Take the 2-minute tour ×
Ask Ubuntu is a question and answer site for Ubuntu users and developers. It's 100% free, no registration required.

I've asked a question about what audio volume to use when recording:

In there, I learned that:

  1. I should avoid too high a volume, to prevent clipping
  2. I should avoid too low a volume, to prevent loss of resolution

The question now is: What is too high a volume? What is too low? I am setting the volume via the GUI for sound config. It has an unamplified setting, a 100% setting, and volumes beyond 100%.

After 100%, is there still resolution loss? How can I tell if there is clipping going on (given that my recording program is the non-GUI ffmpeg)?

share|improve this question
    
You should consider adding the ubuntu-studio tag to this, many studio users are quite knowledgeable about recording :) –  Nate Iverson May 24 at 12:14

3 Answers 3

You can test audio record in Audacity to see if it clips. You generally should be able to hear the clipping if it is present. There is no standard volume that I know just set it at the highest possible level without the signal clipping, that's at least my standard for recording.

To use Audacity to test the audio level:

To use Audacity to test your Audio level you will open a new project. Then locate the two stereo meters then click the arrow next to the microphone and select start monitoring. If no level is shown and you are sure you have audio playing as if you were recording a video then move to the right in the UI and find the second microphone, and click the dropdown menu and select the device you are using, in common cases it simply is "sysdefault."

Then you should be able to see your level if you return to monitoring. Additionally you can simply record a few second clip and see if the sound file spikes to the top (Peak).

share|improve this answer
    
ok. how do I use audacity to check for clipping ? –  josinalvo Sep 25 '12 at 14:38
    
I just keep the volume in the maximum (inside the program) and see if the volume bars exceed their lenght ? –  josinalvo Sep 25 '12 at 14:41
    
You should include some further instruction on how to do this. –  RolandiXor Sep 27 '12 at 2:06

The newer versions of FFmpeg have access to an audio filter called volumedetect which might be well worthwhile having a look at:

https://www.ffmpeg.org/ffmpeg-filters.html#toc-volumedetect

This will give you a number in decibels before clipping occurs (max_volume) and might be worth experimenting with, especially if you are used to using FFmpeg which I see you use for recording...

share|improve this answer

To high a volume while recording is what should be avoided at all cost. Too low of volume is avoided by increasing the bit depth of the recording. So it becomes a question of the bit depth of your audio interface and your audio program.

You should use a program with internal 32bit floating point audio like ardour. This will allow you to adjust levels in post more easily without loosing anything.

In practice 24bit audio on your interface is enough that you can record with plenty of headroom as they call it in the business. This means that you usually won't even worry about the low level being captured because with enough headroom it is there. However you may be concerned about the signal:noise ratio. In practice you should be aware of the dynamic range of the instrument or environment you are recording. Let's say a bass guitar has around 25db of dynamic range. You'd set your peak on the trim so that it peaks 5-10db below that.

The disadvantage of using higher bit depth is you have to dither your recording down to whatever level you want. For instance CD's have 16bit audio. Typically if you decide to record at a lower bit depth you will need to use a mixing tool to compress the dynamic range on the fly. Sometimes people even set a hardwall limiter on each audio track just in case there are one or two sounds that go over 100% (a special form of compression). The disadvantage of a limiter or compression is they can distort the audio, so setting the trim correctly is a matter of experience. This is one of the reasons a band will do sound checks.

Welcome to the wonderful world of audio engineering :)

share|improve this answer

Your Answer

 
discard

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.