Recording Harsh Sound Quality for Higher Notes ?

[rhysonsax]

I record my SATB saxophones, in a small-ish room, using a decent Electro Voice RE 27 N/D dynamic mic, going through a Focusrite Scarlett interface and into Audacity. The mic is slightly above the instrument bell, pointing horizontally at the LH keys and about 50-80 cm from the instrument.

On listening back to recordings, I find that higher notes on the tenor (say G2 upwards) and alto (same sort of range for written notes) sound quite harsh compared with how I hear them playing into a wall or corner.

I suppose this could be caused by the mic's characteristics, the room, the way the mic is positioned and pointed. Or it could be me and how I play when I'm feeling the pressure of recording. Or maybe the reflections from playing against a wall are flattering the sound and I really do sound harsh higher up the range.

I have tried some significant equalisation on the recordings to reduce the gain for higher frequencies, but so far that either doesn't do very much or it makes the whole thing sound muddy and muffled.

Any suggestions for what is causing this and what I can do about it ?

Rhys

 

[altissimo]

trying to track down the cause of things like this can be difficult, but here's a few general suggestions some or all of which you will have already tried -
Check your recording levels - there's often more than one place where you have to set levels - so check levels on the interface and in the software - level meters are often inaccurate so set it so the peak level in the software is about - 8db or less, you can always normalise it later and use some compression etc in the mix
Check your recordings on another device -, sometimes it's the monitors and the acoustics of where the speakers are positioned that's the problem, but if the problem's there on headphones or another set of speakers as well then at least you've eliminated one possibility..
Room acoustics - hang up some coats of something similar along the wall that the mic is facing - internal reflections in small rooms rarely sound pleasant so you want as much stuff in the room as possible to break up and absorb those reflections - bookshelves, thick carpet, soft furnishings, thick curtains etc. Reflections of objects near the mic can sometimes affect things - keep mics away from big flat metal music stands.
Loose screws etc - sometimes it's just a loose screw vibrating on the mic stand or instrument - I've tracked odd sounds down to vibrations in the central heating radiators although that's not likely at the frequency range you've got problems with. I don't think this would be the cause of what you've described, but I'm trying to eliminate all possibilities
Mic position - you seem to have things set up well, but it's worth a fiddle (?) with mic height, angle and distance and try recording in a different part of the room in case there's some internal standing waves emphasising certain frequencies in a particular spot
EQ - if you end up having to do some EQing then use a parametric eq and try to zero in on the troublesome frequency area - try somewhere between 2 and 6 khz with a very tight bandwidth take the gain down to -10bd and sweep the frequency back and forth until you find the bit you want rid of, adjust the gain and bandwidth back to something more sensible. If you're having difficulty finding the nasty harsh frequencies then boost a narrow band of the eq and adjust the frequency until you find the really nasty harshness, then you know where to cut
this article may be of use
How To Use EQ Like a Pro (4 Key Approaches & 10 Top Tips)

a look at the frequency response chart of the EV RE27N/D indicates that there's a peak or two in the 3-6 khz region that may be emphasising harshness - https://www.electrovoice.com/binary/RE27N-D_Engineering_Data_Sheet.PDF so that's a good guide to how the eq your recordings and is the most likely cause of your problems - I should've looked at this first....

 

[rhysonsax]

Thanks @altissimo some very good points there. I used to think that my AKG headphones were emphasising harshness somehow, but I get much the same sound quality from my B&W speakers.

The frequency response chart for my EV RE27 N/D is interesting. This mic is an "updated" version of the much-respected EV RE20 dynamic that is used for voice radio and also recommended for saxophone. But although it has the same external shape, its frequency response has that big lift at higher frequencies. I have the switch selected to reduce the HF lift (the dashed line on the chart), but I have no way of knowing whether it is having the desired effect.

I wonder whether it is possible to generate a graph like that at home, without needing some really significant hardware and/or software..

