2

follow-up question: What can be considered too high or too low volume?


I've been using Ubuntu to record some screencasts (kinda. Khan-style lectures, really)

I want to know what volume setting I should use for the microphone for optimal results.

I am guessing that "base" (in pavucontrol) or "unamplified" (in the ubuntu volume control), and increasing the volume later in ffmpeg would be optimal for quality (as compared to just using a higher setting).

Is that true?

Bonus points: are "base" and "unamplified" the same? What does the "100%" setting in ubuntu volume control mean, given that it is different than "unamplified"?

2 Answers 2

3

When recording, especially to digital formats like this, it is important to monitor the recording levels, whether from a microphone or line input or whatever, and ensure they do not clip. Most recording apps have a level meter that clearly indicates the max level you should allow. By adjusting to meet this level, you will be able to determine whether it needs to be amplified or not.

The other thing is to not have the level too low. This allows background noise/hum/etc to increase in your recording. It also wastes some of the resolution of the audio capture device (the analog to digital conversion). This is why you should avoid having to increase the volume level later in ffmpeg or whatever processing you do.

So, set it to be close, but NOT exceeding the max level before clipping.

4
  • Thanks. I am trying to understand your answer... * What does it mean to 'clip' ? * I'd guess this 'loss of resolution' applies only to levels of volume lower than 'unaplified' (after that, the microphone is already giving its best signal, and just stretching it). Is that true ? * (this question only makes sense if my assumption in the question above is true) If I am already using a level between 'unamplified' and '100%', would it be best to just use a higher level of volume, or to use a lower one and using ffmpeg to increase later ? (as far as quality is concerned. Sure one way is easier ...)
    – josinalvo
    Jul 23, 2012 at 17:07
  • 2
    'Clip' is an analog audio term referring to to the inability of the amplifier to produce a large enough signal when the signal gets too large, so the top (largest part) of the audio signal gets clipped off. In digital recording it also means the analog to digital converter has reached it's maximum number, and effectively the same thing happens. The loss of resolution is also related to the analog to digital conversion, and no matter whether it is amplified or not, it is still whether the signal pushes the level meter close to the max level or not. That's what you must aim for.
    – Jazz
    Jul 23, 2012 at 22:15
  • cool. I just set the volume, and I am trying to see if i'm clipping. Audacity tells me that. However, it seems I am clipping even when I set the volume very low in alsamixer...
    – josinalvo
    Jul 31, 2012 at 14:04
  • (what I really want is to just use audacity as a clipping detector. The recording itself will be done by ffmpeg)
    – josinalvo
    Jul 31, 2012 at 14:32
0

The best setting really depends on your microphone. Unamplified means that the signal from the microphone is caputered without any amplification by the audio card. 100% means full amplification. Everything above is overdrive. You should avoid that. Experiment a little with settings from unamplified to 100%. One in that range should suit you. Listen to your recording and see if you can be understood easily and only little unwanted backround noises are hearable.

PS: I have no idea what 'base' is ;-)

2
  • Thanks! Do you know if finding such a spot, between "unaplified" and 100% is just as good as leaving the control on "unamplified" and boosting the volume after, by software ?
    – josinalvo
    Jul 22, 2012 at 22:20
  • You find the answer to that one in the answer of @ubnewbie2 so that one gets +1 from me ;-) Jul 23, 2012 at 10:35

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .