in our maintainence shop.(Tv network).
Vern

Bob Kagy · Post by **Bob Kagy** » 25 May 2005 9:45 am

Good article Vernon, thanks.

It sounds like it might be a pretty decent way to measure the single coils, but I think it said something about the humbuckers being a different situation?

BK

Randy Beavers · Post by **Randy Beavers** » 25 May 2005 10:20 am

This will be like the "Pro-Tools" engineer in the studio. When asked how something sounds his reply is, "It looks O-K to me."

Actually, I'd be interested in seeing a graph with that curve on it.[This message was edited by Randy Beavers on 25 May 2005 at 11:21 AM.]

Donny Hinson · Post by **Donny Hinson** » 25 May 2005 1:07 pm

Actually Dan, the response differences in similarly designed pickups (by that I mean that a single coil compared to another single coil, or one humbucker compared to another humbucker) are really insignificant compared with variations available elsewhere in the sound chain. Most players don't believe it until I show them, but hearing is believing! Only then do they realize that the differences between a BL-710 and a BL-910 are really small, as are the differences between an E-66 and the BL-710. A one number change in the amps mid-range setting does far more, tonally, than changing to a different pickup. I can prove this because my guitar allows me to swap pickups in about 10 seconds, and it's a real eye-opener!

IMHO, the only really significant pickup change you can make is going from single-coils to humbuckers, or vice-versa, unless you make a really drastic change in the impedance.

So, how significant is a change from, say, an 18k pickup to a 20k pickup? No one on God's green earth could even tell the difference in a live situation.

Believe it!

[This message was edited by Donny Hinson on 25 May 2005 at 02:08 PM.]

Vernon Hester · Post by **Vernon Hester** » 26 May 2005 5:44 am

Donny,
You are so right!!! I just went thru some different pickups on my old D-8 Alum Rickenbacker. Varied the DC res over a 5K range. Magnets make the greatest change in tone and gain.Also I have 23K 710's in my guitar, put some new one's in my Brothers Mullen ,19.5K. Could not hear any difference.
Vern

ed packard · Post by **ed packard** » 27 May 2005 1:09 pm

For those that would like a graph (response curve) for their pickups, and even further into the signal chain, You can do it yourself. If you are eading this, you have a computer and no doubt a sound card. There is a software called TRUE RTA that is a frequency spectrum analyzer and O'scope combined. The poor man's version downloads for free. This will let you see the frequency response of your pickup(s)on your instrument, and on up the signal chain.

If you have a recording software with the spectrum analyzer, you can probably make that work for you.