Audio clip - classical guitar recording - AEA R88 + Soundfield - any comments?

[Douglas Whates]

I'd appreciate some comments on this classical guitar recording:

(First three are a blend of AEA R88 blumlein close mic mixed in with a Soundfield 422 distant mic.)

http://www.naturalstudio.co.uk/misc/...st1_noverb.wav
dry version, just over 30MB, 44KHz/16bit

http://www.naturalstudio.co.uk/misc/...r2010test1.wav
wee bit of added reverb version, again it's just over 30MB, 44KHz/16bit

http://www.naturalstudio.co.uk/misc/...rb_aligned.wav
time aligned version, no added reverb, again it's just over 30MB, 44KHz/16bit

http://www.naturalstudio.co.uk/misc/..._aea88only.wav
only close mic (AEA R88), again it's just over 30MB, 44KHz/16bit


I almost never multi-mic a solo instrument, so I'd be interested to hear what those with more experience recording guitar in this way think.

Thanks in advance for your comments and constructive criticism.

Douglas.

[Corran]

Personally I'd probably take away the added reverb, I think it's a tad too wet. For my tastes anyway. The tone is very nice though.

[Corran]

Just for reference here is a track from a CD I recorded a few months ago. The direct-to-reverb balance was decided on by the player and his friend, another classical guitar player.

[chris319]

Why do you need reverb and a distant mic? It makes it sound like the room acoustics are bad. I'm not sure the close-in ribbon mic is doing you any favors, either.

[Douglas Whates]

Originally Posted by chris319

Why do you need reverb and a distant mic? It makes it sound like the room acoustics are bad. I'm not sure the close-in ribbon mic is doing you any favors, either.

I would almost always prefer just one distant pair, but in this session I had three pairs of microphones - a very close pair of ribbons, a middle distance pair of sub-cardioids, and a distant MS pair. My intention was to use the middle distance pair only and possible use a tiny pinch of the others for a touch of support. Unfortunately, there are weird phase issues with the middle distance pair, so I can't use those at all. The MS distant pair are too distant to use on their own.

This was recorded in a church. The distant mic does pick up a lot of reverb, but I felt it doesn't have much width. So, I added a wide reverb to the close ribbon channel as this was bone dry.

Anyway, I've now uploaded a version without reverb (see original post). What do you think? An improvement, or...?

[Douglas Whates]

Originally Posted by Corran

Just for reference here is a track from a CD I recorded a few months ago. The direct-to-reverb balance was decided on by the player and his friend, another classical guitar player.

This sounds to me like it was recorded in a smallish room with artificial reverb added, no?

[Douglas Whates]

Originally Posted by Corran

Personally I'd probably take away the added reverb, I think it's a tad too wet. For my tastes anyway. The tone is very nice though.

I've uploaded a dry version. See what you think.

[Corran]

Originally Posted by naturalstudio

This sounds to me like it was recorded in a smallish room with artificial reverb added, no?

No, this was in a college auditorium. Not the best acoustic in the world but not bad. Why do you think it sounds like added reverb?

I like the dry version a lot better. It has a better presence to me. Of course, the player should probably have some say in it.

[Douglas Whates]

Originally Posted by Corran

No, this was in a college auditorium. Not the best acoustic in the world but not bad. Why do you think it sounds like added reverb?

I thought it sounded like added reverb because there are some strange early reflections and a bit of hazy imaging/phasiness, which I suppose I mistakenly assumed were early reflection issues from nearby surfaces.

Originally Posted by Corran

I like the dry version a lot better. It has a better presence to me.

Yeah, I'm leaning towards preferring the dry one now too.

[mathieujm]

Originally Posted by naturalstudio

I've uploaded a dry version. See what you think.

This one is very nice. Nothing to add.

How did you delay the close pair ?

JMM

[Douglas Whates]

Originally Posted by mathieujm

This one is very nice. Nothing to add.

How did you delay the close pair ?

JMM

I didn't delay the close pair at all. I'll maybe experiment with that - never thought about it.

