Volume of recorded audio seems very low

Discussion in 'Microphones (live or studio)' started by Curmudgeon76, Nov 24, 2004.

  1. Curmudgeon76

    Curmudgeon76 Guest

    Hey all,

    I'm somewhat new to home recording. I'm using Cubase SX, and M-Audio Audiophile 2496 sound card, and a Behringer 8 channel external mixing board.

    Basically, my mixing board doesn't start clipping audio until +10 db, but in Cubase, clipping occurs at 0 db. I am using VST compression to try to get the loudest possible signal without clipping. I set up each channel on the mixing board, using the "pre-fader listen" function to get the loudest possible input, and then use the faders on the mixer to get the inputs as loud as possible without clipping on the board.

    However, after I record all the tracks, normalize them, and output them to a WAV file, to form one song, I notice I have to turn my computer speakers up all the way to get a decent volume.

    Basically, I can't figure out why the volume seems so low on the finish product.

    Is it bad if the audio is clipping in Cubase? I noticed that when recording, even if there's a lot of clipping in cubase, the sound quality of the finished product still sounds really good.

    Any advice guys? I've been having this problem for awhile and can't figure out how to resolve it. Any input would be greatly appreciated.


  2. vinniesrs

    vinniesrs Active Member

    May 12, 2003
    Hi Justin.

    I'll start by answering your question. Yes, digital clipping is bad, and it sounds bad and it can hurt your gear.

    It might help you out a little to learn about gain structure and the digital process a little.

    First thing to understand is that the db is not a unit of measurement as much as it is a comparison.
    zero db in analog is referenced to a voltage. .775 volts in this case. Since analog circuitry is a physical device with wires and potentiometers and such, analog circuits are capable of producing more voltage. The easiest way to grasp what this means in a practical sense is to consider that an increase of +3db is TWICE the power of whatever you had to begin with. In this example 0db or .775 volts. +3db would be 1.55 volts.

    When your mixer starts to clip at +10db it is clipping because it can no longer produce a higer level of voltage, so you get a corrupt signal.

    In the analog stage the objective is not to get the signal as loud as possible, rather the objective is to hide the noise that is inherent in analog circuitry as a result of radio frequency and magnetic interference which is induced into the circuitry.

    You want to leave some space above the normal signal level, so that if the music becomes more dynamic, and a sudden increase in power is required to represent this, your mixer has enough "headroom" to produce that power and accurately represent the signal.

    So for starters, ease off a bit on the gain in your mixer.

    Next lets look at digital.

    0db does not represent a voltage in digital audio, (although when measured at the analog output of a digital device 0db is usually around +6 analog.) rather 0db represents the loudest sound digital code is capable of representing. There is no +1 db. If digital equipment is forced to record information that exceeds 0db digital, it simply omits any information above 0 since it is not capable of reprensenting it with a value.

    Because of this you need to think differently about how loud you should feed audio into a digital device. There is less to worry about though, digital audio is noise free(basically). The only damges that can accompany your signal in digital are:
    1. clipping, from too high of an input level.
    2. noise carried over from your analog signal path
    3. Tiny errors in sampling caused by inconsistencies in your digital machine's clock.
    4. errors caused from power supply issues.

    To eliminate these things there are a few steps you can take.
    I like to input to digital devices at -15db to -10db. Since I have taken steps to ensure my analog devices are not too noisey and the signal of my musical information is louder than what noise there is, no additional noise will be on the recording, and no clipping will occur.
    That takes care of step 1 and 2
    Don't worry about buying an external clock right now, that makes only a little difference.
    Put all of you digital devices on a power conditioner. One that is capable of supplying additional voltage if the line current drops below, maybe 108v or so.
    If your digital device starves for power it will create problems.

    Hope this helps.
  3. Curmudgeon76

    Curmudgeon76 Guest


    Thanks for giving me some insight into how this works. I have reconfigured my mixer, my sound card, and cubase so that the signal peaks around -6 to -3 decibels and does not clip.

    Like I was saying however, the finished prodcut from recording seems REALLY quiet. If i play it on my computer or thru my stereo the volume seems MUCH lower than a commercially produced CD.

    Is this a mastering issue? I know most artists don't master their own work, but I'm curious as to how I can get a louder finished product that I dont' have to crank my stereo to hear well. Any suggestions?

  4. Yes, this is very much a mastering issue. If you feel the need to crank the hell out of your tracks you'll need to compress the mix, either with a standard compressor set to limiting or a dedicated master limiter like the Waves L3 or the Voxengo whatchamacallit... Lampthruster, methinks.

