Using Har-Bal for mic techniques, mic emulation, PA

[brubart]

Here is a suggested new use of Har-Bal: Make a short recording of a musical instrument or vocal in various microphone positions. Compare the spectra of those positions. This shows the tonal effect (almost like EQ) of various mic techniques.

You could even select a reference mic position where the instrument or vocal sounds most accurate -- say, 2 or 3 feet from an acoustic guitar. Import that recording's wave file into Har-Bal as a reference. Then place the same mic in various close positions (near the 12th fret, near the bridge, near the sound hole) and make short recordings. Import the wave file of each mic position into Har-Bal, and you will see the tonal effects of close microphone placements.

You could even equalize a close-miked recording to sound like the more-accurate distant reference recording. So, when leakage forces you to mike in close, you might be able to equalize the recording to sound more natural, like the distant reference mic.

I did some research on this topic back in 1981. For those who are interested, the research appeared in the Journal of the Audio Engineering Society:

Tonal Effects of Close Microphone Placement
Volume 29 Number 10 pp. 726-738; October 1981

Abstract: The effect of microphone placement on reproduced tonal balance is investigated for acoustic guitar, piano, electric guitar amplifier, and voice. The data presented are based on spectral measurements in various microphone locations, listening tests, and a study of instrument radiation patterns. Suggestions are offered for close microphone placements and equalization to achieve various tonal effects, including the natural timbre of the instrument as heard at a typical listening position.

Author: Bartlett, Bruce A.
E-lib Location: (CD aes4) /jrnl7888/1981/7815.pdf

The same article is Preprint Number 1782, Convention 69 (April 1981).

I hope this new application for Har-Bal leads to a better understanding of microphone techniques.

Another new application: Use Har-Bal as a microphone emulator. Record a voice or instrument with a reference microphone (such as an expensive Neumann mic), and record it again in the same spot with a mic of your choice. Using Har-Bal, EQ the mic of your choice to sound like the reference mic. It won't be a perfect match, but it may be an improvement. You can use this technique to make a lavalier mic sound like a high-quality mic one foot in front of the mouth, and so on.

Maybe Har-Bal could be used in live sound to get a decent-sounding mix. Use a graphic EQ (or console EQ) to make the house spectrum closer to some ideal reference curve. If you couldn't use Har-Bal itself in this application, you could at least use its concept of spectral matching.

Hope this helps,

Bruce Bartlett

[har-bal]

Here is a suggested new use of Har-Bal: Make a short recording of a musical instrument or vocal in various microphone positions. Compare the spectra of those positions. This shows the tonal effect (almost like EQ) of various mic techniques.

You could even select a reference mic position where the instrument or vocal sounds most accurate -- say, 2 or 3 feet from an acoustic guitar. Import that recording's wave file into Har-Bal as a reference. Then place the same mic in various close positions (near the 12th fret, near the bridge, near the sound hole) and make short recordings. Import the wave file of each mic position into Har-Bal, and you will see the tonal effects of close microphone placements.

You could even equalize a close-miked recording to sound like the more-accurate distant reference recording. So, when leakage forces you to mike in close, you might be able to equalize the recording to sound more natural, like the distant reference mic.

I did some research on this topic back in 1981. For those who are interested, the research appeared in the Journal of the Audio Engineering Society:

Tonal Effects of Close Microphone Placement
Volume 29 Number 10 pp. 726-738; October 1981

Abstract: The effect of microphone placement on reproduced tonal balance is investigated for acoustic guitar, piano, electric guitar amplifier, and voice. The data presented are based on spectral measurements in various microphone locations, listening tests, and a study of instrument radiation patterns. Suggestions are offered for close microphone placements and equalization to achieve various tonal effects, including the natural timbre of the instrument as heard at a typical listening position.

Author: Bartlett, Bruce A.
E-lib Location: (CD aes4) /jrnl7888/1981/7815.pdf

The same article is Preprint Number 1782, Convention 69 (April 1981).

I hope this new application for Har-Bal leads to a better understanding of microphone techniques.

Another new application: Use Har-Bal as a microphone emulator. Record a voice or instrument with a reference microphone (such as an expensive Neumann mic), and record it again in the same spot with a mic of your choice. Using Har-Bal, EQ the mic of your choice to sound like the reference mic. It won't be a perfect match, but it may be an improvement. You can use this technique to make a lavalier mic sound like a high-quality mic one foot in front of the mouth, and so on.

