Looking for some EQ'ing tips

[WA2SQQ]

I’ve come to realize that there is quite a bit of talent and knowledge hanging out in these forums. While some might be ashamed to admit, “I don’t know”, I believe that the only dumb questions are those we don’t ask.

I’ve been playing around with doing some EQ’ing (on a ribbon mic) and discovering that slightly attenuating and /or boosting certain frequencies can do much more than making broad adjustments. I’d be very interested in receiving any pointers or links to info on the art of EQ’ing a mic. I'm having some difficulty identifying which specific frequency points are associated with specific nuances of the human voice. I once read that EQ’ing a mic is similar to spices and women’s cosmetic products – “less can be more” and neither should dominate the dish! Thanks
Bob

[N1BCG]

You sure picked a subjective topic, and as such, the results of equalization will be more about cause and effect rather than right and wrong. Whenever I'm asked "How do I sound", my response is always "How do you want to sound?"

Some operators want that "big voice sound" with lots of low end even though their voices may not have a baritone quality. Others want piercing audio (fingernails on a blackboard) that can cut through pileups. A lot also depends on what the transmitter design will actually pass. Presence frequencies (about 3-5 kHz) contribute significantly to intelligibility, and most bandwidth conscious transmitters have no problem with 300-3000 Hz, which is considered communications quality.

But, try cramming massive amounts of bass through undersized modulator components and the audio becomes muddy and distorted. ALCs and DSP circuits often react unfavorably to such frequencies and will either filter them out and/or reduce the modulation level. The same can be said for trying to achieve a Hi-Fi sound by boosting high frequencies.

Transmitter limitations aside, ribbon mics, like all mics, can sound amazing if they are in good condition and positioned properly. By amazing I mean that as transducers they faithfully convert a wide frequency range of sound pressures into electrical energy. This will result in a very natural sound that shouldn't need much colorization unless you desire a particular sound.

Rag-chewers who enjoy casual QSOs on AM often have different objectives than sideband contesters who frequently battle challenging conditions.

All that said, how do you *want* to sound?

Clark

[WA2SQQ]

Ah, yes! I recall hearing that conversation on 75 recently. I realize that there aren't any right or wrong settings, but there are a few narrow parts of the audio spectrum that can make a mic sound "boxy" or "muddy". I'm just trying to identify those pressure points to play with. Than again, we don't hear ourselves the way others do - may some "bone cancellation" tweaks?

The other night I discovered two local marine channels that provide an automated radio check. You transmit and it send back a short transmission of how you sound. Maybe we need that on 75m?? Hmm ... Your points are very well taken. Thanks

[N1BCG]

I believe the automated marine radio aircheckers were set up by SeaTow. You can check your sound by selecting an online SDR station (www.websdr.org) that can receive you well depending on propogation. Some offer the ability to record your audio so that you can playback a brief exchange with another station to hear how you compare. Very handy!

Generally speaking, the "muddy frequencies" are in the 120-240 Hz range. Ironically, this is also the range that most frequently gets boosted. Emphasizing the presence frequencies (3-5 kHz) will go a long way toward adding crispness and clarity, but they also use up more spectrum. Best to be used with care, particularly on crowded bands.

It's important to point out that all this only affects the sound of the transmission. If everyone is listening on limited bandwidth receivers with the filters switched on, your finest efforts toward quality audio will be merely for your own amusement. Only those with wideband receivers and decent speakers will appreciate quality audio.

Keep in mind that amateur radio manufacturers put their efforts toward designing equipment for effective DXing, where fidelity doesn't play much of a role.

[WD5JKO]

Always a good read from the one and only W2WLR, George Bonadio:

http://www.amwindow.org/tech/pdf/eam.pdf

Jim
Wd5JKO

[W9ZSL]

Here is a bit of a guide-line.  Orban Parasound made a processor for AM broadcast that broke the audio spectrum down into 6 frequencies which were compressed / processed separately.  The frequencies were 100, 450 and 700 Hz plus 1.6, 3.7 and 6.5 KHz.  In addition to being a Ham Op, I've done professional studio engineering since 1965.  I also was a DJ since the early 60's at the same time.  EQ is a funny thing.  A little goes a long way.  Since music is not a consideration and the voice is, consider the frequency range of normal speech.  Anything above 6 KHz is not as important as those between 200 to 300 Hz up to 6 K.

One mistake many people make is BOOSTING certain frequencies rather than ROLLING OFF adjacent frequencies to tailor the response you are looking for.  A good example is EQ where bass guitar and kick drum is concerned.  With guitar, if selected frequencies below around 150 Hz are rolled off, you're left with the punch of not only that frequency but depending on the band-width of the equalizer per frequency, you retain frequencies above and below 150.  If you had a parametric or semi-parametric EQ, you can actually dial in the amount of bandwidth in addition to the frequency.  Those things are great because you can select a frequency and tune the response by increasing or decreasing the band-width around the fundamental.