[chris319]

Yes, definitely a noticeable improvement without the reverb. Just for fun, can you post the close-in mics all by themselves? I'm concerned that there are phasing issues using more than one mic on a single monaural source. Here, read about the three to one rule:

THE THREE TO ONE RULE & PHASE CANCELLATION FULLY EXPLAINED

[rumleymusic]

Love the dry version. Doesn't need any reverb.

[Douglas Whates]

Originally Posted by chris319

Yes, definitely a noticeable improvement without the reverb. Just for fun, can you post the close-in mics all by themselves? I'm concerned that there are phasing issues using more than one mic on a single monaural source. Here, read about the three to one rule:

THE THREE TO ONE RULE & PHASE CANCELLATION FULLY EXPLAINED

I am perfectly aware of the three to one rule!

I'll post the close mics, sure - will post them in a few minutes (EDIT: that's them there now, see original post). What phasing issues do you hear?

Douglas.

[Douglas Whates]

Originally Posted by rumleymusic

Love the dry version. Doesn't need any reverb.

Glad you like it.

[chris319]

I listened over and over to both dry recordings and frankly couldn't hear that the second set of mics added anything. I think you were just as well off with the single close-in pair.

[Douglas Whates]

Originally Posted by chris319

I listened over and over to both dry recordings and frankly couldn't hear that the second set of mics added anything. I think you were just as well off with the single close-in pair.

EDIT: Oops - I had mis-named the files - they were both identical! Fixed now!

[Douglas Whates]

Time-aligned version added (see original post)

[chris319]

Well you snuck in a pretty good check against placebo effect there!

The time aligned version has a bit more depth but the single pair is still a satisfying recording. The R88 consists of two figure-8 mics so you're going to pick up some measure of room with the single mic (pair) regardless. However, both dual-pair versions have noticeably more hiss and this detracts. I tried attaching a version with hiss reduction applied but the gearslutz upload thing doesn't like me today.

[Douglas Whates]

For a classical release, I feel the close pair alone is *far* too dry. It sounds nothing like the room and lends no support to the guitar. If I was going for a studio sound, then maybe it's okay.

Also, contrary to you, I feel the time-aligned pair loses a bit of depth.

Yes, there is quite a bit more background noise on the dual pair recordings, but I'd almost always prefer a bit of hiss over artefacts from noise reduction. Is the hiss really that distracting when listening at normal levels?

[CGBrian.M]

I'm a professional, conservatory trained classical guitarist (hence, the CG in my user name), and I make my entire living off teaching and playing the classical guitar. This is completely a matter of taste, but I generally don't like ribbons for classical guitar (and little else). I find them highly colored, muffled, hard, and compressed sounding. Very often, there is a noise issue, as there is here.
Having said that, I also much prefer the 1st no reverb version you posted. Your recording also has a very professional sound. I have several classical guitar CD's I own that don't sound anywhere near this good. The no reverb version has more space, and I can hear the decay of notes much better. For instance, the high C# about 40% of the way through that is ever so sweet.
The other versions, while not bad, are thinner & harder sounding me.
Overall, good job, I say!

Brian

[chris319]

Is the hiss really that distracting when listening at normal levels?

Yes. I would not consider it an acceptable recording.

I'll try attaching the file again, but I heard nothing in the way of artifacts after applying the hiss reduction.

[CGBrian.M]

Originally Posted by naturalstudio

For a classical release, I feel the close pair alone is *far* too dry. It sounds nothing like the room and lends no support to the guitar. If I was going for a studio sound, then maybe it's okay.

Also, contrary to you, I feel the time-aligned pair loses a bit of depth.

Yes, there is quite a bit more background noise on the dual pair recordings, but I'd almost always prefer a bit of hiss over artefacts from noise reduction. Is the hiss really that distracting when listening at normal levels?

I agree with the above statements completely.
I would much rather not have the hiss, as it is quite present. But the playing is so good, and the other elements done so well that I probably wouldn't focus on it to much.

Brian

[David Spearritt]

Douglas, the dry one magnificent, a "real" recording of a guitar. I feel it needs a little reverb, but not quite as much as in the first sample.

Well done, classical guitar is mostly badly recorded, this is excellent. Sounds like a real instrument, and there is a wonderful rock solid phantom image in the middle of the speakers as there should be.

[chris319]

[clips deleted by poster]

I agree that it is a good performance, and ribbons wouldn't be my go-to mics for classical guitar.