Maybe Har-Bal could be used in live sound to get a decent-sounding mix. Use a graphic EQ (or console EQ) to make the house spectrum closer to some ideal reference curve. If you couldn't use Har-Bal itself in this application, you could at least use its concept of spectral matching.

Hope this helps,

Bruce Bartlett

Is this the Bruce Bartlett who is the author of the book "Practical Recording Techniques" as well as others? If it is we all humbly bow down to you. You have a tremendous history in the recording field and might I add what a treat this is for those folks who don't know who you are.

Welcome to our Har-Bal family and thank you for those fantastic tips! We are truly honored.

Cheers

Earle

[HarBal]

Bruce,

I love your ideas on emulating/correcting for mic placement. They're brilliant! Natural sounding acoustic recordings done with distant mic'ing typically require pretty expensive Mics to avoid having your music buried in Mic self noise. With close Mic'ing you can get around the noise problem of cheaper mics but you end up with altered timbre. Your suggestion gives the best of both worlds.

I'll have to read your paper. It sounds like it'll be filled with great information.

Thanks!


Paavo.

[brubart]

Hi Earle and Paavo,

Thanks for the kind words! I am mentioning Har-Bal several times in Practical Recording Techniques 4th Ed. which is due out in September.

I'm going to explore mic techniques with Har-Bal as a research tool. When you want to achieve a certain sound on an instrument, EQ is not a substitute for optimum mic placement, but sometimes EQ is the only option

Har-Bal is the first piece of software in a long time that has made me rethink what I'm doing. I often suggest that it's a good idea to reference your mixes against commercial CDs, and Har-Bal makes it so much easier.

Also, matching levels of songs in an album by ear can be a long, tedious process, and Har-Bal helps streamline that as well.

After using Har-Bal for a few days now, I fully agree with your advice not to match the reference curve exactly. You still need to apply musical judgement after Har-Bal gets you in the ballpark.

I was trying to figure out why the particular reference curves you supply tend to result in mixes that sound good on a wide variety of playback systems. I think it's because the ear can tolerate small amounts of tonal coloration on a particular monitor speaker, but other speakers with different frequency response can exaggerate the colorations if those speakers have response peaks or dips in the same area, making them more audible.

Since Har-Bal helps recording engineers create mixes with spectra that are free of obvious peaks or dips, hearing those mixes on several different speakers with different frequency responses does not result in additive coloration. I hope I'm explaining this okay.

It's interesting that the general shape of the reference spectra is something like the recommended house curve for PA systems when playing pink noise. The house curve (flat up to 1 or 2 kHz, with a smooth rolloff above that) results in subjectively "flat" response. But a flat pink-noise spectrum, as picked up by a flat free-field microphone in the audience area, sounds very bright.

Harmonics of musical instruments have a natural fall-off with frequency, so I guess that's why the spectra of musical groups heard live has the same characteristic. In our mixes we are aiming to duplicate that natural HF rolloff.

The Journal of the Audio Engineering Society has some papers on the measured spectra of various genres of music. I'll have to do a search on that.

Thanks for a great product!

Bruce Bartlett

[HarBal]

Hi Bruce,

I think your explanation on why having a "relatively" smooth spectrum helps in transportability is spot on. As a "past" home loudspeaker designer-constructor the additive colouration of loudspeakers was something that always confused me and had me continually tweak my crossovers to try an make them sound good with all material. But invariably, and to my frustration, I found that the tweaked crossovers only worked for some and not all recordings. Post "Har-Bal" I see things a bit differently.

The pink noise issue with PA tuning is something that has puzzled me too. On reflection of some of things you said concerning mic placement and the difference between distant and near field mic'ing, I think you'll find an answer. All acoustic instruments have a polar radiation reponse that is pretty directional at the HF end but uni-directional at the LF end. Naturally then, when you listen to instruments played in rooms and concert halls you'll hear more sound at the LF end than the HF end because of that polar radiation response. When you close Mic everything you are bypassing the tonal modification of the ambient field and the Mic pics up much more HF energy than it would in the far field. If you then play that back though a PA set up with flat pink noise response it seems logical that it will sound bright. Probably even more so if the polar response of the loudspeakers is not great at the HF end, cos the EQ applied to the PA would probably have to be quite extreme to compensate.

That's the way I see it though I'm not sure my reasoning is entirely correct. A few ideas for you to ponder. Look forward to seeing your new edition.

Paavo.