Weird noises when recording

Louie Louie

New member
Hello lads,

I'm new around here and I've registered because I'm having a problem that I just can't fix. I'm recording my acoustic guitar and I'm having really weird noises. It sounds like a resonance issue, kinda flanging, kinda phasing, kinda wah noises. It happens in all frequencies. After I cut out all the noises from all the frequencies with the EQ, there was no acoustic guitar sound left!

I've researched a lot, but I'm still in the dark.

First of all, I've tried thousands of mic positions. Best one I could find was 30º angle, pointed to the 12th fret, 1.5ft away from the acoustic guitar. But still didn't solved the problem.

Then I thought it was the acoustics of the room. I went to a studio with proper acoustic treatment, though the problem persisted.

Finally, I assumed that it is probably the cheap quality of my instrument.

However, I've read some posts where people had the same issue with good guitars - which gave me some hope of figuring out how to avoid this. Even so, I'm a bit lost and I don't know how to fix it.

Can someone help me? I'm attaching a file with my recording without any effect or EQ.

My current gear is:

PC: Dell Inspiron
DAW: Reaper
Interface: Behringer U-Phoria um2
Mic: Behringer C1
Guitar: Memphis MD-18

Cheapest gear money can buy (around $200 the full set)!

I thank you all in advance! Cheers mates!


View attachment test.mp3
 
Maybe you should also describe how you record into your computer...the setup, the settings.
It could be something there, and not with the mic or guitar.
 
Well, I record mono, with an xlr cable. I usually record between -12 or -6 dbs.
Out is set to 2x 32-48kHz, 16Bits. In is set to 2x 11.0-48kHz, 16 bits. Buffer size is 2048 samples. Latency compensation is 32 samples for both, in and out. Kernel buffers is set to 2. Resemple 44.1kHz <-> 48kHz. My audio thread priority is ASIO default/MMCSS Pro Audio/Time Critical. I've already recorded with and without direct monitoring. Matter of fact, those are the standard configurations - never thought about changing though... Can that be the problem?
 
I don' think it's technical, but just a boomy guitar that has a peak the microphone hears quite well. EQ will probably just make it thin. I'd start by a mic swap if at all possible to remove the mic from the equation. The mic and the guitar could have a similar peak in their response - another mic will soon answer this. Then maybe move the mic out to double the distance and though this will make the sound much thinner - it will either still reveal the resonance, or not. If at the distance it vanishes, then you just need to try lots of different positions. I'd get somebody else to play, stick some headphones on and then move it around while they play. Your settings are quite typical - I suspect miroslav was thinking maybe you were processing during record - dynamics or frequency stuff, but your interface isn't known for being coloured, or the mic really. You just need to borrow another mic to be dead sure, but my money is on the guitar just having a resonance that the mic picks up too well!
 
I'm not actually hearing anything that is killing the tone. If anything I hear the resonant frequency of the guitar popping out around 250, but yet no low end. That is likely the guitar itself. Not necessarily the issue tho.

The string/pick noise is a bit much, but there is a ton of background noise. That will be mic placement and maybe mic too, but overall I am not finding it anywhere near bad. I would absolutely suggest a LDC on the bridge and a LDC on the 12th fret.

What is your goal here? Are you planning on adding other instruments?
 
Looks like you have a HPF on the track. I have nothing showing below 100hz. A recorded track of acoustic should have that.
 
I still hear the background noise level quite high, but that may just be the U-Ph preamps being a bit high noise or gin up a bit past the ideal point. I would again suggest 2 mics.
 
Thanks a million for your opinions lads! Undoubtedly, I'll engage in this fórum - it's jut amazing have people helping you like that!

rob ayelstone: I thought the guitar resonance was the issue as well. It seems quite worthy to buy a better one. However, I have a bit of a drama around here. I'm currently living in Brazil and everything around here is ridiculously expensive due to tax policy. I chose to buy the cheapest guitar (around $70) because it does not pay to invest in a better one. For instance, a simple Martin DXMAE, which is probably the cheapest Martin guitar, costs around $599 (something I can easly afford). Here, a DXMAE is bizarrely costing $2.213,54! So, I'll have to wait for my next trip abroad to buy a better one.

jimmys69: I have only one behringer c1, which is a LCD if I'm not wrong. Otherwise I would definiely use another mic on the bridge. I didn't use any HPF on this track. Matter of fact, if I use a HPF nothing changes until 130hz. That's just the cheap guitar speaking, I guess. I'll work on mic placement to avoid the sting/pick noise (thanks for the tip!), but today the background noise was a bit too much in my building. My goal is too record a very simple demo with my songs - it is a songwriter project, with only voice and acoustic guitar. Here is an example of what I'm trying to do (I recorded a few weeks ago):

View attachment 1-song.mp3

Your "little fun" is way better than the original and way beyond my newbie mixing skills! I'm still learning! But, maybe my speakers/headphone just accentuate a bit too much the noise! That might be one good hypothesis as well and the reason why I can properly mix my tracks and fix the noises...

Anyway, thanks again, both of you! I'll keep trying and once I've achieved a better quality level I'll post around here!

Cheers!
 
It likely the the guitar then. The C1 in spec has the ability to pick up the bottom end of the guitar to 40hz. But a $70 guitar is , well, a $70 guitar. Sometimes one can find one that sounds great, most times not so much.

But, your playing seems solid so you have that going for you. If you indeed used the C1 at the bridge and there is no low end, then I suspect the guitar or an inadvertent placement of a HPF on the track. I would double check that first.

