Is it possible to even out the audio pitch?

Audacity Help Forum Windows
1

So I couldn’t find this question anywhere, but I have a group of audio recordings that I’m putting together, the problem is that they all have varying audio pitches. I’ve tried adjusting the pitch of each piece of audio individually but I can’t manage to get them to balance up correctly. Is there some way to highlight all the audio and click some setting that will automatically make it so that all the audio is a similar pitch (I don’t need exact, but once they’re all a similar pitch I could just use the change pitch setting to get the sound I want)

I’m currently using Audacity Version 2.1.1 on Windows 10. I don’t have the installer for Audacity on my computer so I can’t tell you whether I downloaded this through the .exe or zip. If you need more information, feel free to ask. Also if I can’t do it in audacity, could you advise a program that would be able to do this?

JamieDaNerdKing

2

I presume these are music recordings?
Are they single instrument recordings or multiple instruments?
Are they original recordings (voice / instrument played by you) or music that you have bought/downloaded?
Are the recordings just “a bit out of tune” with each other, or are they in different musical keys?

3

Yes, they are all music recordings of a single instrument (Though more specifically, it’s of a person humming) I made the audio myself with audacity and I was curious if they could be made the same pitch, thus why I asked. Regarding how out of tune the audio is, if I could describe it in numbers (Because I’m not that into the musical side of things, nor do I have any idea what pitch they are set to already), the first audio would have a pitch of 1, the second has a pitch of 4, and the final piece of audio has a pitch of 2. So describing it in this fashion, I would be trying to balance it out so I had an average volume pitch of 2.33. Is there any way that I do this, or have I missed something where if you highlight the recording it tells you the average audio pitch of it, and then you can balance it out yourself.

Hopefully this information helps you out steve

4

If you are humming a tune (more than one note) then getting the “average frequency” will not help. You can use “Plot Spectrum” (in the “Analyze” menu) to see the frequencies that are contained in the currently selected audio. If you select just one note from a recording, you will see that there are many peaks - lots of different frequencies which create the timbre of the sound.
fullwindow-Frequency Analysis-000.png
By placing your mouse pointer over the spectrum plot, you can see what the frequency is for different peaks. In the picture above (me humming an “E” note) you can see that the dominant peak is at 355 Hz. If I move my mouse over the lower frequency big peak, the frequency shows as 168 Hz, which is an “E” one octave (half the frequency) below the tallest peak.

Note the settings in the Plot Spectrum picture, in particular the “Axis” setting which spreads out the display so that you can see the peaks more easily.

If you repeat this with an “E” note for another recording, you may find that the peaks are in slightly different places, perhaps at 326 and 163 Hz, which is still approximately an “E”, but the two recording are not in tune with each other.

The standard frequency for an “E” is 329.63 Hz / 164.81 hz (see: http://www.phy.mtu.edu/~suits/notefreqs.html). This is the “A 440” standard, which is the most commonly used standard. It is also known as “concert pitch”.

To make the two recording “in tune” with each other, I could select the first recording and use the “Change Pitch” effect (http://manual.audacityteam.org/o/man/change_pitch.html) and use the “Frequency: from / to” text box controls with a “from” frequency of 355 and a “to” frequency of 329.63. I would repeat this with the second recording and change from 326 to 329.63Hz. Both recordings should now be in standard (concert) pitch tuning and should sound “in-tune” with each other.

5

Alright, that seems to have fixed the problem, thanks for the help steve. Since we’re on the topic, would this work if I had recordings of myself talking in different pitches and wanted to fix them, or would I have to use a different method from this?
(Note: This is assuming that I have no idea what pitch my recordings are at already.)

6

The basic idea works for any sound, but in the case of “talking” it may be difficult or impossible to see any definite pitch - “talking in clearly defined pitch” is called “singing” :wink:

A tip: It’s a bit harder to use, but the “Sliding Time Scale / Pitch Shift” effect can change the pitch with better sound quality than the more simple “Change Pitch” effect. Note however that the “percent change” setting as determined by the Change Pitch effect may be typed into the “pitch shift: %” controls in “Sliding Time Scale / Pitch Shift” http://manual.audacityteam.org/o/man/sliding_time_scale_pitch_shift.html

7

It might be easier with talking because the natural pitch of your voice is more-or-less constant. If it’s not constant from day-to-day, it may not be the pitch, but the tembre (the harmonics & overtones that make different people sound different, even when the pitch matches). If you have a cold, the pitch of your voice may change. If you’re under stress, the pitch and/or tembre may change, etc.

With humming, there’s a good chance you’re hitting pitches in-between notes (especially if you are not a trained singer/musician) and although you can pitch-shift to get some notes on-pitch, other notes will be off-pitch.

If you are a trained singer/musician, or if you just have a “good ear”, the relative pitch between notes may be OK, and you can pitch-shift to make all all of the notes correct. If you are very lucky and you have perfect-pitch, you can sing/hum the exact notes without a reference or backing music as a reference.

There are applications such as AutoTune or Melodyne (and some free ones) designed to make note-by-note corrections. They can automatically pick the closest actual note (in the selected scale) or the notes can be corrected one at a time, manually or with a MIDI reference.

Pitch refers to the perceived frequency of a particular note. All real-world sounds are composed of many frequencies, and the pitch is usually determined by the lowest, or dominant frequency-component. A note has a pitch. A song doesn’t have a single pitch. A song has a key, which is (usually) the dominant or starting-note (or chord) in the song.

There are 12 notes in an octave (in American/European music) but most music is written in a particular scale, which is a subset of the “available” notes. And, some instruments can’t play all 12 notes.

The A = 440Hz standard is a tuning-standard. So, an A on one instrument matches an A on another instrument (and all of the other notes match too). That doesn’t mean the song has to be composed or performed in the key of A. 220Hz is also an A, one octave down, and 880Hz is an A, one octave up.

8

A free plugin called DTBLKFX can force the pitch of the left channel to follow the pitch on the right channel , (a bit like auto-tune) …

DtBlkFx settings to shift pitch of left channel to match right.png