    Just remember to keep a copy of the un"mastered" 24-bit mix lying around - whatever you do to your song WILL cause permanent damage (in that it will sound like (really loud) crap when you're done) and some day you'll probably wanna redo it/have someone else do it for you.
  5. Curmudgeon76

    Curmudgeon76 Guest


    Yeah, I've read some stuff about limiting and what not. Basically, I'm using a software compressor when recording the tracks to try and get the loudest sound possible without clipping.

    It's not that I feel the need to crank my songs, I just want them to have an acceptable volume level so I don't NEED to turn them up loud to hear them.

    I'm using software compression, like I said, In cubase when recording tracks. I am going to buy a hardware compressor eventually. One question I have is, do I run the pure signal through the compressor first, and then thru the mixer, or thru the mixer first and then the compressor? Also, if anyone has any suggestions on using a software compressor to improve volume levels and tighten up the mix, that'd be great.

    Like I said, basically I'm just trying to get the best volume possible without clipping, and I'm keeping the signal on my mixing board, sound card, and Cubase around -6 to -3 decibels. Still, I end up with a really quiet finished product.

  6. Massive Mastering

    Massive Mastering Well-Known Member

    Jul 18, 2004
    Chicago area, IL, USA
    Home Page:
    Stop, stop, STOP doing this.

    Try something -

    Start a new 24-bit recording of a song you've done before. Make each track, kick, snare, guitars, bass, vocal, whatever - PEAK at -6dBfs. Throw up a mix, compress or limit at the track level where it's needed - NEVER EVER for the sake of sheer volume. You compress when the dynamic range is too wide. Not to "squeeze bits and make it loud."

    Make that mix PEAK -6dBfs. No limiting, no excessive bus compression (see the previous point). Marvel at much more open and dynamic it is than your other mixes.

    NOW start compressing and/or limiting the mix, and marvel on how much louder you can make it.
  7. vinniesrs

    vinniesrs Active Member

    May 12, 2003

    Justin. compression is an art. It will take you some time to get the hang of it, but this may help you out.

    Consider that you are not making your mix louder, technically. (well, technically you're the rms level higher.) The AVERAGE volume will be greater, but since it's digital you can't exceed the 0db mark, so your peak level shouldn't change much.
    Commonly compressors are over-used or improperly used, resulting in a dull, or squashed sound.

    If you can, use bandpass compression or limiting. This will allow you to apply different attack times and compression ratios to different frequency ranges in the song. For example, you could use a lower compression ratio for the bass sounds, with a slower attack and a fast release time, allowing for more dynamic, punchy bass. Other parts of the music which are less dynamic could recieve a faster attack time, higher compression ratio, and slower release time to improve the apparent loudness. When dealing with transients like a snare drum for instance, you need to allow enough delay on the attack time, to allow the initial sound of the snare to "pop". If the compressor "attacks" or starts compressing too soon, you will lose this important part of the sound. Sounds like snares that have a very brief but loud sound characteristic are the sounds that make a mix ballsy. The same applies to all dynamic musical information.

    Dont get carried away, though. Be creative and move slowly. With some experience and careful listening I think you will be happy with the results.
  8. Curmudgeon76

    Curmudgeon76 Guest

    Thanks for the info,

    I really like recording my own music, and want to get better at it. Any insight you can give me is greatly appreciated.

    As far as an external compressor (which i plan to buy)...do I run the signal thru that FIRST and THEN the mixer, or the mixer first, then the compressor, then the comp's sound card?

    Are Behringer compressors any good? I have a Behringer mixing board and it's ballsy, for what I'm using it for.

  9. Curmudgeon76

    Curmudgeon76 Guest


    OK, I get the idea. I appreciate your help, but please be patient with me because I'm learning about all this, for the most part, on the fly. So record each track so that it peaks at -6db. As for compressing individual tracks if the dynamic range is too wide: how would I know if the dynamic range is too wide? Are there certain instruments that benefit more from compression than others, and is their other criteria I should consider?

    After I've recorded all the tracks, each peaking around -6db, I then mix down with that as the peak, correct? And then I reopen the mix, and compress it on the mix level, to tighten it up and make it louder? Is that right?

    Can you recommend any general guideliness when comrpessing/limiting the mix to get optimal volume while still preserving a good dynamic range, i.e., what settings might be good in the compressor/limiter section? I'm using the "VST Dynamics" plug in in Cubase, by the way. Do not have a hardware compressor as of yet. Thanks,


Share This Page