A graphic EQ is fine if you want to emphasize or  de-emphasize a particular frequency.  An application would be getting rid of a 60 Hz hum by rolling off that frequency in particular without effecting adjacent frequencies since graphics can be quite narrow-banded.  I've used them when doing commercials that call for an effect like a telephone which has a relatively narrow frequency response that emphasizes the most effective range of the human voice.

So, in a nutshell.  Depending on your equalizer, say it is a graphic, dump everything above 6.5 KHz for starters.  Next dump everything below 300 Hz and begin tuning from there.  By "dump" I mean roll them off completely.  That gives you 300 Hz to 6.5 KHz bandwidth to play with.  Leave a frequency around 3.7 KHz alone and then start rolling off frequencies either side of that bench mark.  Do it very gradually; a db or so at a time.  Once you get that dialed in, you can always boost adjacent frequencies until you find a sweet spot.  Bottom line: the natural instinct is to boost frequencies when professional audio engineers do just the opposite.  That "sweet spot" should be neither boosted or rolled off but left in the "flat" position.

[K1JJ]

Hi Bob,

Good info.

Here's some more observations from my Bizzaro Whirl viewpoint:

Assuming we start with a clean, flat transmitter with transparent audio, I've found that the difficulty and time required to get a "desirable" sound when EQing is inversely proportional to the resonance and broadcash quality of our voice. IE, if you have a great voice for radio, almost anything (even flat settings) can be made to sound excellent in short order.

However, if we have an "average" voice (like most of us do by default) then to approach the resonant broadcash sound will take much expertise and effort.  It will be more challenging if we have mid-bassy voices  or have higher pitched ranges with no low end. We are basically stuck with what we have.  Most of us can use some "make-up" like a hot babe, no? What ham doesn't want to sound his best on the air?  It was once suggested to me that I should consider a larynx transplant if I wanted to sound any better.

Another challenge is that voices change over time, day to day and even year to year. I've found that great voices  usually can tolerate a wider range of settings - more stable - settings can remain fixed for longer periods of time. However, an average voice that has been electronically "optimized" to sound great will fall out of this band more quickly and require resetting the controls to become optimized again.  (EQ, processing, etc)   An extreme example is how we sound for a few days with a deep sinus and chest cold. We sound like a million bux! It's amazing how we sometimes sound the most resonant when we are sick...

So, bottom line is to first decide what kind of voice we have. This is easy - just axe friends with good ears that you know. You probably know the answer already. Then get a good monitoring system, voice recorder, etc and start playing with your audio settings until you get the sound you like.  Keep on top of it periodically since your voice "drifts" over time. For example your voice will sound much different from age 22 to age 70.

Only a small part of the people voice bell curve have the broadcash sound. It must always be this way. We tend to admire things that the majority do not have, like good looks, muscular builds, blond hair on women, etc. Same for voices.

BTW, in this day and age, there are many top announcers on radio and TV with average voices. But when a true-broadcash super voice is heard, we all still go crazy in disbelief.  Remember the Ted Williams Homeless episode?
https://www.youtube.com/watch?v=6rPFvLUWkzs

T

[W9ZSL]

T, so true!  I've had bronchitis for over a month.  My voice dropped in low frequency.  I never had one of those "Voices of God" but had hair to match at times.  Yet I sang tenor, then baritone, then bass and now I'm stuck between tenor and baritone.  When I sing in church I shift octaves all the time!  I was big in chorus and solo.  I'm 69...that's age...not sexual preference!

The main advice I can give is less is more. Unless you are dealing with broadcast quality AM, the number of variables in Ham Radio AM transmission is a total crap shoot. I worked for one station in Minneapolis as Production Manager and the Operations Manager worked with me to EQ my voice to the best advantage.  Of course, he made me change my name because the station already had someone named Mike.  Hated that.

Here is one idea.  Radio Shack actually sold a cheap graphic EQ with a built-in spectrum analyzer.  It had a LED scale in the center which would show the spikes of the frequency input.  That allows me to back off on that particular frequency either before the graphic or inline to flatten my overall response curve.  If you could score one of those, it would show you the bandwidth of your voice and where it has a primary resonance.

Don't go nuts trying to get a perfect sound with EQ.  It will drive you crazy.  If you don't have the "Voice of God" to begin with, settle for the hair!

[WA2SQQ]

I actually have a Behringer DEQ2496, so the spectral display is something I already have. I'll be playing with the ribbon mic this weekend. My Heil PR-40 already sounds just fine, so this "project" is nothing more than seeing if I can broaden out the response of the ribbon mic. The nulls of the sides of the ribbon are quite sharp and I'm hoping to be able to use that to null out some of the laundry room QRM!

Thanks to all

[W9ZSL]

Another thing you can do is Google that mic and see if there is a response graph out there.  That way you can get a better idea of the frequencies to concentrate on.