Topic: High Frequencies In Outdoor Recordings

[chris319]

I recently recorded a string ensemble in an outdoor venue. I tried eight different mics (I was shooting out mics that day). For some reason, all the recordings sound like they have the highs sucked right out of them. They sound wooden. They sound rolled off at 8 kHz, kind of like AM radio. I was seated in the front row, about eight to ten feet behind the conductor, with the mics about waist height (the height of an unextended floor stand).

The mics were:

CAD e70, omni capsule
MXL 991 with capacitor mod, omni capsule
MXL 991 unmodified, omni capsule
Studio Projects B-1
Heil Sound PR40
E-V RE27
E-V RE55
E-V 655C

Here is a clip from the Studio Projects B-1 recording, one of the least worst:

http://www.miclisteningroom.org/MSLMYSO/SP_B-1_short.wav

Why do my recordings sound so deficient in the highs? Is it the outdoor venue? The position of the mics? I have swept my system and I know it is flat to 20 kHz.

[boojum]

Waist height?  Sounds like a problem right there.  What was in front of the mics if anything.  Waist height is awfully low.  They should normally be 10 - 12' up and aimed into the group.  Could they have been facing backwards?

For a test mount one and walk all the way around it while the recorder is running to see how the sound changes as you move.

[raymonda]

I will take a listen and also look at the wave but if you were only recording strings I wouldn't expect much information beyong 8,000.

[raymonda]

The wave looks fine and overall it would have sounded fine but you have low frequency wind noise that causes a lot of problems. You should have used a wind screen or a more beefier one at least.

The primary frequency of strings do not extend that far. That is why you don't see much information up top.

Try your shoot out again but use screens or move it in doors.

[chris319]

For a test mount one and walk all the way around it while the recorder is running to see how the sound changes as you move.

You can't do that during a live performance in front of an audience

The wave looks fine and overall it would have sounded fine but you have low frequency wind noise that causes a lot of problems.

I know about the wind noise. Moving it indoors was DEFINITELY not an option. I'm talking about the sound of the instruments. Listen to the violins. There is no high-frequency detail.

[chris319]

if you're in the front row in a normal auditorium then you are below the orchestra

This was outdoors, not in an auditorium. The audience and musicians were on the same flat surface. Waist height means waist height to the conductor. Sorry, I should have been clearer.

the B1 is gonna drop high end detail over 13kHz or so by a function of being a large diaphragm mic

I did not know that. Now how do you explain the other seven mics? They don't sound any better. The B-1 was the only LDC among the eight. Must be the height.

[raymonda]

A violin's frequencies fall within 200 to 3.5khz. The table attached should help you out a bit on why there is little information above 6,000 hz on your recording.

Sure, getting your mics up about the instrument is the way to go. But what I hear as your primary problem is the accessive wind noise causing large low frequency rumble. Do you have full range speakers, or subs, that go down to 20 hz (-5dbs) or so. If not, it may explain why this is not obvious to you. But with woofers and speakers flapping like this it is hard for them to reproduce other frequencies reliably.

BTW, I wasn't suggesting that you ask the performers to move inside, but rather, try you rshoot out in at an indoor concert event.
Frequency Range of Vocals and Musical Instruments        Typical Sound Pressure Levels         

