Hi everyone! I am new to audio editing, and I had come across this problem yesterday. I have a violin audio that I’d like to piece together with different takes. I used a video editing tool (Wondershare) to try to piece together the video initially, but there is a click sound in between the two excerpts that I’m piecing together. I went into Audacity and tried to select the “click” portion, use the Z key, and delete, but whichever way I cut the click sound, a new one qppears in the new transition between the audio. How do you guys cut an unwanted noise without creating new ones?
Also, I tried click removal, but the noise in my audio isn’t affected.
I attached the audio file; the noise is approximately at 2:10
Video editors don’t actually cut audio. They do a rapid dissolve which solves a lot of these problems. They also have the problem I’m sure you noticed that you can only cut on video frame boundaries.
I’m always joining bits and pieces of solo takes together. Z key on initial selection or repair usually works for me but I also use the following method
I place each segment I want on a different track in Audacity, then slide each into position.
Zoom in a long way and apply fade in/fade at either end respectively. Just fade the shortest length possible (it won’t be audible on playback and will smooth the transitions).
Zoom right in when you slide each segment (each on a separate track) into place until you end up with an inaudible nanoo second overlap
But if you do end up with clicks, the easiest way I know to find them (in Audacity) is to roughly select where the click is in normal view, then switch to spectrogram view (in the dropdown menu by the track name)
and progressively zoom in from there - you’ll likely see a line or spike where the click is - which broadens as you zoom in. Then switch back to waveform view.
Highlight the deformity and then use “Repair” from the effect menu.
Tapeheads advice, I think, is the best solution. Placing different parts of files on separate tracks, then editing them together via the use of fade-ins and fade-out has worked. It can be consuming, but it can also work wonders.
That said, if you go this route, I’d advise you to be sure to save your work as an Audacity Project (AUP) file so you can get back into it and tweak it, if needed. There is a reason for that: Occasionally when we’re editing this way, we get too close to the file and we start to miss minute differences in timing that creep in by blending tracks.
So what I do is finish my edits, then go out for a few hours and do something. Then I return re-listen to the file. It’s only then that small unintentional timing errors sometimes become evident. Since I have every segment on a discrete track, these are then easily corrected. I learned the hard way not to scrunch it all down to a stereo WAV file as soon as I (thought I) was done!
How I wish Undo could be saved! I’ll take the hit on HD space.
I heartily agree with taking a long break Tony - I don’t always follow it though.
Yes timing can be tricky. Several times I’ve edited a solo together over a backing track and thought it sounded O’K
But then I’d do the same thing using the freeware dj software Mixxx which has excellent beatmatching and I’m often amazed at the difference (I’m not into dj style music or mashups far from it).
The beta 1.12 version from their support forum is far better at this than the quite old now official release.
So even if the timing of the joined up solo or instrumentation is ever so slightly iffy Mixxx will adjust the timing to compensate and you can get a good reference to manually make further adjustments from.
But this only works for up to 40 seconds or a minute or so because accurate beatmatching isn’t possible for an entire song yet. I mention this because I’ve tried Mixxx and other softs to retrospectively add drums to antique rhythm tracks I recorded without a back beat for reference. It is possible but it takes around three runs of Mixxx starting at different stages of the track and then crossfading the resultant Mixxx mixes together.
Ahh kozikowski, I see your trying to do AFSK as well it looks like(and I would love to know how you got that)…
anyways, the disruption of the waveform in signal processing usually for me is the sample rate/filtering. It stinks. Thats usually what I have to deal with on a daily basis