Rhys

 

[altissimo]

I think the main reason for the RE20 being popular for sax use is that it's almost indestructable and nearly impossible to overload and therefore gets pointed in the direction of any instruments that sound engineers are afraid of, so they're popular as bass drum mics and that keeps secondhand prices high.. If I wanted one mic for stage and recording use I'd prefer a Beyer Dynamic M201
I think by 'updated' you mean cheaper... I assume the purpose of it's boosted upper mids is to help vocals cut through, a bit like the similar hump in the response of an SM57.
I use a cheap Audio Technica At3031 condenser mic for recording, but it isn't made any more.,Given the relatively limited frequency range of the sax it's not a difficult instrument to find a mic for as long as the response is fairly flat and doesn't have those 'presence peaks' that seem popular in vocal mics.
If you want to make acoustic measurements of the combined effect of your room and mic then put some white noise through your B&W speakers at a fairly loud level and record it and find a freeware specrrum analyser plugin for your recording software if it's not got one already and you'll get a rough idea of the peaks in your room and mic..
Meanwhile, get used to using a parametric EQ to tame that lump in the mic's response - it's what EQ is there for.
The only problem with Audacity is that you can only listen to a 30 second sample of the track you're EQing, so you have to select a representative part of the track, fiddle with the EQ until it sounds right and then select the entire track and apply it to the whole thing and hope it's ok - a pain in the a*&% that makes some of us prefer different software like Cubase, Reaper or the free copy of Pro Tools that you may have got with your Scarlett interface

 

[Colin the Bear]

If you remember while you're playing, that your high notes are coming out harsh in the recording, try to soften them while playing. You may have to adapt your style to record with this gear in this room.

If a piece has a lot of high notes, I would position the mic to the side, below the bell. The highs seem to come bowling out of the upper tone holes and drown the mic. Also try to play them a little quieter. Maybe more soft furnishings will help. I find I get a noticeable difference if I draw the curtains and leave the door open in my small room.

Perhaps a screen in front of the mic.

 

[brianr]

Any suggestions for what is causing this and what I can do about it ?

I think you know the answer but are just putting your head in the sand.

Its obvious that you need to spend at least 5k each on a new tenor and alto, and a couple of new, boutique mouthpieces, at 500 each would almost certainly help.

Go on, you know it makes sense.

 

[jbtsax]

There is a principle of woodwind acoustics that applies in this situation. All of the harmonics whose frequencies are above the instrument's "cutoff frequency" which is in the neighborhood of F#3 for most saxes go straight out the bell. In Benade's words, they don't even see the open toneholes. This means that a mic close to the bell opening is picking up more of these high freqencies. A good portion of the sound waves of the fundamental and its lower harmonics are disseminated into the room through the open tonehole(s).

 

[altissimo]

There is a principle of woodwind acoustics that applies in this situation. All of the harmonics whose frequencies are above the instrument's "cutoff frequency" which is in the neighborhood of F#3 for most saxes go straight out the bell. In Benade's words, they don't even see the open toneholes. This means that a mic close to the bell opening is picking up more of these high freqencies. A good portion of the sound waves of the fundamental and its lower harmonics are disseminated into the room through the open tonehole(s).

"The mic is slightly above the instrument bell, pointing horizontally at the LH keys and about 50-80 cm from the instrument. "
if it's 50 -80 centimetres from the instrument then we can interpret that a meaning just above the height of the bell and 50-80cm in front of the instrument and not positioned just above the bell itself.

 

[saxyjt]

On listening back to recordings, I find that higher notes on the tenor (say G2 upwards) and alto (same sort of range for written notes) sound quite harsh compared with how I hear them playing into a wall or corner.

Excuse the silly question, but did you try recording from where you hear yourself? Near your head/ears, in the same conditions as when it sounds good to you.

 

[jbtsax]

"The mic is slightly above the instrument bell, pointing horizontally at the LH keys and about 50-80 cm from the instrument. "
if it's 50 -80 centimetres from the instrument then we can interpret that a meaning just above the height of the bell and 50-80cm in front of the instrument and not positioned just above the bell itself.

Your point is well taken. The pickup pattern of the mic also determines how the upper frequencies coming from the bell are recorded.

 

[altissimo]

Your point is well taken. The pickup pattern of the mic also determines how the upper frequencies coming from the bell are recorded.

which pickup patterns have difficulty picking up upper frequency sounds directly in front of them? It's usually the polar pattern to the sides that becomes a bit erratic,, particularly hypercardioid and fig 8.
The RE27 N-D is cardioid and the polar response chart is given in the engineering data sheet in the link I posted above

 

[jbtsax]

which pickup patterns have difficulty picking up upper frequency sounds directly in front of them? It's usually the polar pattern to the sides that becomes a bit erratic,, particularly hypercardioid and fig 8.
The RE27 N-D is cardioid and the polar response chart is given in the engineering data sheet in the link I posted above

I have much to learn from you and others on this subject. Thanks for sharing your knowledge and experience with us.