Vocal Woodwind Brass Strings Keyboards Percussion               

VOCAL   Approximate Frequency Range   Sound Pressure (N/m2)   Sound Pressure Level (dB)   Environment   Description
Soprano   250Hz - 1K   200   140   Military Aircraft Taking Off   Intolerable
Contralto   200Hz - 700Hz   63   130   Pneumatic Drill   Very Loud
Baritone   110Hz - 425Hz   20   120   Ships Engine Room   
Bass   80Hz - 350Hz   6.3   110   Factory   
WOODWIND   Top   6.3 X 10-1   90   Heavy Lorry (6m)   
Piccolo   630Hz - 5K   2 X 10-1   80   Busy Road (kerbside)   Loud
Flute   250Hz - 2.5K   2 X 10-2   60   Restaurant   Noisy
Oboe   250Hz - 1.5K   6.3 X 10-3   50   Conversational Speech   Quiet
Clarinet (B flat or A)   125Hz - 2K   2 X 10-3   40   Nightime (residential area)   
Clarinet (E flat)   200Hz - 2K   2  X 10-5   0   Normal Hearing Threshold   
Bass Clarinet   75Hz - 800Hz            
Basset Horn   90Hz - 1K   Reverberation Targets      
Cor Anglais   160Hz - 1K         
Bassoon   55Hz - 575Hz      Reverb time   Function
Double Bassoon   25Hz - 200Hz   Only Speech   0.6 - 1.2 secs   Lecture/Conference Halls
Soprano Saxophone   225Hz - 1K   Speakers (trained)   1.0 - 1.4 secs   Theatres
Alto Saxophone   125Hz - 900Hz   Reproduced Sound   0.8 - 1.2 secs   Cinemas
Tenor Saxophone   110Hz - 630Hz   Multi-purpose use   1.0 - 1.5 secs   Multi- purpose Halls
Baritone Saxophone   70Hz - 450Hz   Opera   1.0 - 1.6 secs   Opera Houses
Bass Saxophone   55Hz - 315Hz   Solo Instruments   1.2 - 1.6 secs   Recital Halls
BRASS   Top   Orchestras   1.6 - 2.2 secs   Concert Halls
Trumpet (C)   170Hz - 1K   Choirs/Organs   2.0 - 4.0 secs   Churches
Trumpet (F)   300Hz - 1K         
Alto Trombone   110Hz - 630Hz   Musical Range and Frequency       
Tenor Trombone   80Hz - 600Hz         
Bass Trombone   63Hz - 400Hz   (Piano Keyboard Equal Temperament)      
Tuba   45Hz - 375Hz         
Valve Horn   63Hz - 700Hz   Note   1   2   3   4   5   6   7   8
STRINGS   Top   A   28Hz   55Hz   110Hz   220Hz   440Hz   880Hz   1.760K   3.520K
Violin   200Hz - 3.5K   B   31Hz   62Hz   123Hz   247Hz   494Hz   987Hz   1.974K   3.949K
Viola   125Hz -1K   C   33Hz   65Hz   131Hz   262Hz   523Hz   1.47K   2.093K   4.186K
Cello   63Hz - 630Hz   D   37Hz   73Hz   147Hz   294Hz   587Hz   1.175K   2.350K   
Double Bass   40Hz - 200Hz   E   41Hz   82Hz   165Hz   330Hz   659Hz   1.318K   2.637K   
Guitar   80Hz - 630Hz   F   44Hz   87Hz   175Hz   349Hz   698Hz   1.397K   2.794K   
KEYBOARDS   Top   G   49Hz   98Hz   196Hz   392Hz   784Hz   1.568K   3.136K   
Piano   28Hz - 4.1K                           
Organ   20Hz - 7K                           
PERCUSSION   Top                           
Celeste   260Hz - 3.5K                           
Timpani   90Hz - 180Hz                           
Glockenspiel   63Hz - 180Hz                           
Xylophone   700Hz - 3.5K                           

[raymonda]

I love the comment about strings having no highs; try having three daughters learning violin and get back to me on that    Please consider the difference between fundamental tones and overtones--strings have a LOT of overtones).[/size]  So most of the highs are going over your head, especially if your mics are at waist level.

I believe I stated the primary frequency, I was not discussing overtones or upper harmonics, e.g 2nd, 3rd, 4th, or 5th, etc.......  Any bowed string instrument played by a novice or beginner can sound annoying. Relative to a paino a violin does not cover nearly rhe same range. Sorry, but that's just a fact.

[mfrench]

I regularly record a violin that aurally disappears at 14.5k, and is probably still making waves above where my cloth ears can hear them.

[raymonda]

I regularly record a violin that aurally disappears at 14.5k, and is probably still making waves above where my cloth ears can hear them.

Overtones my friend, overtones. It does not have any fundimentals up there. And, it should be relatively low in nature when copared to the primary note. 3,500 second harmonic is 7,000.... 3rd harmonic is 14,000 and 4th is 28,000.......etc...... If you are getting the db level as high as the fundimental note, there are some problems which will obscure the timbre.

[raymonda]

If you look at the frequency graph of he sample provided you will notice a 15-20 climb going down from 175 to 11hz. This is all wind noise muddling up things. Because of this, the conditions the recording was made under provides limited information regarding the mics and the shoot out.

[boojum]

Chris, pardon me for making a wild assumption.  I thought you might test the mics some other place or before the performance.  I guess I should have spelled that out for you.  Next time.

[raymonda]

raymonda, please, that table is really a joke.  Citing a fundamental frequency range of an instrument has 0% to do with the overtone series and therefore the high-frequency content of an instrument.  Many, many instruments have a LOUDER first overtone than fundamental.  Some of them have second, third, even fourth as strong as the fundamental.  The general rule of full-range recording is it tends to follow a "pink-noise" distribution with descending amplitude at increasing frequency.  There are many exceptions that rule, but the recording in question clearly shows a rolloff above 4kHz that exceeds the norm significantly.

Trust me, I have studied the overtone series of a violin.  To be blunt, you don't have any idea what you are talking about.  I can arrange a quick demonstration of many instruments here in my studio tonight that will prove this conclusively.


I would say that I have a bit of an idea of what I'm talking about and I stressed primary and fundimentals from the onset. I obviously have a good understanding of overtones and the effects it has on timbre and tone. Close mic'd in a studio will certainly provided a different frequency plot than far field, outdoors with an ensemble.

The decay of the note close mic creates very interesting changes. Pluck any string under a strobe an you can see the effects. I don't really see where we part thoughts, other than I think the recording is more effected by being recorded out doors with significant wind problems. Absolutely, having the mics up would be ideal. But with this recording it is secondary to the wind problem.

[raymonda]

Chris, pardon me for making a wild assumption.  I thought you might test the mics some other place or before the performance.  I guess I should have spelled that out for you.  Next time.

  Yes, there are always exceptions.

[mfrench]

this is a frequency scan of a short solo - that had some backing ensemble accompaniment - note the sharp spike at 20k+ :)