We will be around man. Keep posting your recordings.

I am always here to help if I can. :)
 
When I record acoustic, I always start with the mic straight out from the neck joint (or about the 12th-14th fret), angled toward the upper bout and soundhole. You can increase or reduce the bass (or woofiness, if it mics up that way) by adjusting the angle and distance from the guitar. This is not the only way to record, of course, or the only way I've recorded, but it usually turns out pretty good if I spend time getting it right. Having 2 mics does make it less critical where each one is, since you can mix them to achieve the sound you want. I use a pair of SDCs usually in those cases.

Bass is going to improve as you capture more of the entire top, and to some degree (depending on the guitar and mic) soundhole. It's always some learning, with the guitar being a big variable, as well as your playing style. If you play softly, you won't really move a lot of air, and you'll have to mic closer to optimize s/n, but then you'll lose some bass and pick up more mechanical noises from your playing. Softer playing really requires a responsive guitar, a good mic, and a good room.

So, work on mic placement and don't mess with the mixing until you've really optimized the recording part. It's less important in a big mix, but solo or guitar+voice, it needs to be as good as possible.

The 16-bit limit is working against you, too.
 
So, quick update!

I've tried again, with a better acoustics and another mic placement. I'm attaching two files: first one with no EQ at all, second with some Noise Gate, EQ and compression. I still had to do a lot of cuts, but at least I think it sound a bit better. See what you think:

View attachment test2.mp3

View attachment test2EQ.mp3

Just to let you know how I EQ this one:

- I've add a noise gate at -40db.
- I've used ReaEQ to insert HPF at 80Hz and LPF at 20kHz.
- I've cut: 226.7Hz (-3dB); 390.8Hz (-7dB); 505.5Hz (-7dB); 700Hz (-7dB); 1063.9Hz (-5dB); 1570.4Hz (-5dB); 2981.7Hz (-5dB); 8083.9Hz (-3dB).
- I compressed really smoothly: threshold -30.1dB, attack 150ms, release 70ms, ratio 4:1, output +8.
- Then I've used another EQ (TDR Nova by Tokyo Dawn Labs) to boost 2dB at 450Hz, to cut some more at 700Hz (-2dB) and to boost 1dB at 3.5kHz and 1.5dB at 12kHz.

As I told you before, I'm new on mixing, so I might had made some pretty weird things!

Looking forward for your opinions!

Cheers mates!
 
That's a lot of EQ manipulation.

I guess I'd want to hear this guitar in a mix, i.e., what's the final piece you're working on? For me, those kinds of "surgical" cuts are something I'd defer until I knew what else was going on around it. This is (I assume) not a solo guitar work, so how it sounds alone is not so relevant. Of course, HPF and LPF are things I probably apply fairly blindly to a rhythm guitar track (well, almost any track) and if it's a guitar I know that will be used with any other instruments I'll have a low shelf ready, and if my croaking is present, there will be a fairly broad EQ down low ready for use. But, not all of those things going on.

So, to me it sounds like you've squashed it a bit much and taken some of the sparkle out of the performance. Just because you see something peaking in a track doesn't mean it immediately needs to be knocked down. Wait until you've got more of the song together and then listen to the whole.

And, finally, ALWAYS compare before and after at the same loudness level to see whether what you have done is really an improvement, or just louder, which will sound like an improvement 90% of the time, but that's your ears/brain fooling you. I mean, it might be better, but you can't be sure until you eliminate the loudness bias.

Here are your 2 MP3s with a limiter that I applied to push them to ~-14dBLUFS, and then the original with just the kind of EQs I might start with, and compression, with the same limiting. Now, a lot of folks would find the guitar needing something, but they say that about my guitar playing too :)... You have to decide what your sound is, and how it will be used before you hack away mercilessly. You might cut out something that's actually worth keeping.
View attachment test2-lim.mp3
View attachment test2EQ-lim.mp3
View attachment test2kr.mp3

P.S. I'm working on my old laptop with ancient MDR-V6 cans - son has reclaimed his bedroom for his Christmas visit, and all my stuff is powered down. No idea what this really sounds like ;)

P.P.S. Some of that "resonance" I've seen mentioned is, to me, as much technique as the guitar itself.
 
Yeah, what is the final goal for this? Meaning are you going just acoustic and vocal? Or adding bass guitar and drum as well?

Doing two tracks of the acoustic will help if you are looking for a bigger sound. There are many options depending on what you are looking towards for final production.
 
Looks like you have a HPF on the track. I have nothing showing below 100hz. A recorded track of acoustic should have that.

Yes, guitar does go own to 80 Hz and there should be some "chunk" below even that. Of more concern* is the cliff edge at 8kHz. That could be the result of the MP3 smash (I like an attachment of the best 320k MP3 you can do) but if in the original you have lost all the "air".

But, I actually agree with Jim, not really a lot wrong with the basic sound. 16 bits should not really add too much noise, CD noise floor anyway, but many cheap 16 bit converters need Windows setting up properly for them. You need to make sure to have the record gain very low.

Dave.
 

Attachments

  • lous git.png
    lous git.png
    24.3 KB · Views: 11
Regarding the frequencies in the recording, the guitar's low E string stops at 82Hz (E2) so maybe over-agressive HPF; but beyond that, this sounds like a guitar with a capo on the 3rd fret (playing Am-C "open" chords), so I'd not expect much low frequency now that it's stopped around 98Hz (G2)!
 
Back
Top