Just a request for a minor feature ‘completion’ of sorts. In the AMPLIFY dialog the decibel scale could use a concurrent % conversion translation, I just don’t relate to decibels that well but can guess at perceived volume changes as a %. For example I have an audio-video receiver that shows volume increases as decreasing decibels, I don’t think I’ve ever seen anything more stupid and it might soon get a mouthful of my axe just to illustrate the point
It’s not stupid. It may not be the way you like to work.
Sound doesn’t work in percent. If you like working with the Audacity blue waves (also in percent) please note that they only show the loudest third of audible sounds. It happens to be convenient when trying to find that one drumbeat or music sound to do it that way, but it’s not unusual to completely miss things because they’re in the bottom 2/3 that you can hear but not see.
Maybe the team could make it a user-selectable option, but I can’t see them changing it wholesale.
I have an audio-video receiver that shows volume increases as decreasing decibels
Do you like the way it sounds? Put black tape over the readout.
My guitar control knobs and all my amp ones too all show like 1-10 index markings with no decibels and do the job very nicely for millions of users all over the planet. Old stereos did the same but if I just tape this digital readout over than I will have no way to set exactly the same level for the same recording each time. So there, as for the wholesale rewrite I wouldn’t ask for that knowing full well that there are people who have a use for the db numbers. The suggested change might take the form of Db on top of the slider and % on the bottom so I can set say a 10% drop or increase that I may want. All I want is to adjust the volumes on my backing tracks on the basis of listening to them and deciding if they need to go up or down in the production copy.
What kind of amplifier do you have that doesn’t go to 11?
Made before 1984? Any trouble getting tubes for it?
A guy at work complained how wussy his amp was sounding and it turned out he had never changed the tubes. We got a new, matched set for him. I took out the old ones with surgical clamps (the bases fell apart from heat damage) and carefully put the new ones in.
He was jumping up and down happy. His amp went straight back to factory specifications. And yes, it was old enough so his Audio Taper Volume Controls only went up to 10.
There’s three variations in volume control:
Linear which is how most Windows Soundcards worked for a long time. That’s where you turn forever and nothing happens and then suddenly it does. All the control was on the last 10% of rotation. We in engineering called that a mistake, but we also assume it was cheap to do it that way. Nobody buying a soundcard ever said, “Let’s see what the highest price is.”
Logarithmic Control. You’d think this would be terrific because this is roughly how your ear works, but you can’t use that for an amp because it doesn’t go off. One of these in your amp wouldn’t let you turn it off to answer the phone.
Audio Taper. This is the one in your amp. It’s a fudge job. It’s roughly Log taper in the middle; that’s the part you’re referring to, but it’s linear taper at the bottom so it goes off when you twist it off and many of them have an “11” boost at the top even if they only say 10.
So that’s the problem. The first two are math derived and have very specific behavior as you adjust them and those are the ones in Audacity. The last one, Audio Taper, changes depending on who made it. Vishay, Bournes, BA-Technologies. All different.
We have enough vast international discussions about how to change Audacity without a Vishay/Bournes fight.
When amplifying, you need to have two reference levels.
On a typical guitar amp there are multiple controls that affect the output level. Gain, Drive, Bass, Mid, Treble, Master… will all affect the output level, as will the input level of the signal being fed into the amp. The “Master” volume cannot be calibrated to an actual measurement because the output level depends on all of the other controls and other factors. Instead of providing a measurement, it simply provides two reference points:
- 0 = silence (infinite attenuation)
- 10 (or 11) = “maximum”
Note that the upper reference is an arbitrary figure (something that Nigel Tufnel failed to appreciate). It could just as easily be 8, or 100, or 42, or it could even be A,B,C,D,E,F…
For amplification in dB, the two reference levels are:
- -infinity = silence (infinite attenuation)
- 0 = No change (same level).
Note that in this case, both reference points are fully defined. “0” is not arbitrary but means “zero” change. The meaning is not ambiguous, “6 dB” is “6 dB” regardless of the make/model of amplifier and regardless of what signal is being processed. The dB scale represents a ratio, thus one signal can be said to be …dB greater or less than the other.
One of the problems with using a percentage range is that the numbers are not very practical.
Let’s say you want to double the amplitude:
In dB that is +6 dB
In % that is 200%
Halving the amplitude:
Reducing the amplitude to 1/10th
So probably a reasonable amplification range would be 0 to 200%? (if we want to make a signal bigger then we need greater than 100%).
So let’s say that we have a waveform that has been recorded rather too quietly and we need to amplify it. Say the initial level is -24 dB.
- We will need to amplify by +24 dB to bring the level up to 0. (note that we only need to change the sign from “-” to “+” to work out the required amplification).
- The same level represented on a linear scale is 0.063096. To bring this up to 100% we need to amplify by 1584.89 %. Not only is that harder to work out, it is also much bigger than our “reasonable” range of 200%.
For many audio processes, “dB” is a practical and universally accepted way of representing the level of a signal with reference to either another signal, or a fixed reference point. In many of these cases “%” would just be impractical.
Examples of how “dB” is more practical than “%” abound.
I have a microphone which has a noise floor (the “hiss” level) of -72 dB. As a percentage that works out as about 0.0251%.
For sounds, if we use “dB” then for most practical purposes we have numbers between about -100 (dB) which is very very small, and +100 (dB) which is very very large. If we use percentages instead of dB, to cover the same range we would need “silly numbers”: 0.001% = -100dB. 10000000% = 100dB.
Whether you like it or not, if you wish to do more than very basic audio editing you need to become familiar and comfortable with using “dB”.
Of course, many users do not want to do more than very basic audio editing, which is why I made the plug-in: Amplify by percent
If anyone wants to comment about the plug-in mentioned in the previous post, please do so HERE https://forum.audacityteam.org/t/amplify-by-percent/33092/1