 

[altissimo]

it's important to realise that microphones don't 'hear' sound the same way we do and all a mic can do is give a reasonably accurate indication of the variations in pressure on the surface of it's diaphragm and as such it's turning a three dimensional phenomena into a two dimensional voltage signal. So recorded sound is not the same as real sound and you can't really capture the full resonance of a musical instrument making the air vibrate around it. But our ears become accustomed to recorded music and we mistake it for reality. My sound recording lecturer used to remark, while watching us busily putting mics on all the instruments "remember, you're creating an illusion" meaning that there's nowhere you can stand in the room and hear the multi miked sound that we were recording and would always encourage us to use our ears and listen for the sweet spots in the room.
When you're recording yourself at home you don't have anyone else to help you find the right position for the mic, so a lot of trial and error is needed. I tend to mic from the right hand side about stomach height, so I can move around a bit and not get the variations in level you might get if you're bobbing around with the mic in front of you, but recently got good results with the mic at about head height about 12 inches away from me, pointing down towards the L/H table and got a crisper sound than I usually get from a Martin alto, I've also got good results miking from low down pointing up and directly overhead. I've tried using more than one mic, but it doesn't make a great deal of difference and eventually I have to conclude that it always sounds like me.

 

[rhysonsax]

Excuse the silly question, but did you try recording from where you hear yourself? Near your head/ears, in the same conditions as when it sounds good to you.

Playing into a wall or a corner of two walls and hearing the reflection is how I hear myself. I don't think I want to record like that, but maybe raising the mic higher up to head height and pointing it slightly downwards towards the LH keys from further away could be something to try. At the moment it is roughly level with my left hand and pointing in horizontally.

Rhys

 

[Colin the Bear]

A good monitor must help. I haven't got one but if the sound in the headphones is the same as being recorded then you can adjust the way you play to fit the situation. I often record, playback and rerecord trying to imagine what's going down and smooth out the bits that grate.

 

[altissimo]

your mic positioning is perfectly sensible and similar to that used by Rudy Van Gelder on some of those old Blue Note sessions.

Having the mic higher up and pointing down may tend to pick up more of the higher harmonics from the bell, but you don't know until you try it - I once recorded with the mic above and slightly behind my head in the hope that it'd pick up the sound from the player's point of view. It didn't do that, but it sounded ok.
Tenor records differently to alto and sometimes you just have to spend a day trying the mic at different heights and angles... I remember having the mic about two feet above floor level on the right hand side pointing up towards the middle of the sax for one session with a tenor player., so really anything goes...

 

[rhysonsax]

I still record in the same room, with the same gear and although I have slightly altered the mic position, it has only slightly softened the harshness I hear at higher frequencies.

I have begun to wonder whether there is a resonant frequency due to some dimension in the room that is causing the problem. Although the walls are covered in books and soft finishes, the ceiling and floor are both hard and flat and parallel (unsurprisingly). The room is approximately 18 x 7 x 8 feet and box-shaped.

Can room resonance be a significant issue for recording the saxophone (or other, lesser acoustic instruments) and what can be done about it ?

I have seen various "reflection filters" that go around microphones, typically for recording the human voice in a "dry" way. This sort of thing:

But searching on the internet, it seems that opinions are divided about how useful such reflection filters are for voice recording and I've not yet seen a saxophone being recorded with such a filter in place.

Any thoughts ?

Rhys

 

[nigeld]

the ceiling and floor are both hard and flat and parallel

I guess that a carpet would be an obvious thing to try.

 

[Pete Thomas]

I have tried some significant equalisation on the recordings to reduce the gain for higher frequencies, but so far that either doesn't do very much or it makes the whole thing sound muddy and muffled.

The answer here may be to use a filter (Low pass) rather than normal EQ that would just boost or reduce certain frequencies.

 

[Jazzaferri]

Are you able to borrow beg or otherwise liberate other mic's to see how they sound. Some of my mic's are OK with vocal and suck on sax some are great on vocal and ok on sax but suck on guitar. Every good recording engineer has a number of mics that work for different instruments in their studio. Finding a nearfield or midfield position that works well can take a lot of trial and error. I have spent half an hour playing around with mic positioning before finding the sweet spot with some.

I am not a fan of dynamic mics for recording sax but that is a personal thing.