Is there a reason these files are 16 bits? Easier to send over the Internet?

[Douglas Whates]

Originally Posted by chris319

I would not consider it an acceptable recording.

Does anyone else agree this this (with regards to the level of hiss).

[Douglas Whates]

Originally Posted by chris319

Well, screw the gearslutz file attachment deal. There are two files which can be found here:

http://www.miclisteningroom.org/octo...hissreduc1.wav

http://www.miclisteningroom.org/octo...hissreduc2.wav

#1 has full noise reduction and #2 has half as much. If you can hear artifacts in either one, it's either placebo effect or your hearing is much better than mine.

I agree that it is a good performance, and ribbons wouldn't be my go-to mics for classical guitar.

Is there a reason these files are 16 bits? Easier to send over the Internet?

Sometimes it is difficult to objectively analyse a noise reduced sample vs. non-noise reduced sample, as it is too obvious which one has the hiss and which doesn't.

However, in this case I am quite confident that I prefer the non-noise reduced version. I feel both of your noise reduced examples have not only removed a lot of the subtle nail and ricochet noises, it also seems to have resulted in a flatter sound with less depth and less clarity - more veiled than the original.

Although, I would say, as noise reduction goes - it's certainly not too unpleasant.

(Yes, 16bit because that's easier for most people)

[Douglas Whates]

Originally Posted by CGBrian.M

I'm a professional, conservatory trained classical guitarist (hence, the CG in my user name), and I make my entire living off teaching and playing the classical guitar. This is completely a matter of taste, but I generally don't like ribbons for classical guitar (and little else). I find them highly colored, muffled, hard, and compressed sounding. Very often, there is a noise issue, as there is here.
Having said that, I also much prefer the 1st no reverb version you posted. Your recording also has a very professional sound. I have several classical guitar CD's I own that don't sound anywhere near this good. The no reverb version has more space, and I can hear the decay of notes much better. For instance, the high C# about 40% of the way through that is ever so sweet.
The other versions, while not bad, are thinner & harder sounding me.
Overall, good job, I say!

Brian

Thanks very much for your comments. Yes, I decided that I didn't really like ribbons on guitar, but for this recording we were attempting a bit more robust, perhaps less realistic sound. Glad you like the overall results and guitar sound - PM me if you'd like more info about performer and instrument.

[Douglas Whates]

Originally Posted by David Spearritt

Douglas, the dry one magnificent, a "real" recording of a guitar. I feel it needs a little reverb, but not quite as much as in the first sample.

Well done, classical guitar is mostly badly recorded, this is excellent. Sounds like a real instrument, and there is a wonderful rock solid phantom image in the middle of the speakers as there should be.

Thanks for the comments - yes, it's a nice strong image. Maybe even a bit stronger than I am used to.

I too might like a bit more reverb, or some more space around the sound, or something. I don't know - it's just a touch closed in for me. But, I suppose it's because I know what the original space sounds like.

[Teddy Ray]

for a bit of reference..


Julian Bream, Julian Bream Edition, Vol. 12: 20th-Century Guitar I, RCA 9026-61595 (1995).
Miguel Llobet, Miguel Llobet - The Guitar Recordings 1925-29, Chanterelle CHR 001 (*year).
Paco de Lucia, Siroco, Mercury 830913-2 (1987).
Ramon Montoya, Grandes Figures du Flamenco, Vol. 5, Le Chant du Monde LDX 274879 (1988). Distributed by Harmonia Mundi.
Presti and Lagoya, Baroque Music for Guitars, Philips 422-285-4 (recorded 1962–1965).
Niño de Ricardo, Grandes Figures du Flamenco, Vol. 11, Le Chant du Monde LDX 274927 (1991).
Sabicas, La Guitarra Flamenca, Alfa AF-CD10 (*year). Distributed by FTC.
Andrés Segovia, The Segovia Collection (Vol. 1) Bach, MCA Classics 42068 (1987).
Andrés Segovia, The Segovia Collection, Vol. 3: My Favorite Works, MCA Classics 42069 (1988).
John Williams, From the Jungles of Paraguay: John Williams Plays Barrios, Sony SK64-396 (1995).