Take the 2-minute tour ×
Sound Design Stack Exchange is a question and answer site for sound engineers, producers, editors, and enthusiasts. It's 100% free, no registration required.

I'm trying to get into home music recording of music that will now and then have some very quiet passages, quiet vocals, acoustic guitar--delicate stuff in addition to louder sections. My current setup is this:

  • Laptop PC running Audacity (though am open to any recording software)
  • MAudio Firewire digital interface to accept XLR mic in and instrument lead in.
  • 1 Shure SM57 dynamic mic
  • 1 Audix OM2 dynamic mic (similar to Shure SM58)
  • Concrete basement room with no audio treatment of walls, no rug. Not obviously echoic, but obviously not an audio studio!
  • Various acoustic and electric instruments, drum machine, etc.
  • $2 ear buds for listening--for now (I know).
  • My larynx :D

The problems are two:

1) Boosting quiet recording results in hiss: When using the mics to record vocals or just acoustic guitar, the sound level I am getting is "not all that loud" (as judged by listening through the ear buds or the computer's speakers...or the audio signal on Audacity) even if my singing or guitar playing is nothing close to "quiet" if a person were sitting anywhere in the room with me. Therefore, to get what seems like a better level, I use Audacity to increase the decibels of the recording. But, then there is an obvious "hiss" during the silent parts of the recording (actually it's there throughout, of course...just more noticeable when there is no other sound playing). I tried to use Audacity's "Remove Noise" algorithm, but even with a long sample of a silent period and the most conservative noise removal level, it creates an obvious artifact (like "tinkling digital crystals") that sounds pretty grim and is just unacceptable.

2) The sound recorded through the mics is "muddy": I just don't hear the crispness of the voice or guitar strings when recorded through the mics. So, of course I tried using the equalizer on Audacity to boost the higher frequencies--but then we're back to Hisstown, USA.

It seems to me that if I could get a higher gain from the start, picking up louder and higher frequencies and bringing that into the digital recording, I would not have to do much (or any!) EQ'ing or boosting and I'd get a very clean recording.

What I want to know is: How do I quantify this and then what should I expect? How do I know what I "should" expect given this setup, or what is possible with a, say, condenser mic or other things people might suggest? I don't want to just base things on anecdote and personal judgment: I want to be able to measure the noise and know if I'm coming up against "acceptable limits". My training was as an experimental scientist, so I still tend to think in terms of what I can measure and what are acceptable benchmarks for quality. I know it is unreasonable to expect the highest level of sound quality when recording in a cheap home setup as if it were a top of the line professional studio with $10k mics, but on other hand I suspect that I can do a lot better than this with some adjustments or not-that-costly purchases.

Secondly, while you're at it, if you can suggest what will improve this "quiet and muddy or hissy, you choose" problem, that would also be helpful.

share|improve this question
add comment

migrated from avp.stackexchange.com Jan 24 at 12:01

This question came from our site for engineers, producers, editors, and enthusiasts spanning the fields of video, and media creation.

1 Answer

up vote 3 down vote accepted

To quantify this problem, you want to know the difference between your recording level ("how high the meters are when you're making sound into your mic") and the noise level ("how loud things are when you aren't making noise"). In simple terms, the difference between these is called signal/noise ratio, or SNR. When you boost the signal, you are also boosting the noise.

You should expect a relatively high SNR - when recording my voice through a similar setup (Shure SM57 through an M-Audio Fast Track Ultra), I could peak at -6dBFS while the noise floor with the mic attached was around -50dBFS due to ambient noise in my apartment. These numbers were measured using the meters in my recording software.

For reference, dBFS refers to decibels full-scale and can be measured with any meter anywhere in the digital signal chain. I know Audacity has such a meter, and your M-Audio drivers probably do too.

The individual components can each add noise - it could be coming from the mic, the cables, or the interface. Generally, when I've seen problems like this in my hobby recordings, it's either the mic or the interface. Generally, the interface inputs are not turned up appropriately, so the signal reported to the computer is quite low and very close to the noise floor and the solution is to turn up the input gain on the interface and disable any padding. You want to raise the level of your audio so that it is above the noise level so that the details are not buried by the hiss.

You absolutely can achieve usably good results with those mics and a basic interface like you have - no other purchase should be necessary. The problem is almost assuredly in configuration and usage. While I can't authoritatively solve your issue from here, I will say to check your driver settings and routing for your interface. M-Audio's drivers include a utility to internally route and manage audio streams, and it is possible to have the levels turned down within the drivers. If you are having the same problem both with mic'd audio and with direct inputs (such as from your drum machine), then I'm inclined to think that the issue is something like this.

share|improve this answer
    
Thanks, but how can I measure dBFS? –  Chelonian Aug 7 '12 at 5:41
    
@Chelonian In short, with a meter, usually in software. I edited my answer to clarify. –  Warrior Bob Aug 7 '12 at 14:31
    
Excellent addition. So I've done this now and realize I don't know how to really use the meter (many settings to choose from) but just using the default settings gives me something close to 0 dBFS around peak (guessing where peak is by where I see it gets cut off) and about -65 to -70 dBFS in a silent room. Which you find good SNR, right? I switched to the Audix mic for vocals and it's a bit less muddy and I think the room echo is part of the problem. I think I can just about live with this now. Thanks for the reality check! –  Chelonian Aug 10 '12 at 17:20
    
That's right. In Audacity there's no separate peak-tracking system, so you just have to eyeball it. 70dB of useful range is plenty good for home recording, especially in an untreated room. See if you can do something about those echoes, or perhaps find a different room! –  Warrior Bob Aug 10 '12 at 17:26
    
Actually, it seems very unlikely to me that the problem has to do with the interface's internal routing. Those audio streams should be 24-bit digital, so even if they're turned something like 40 dB too low the introduced noise isn't anwhere comparable to the preamp and AD-converter noise level. But if you've had such an experience... — I really think condenser microphones (not necessarily expensive ones) are the way to solve the hiss/crisp problem, the much better SNR at high frequencies you get with them (and without expensive discrete low-noise mic preamps) can easily be measured. –  leftaroundabout Aug 13 '12 at 11:04
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.