Hi there,
A quick question, I’m using Spectralayers 10 Pro.
I have an audio file I’m working on that has a drop out (about a tenth of a second), I realise that there is a heal function in Spectralayers (which I did try out of curiosity, but I think the dropout is too large).
What my question is - I have a version of this audio I copied (… 12 years ago) before my tape was damaged. What I’d like to do is paste the section from the good recording over the top of my damaged recording
What I’m imagining is that as the spectral signature for each section is like a fingerprint and unique, I’d like to be able to copy a section from the good audio file and then hover my selection over the damaged area until they line up. A bit like having two negatives or transparencies and and moving them sideways until they match.
(I did have both audio files opened in the same Spectralayers session and I did work out how to export the “good” section of audio and import it as a layer into the damaged track, but as there doesn’t seem to be an easy way to slide my imported selection sideways – I’m kind of stumped – I’d like to be able to do this visually if I can.
In my case both audio files are different in volume and one starts earlier that the other etc (so I know there are other issues I have) and I can probably “paste” in my good audio for the drop out with an external audio editor, but I thought I’d pass this suggestion on as the idea may be useful. (Please let me know if I’m missing something obvious).
Anyway, that you very much for the work you have done on Spectralayers 10
Assume the non-damaged version isn’t an identical file, you can EQ match the non-damaged to match the damaged, then cut and merge into the space and sample align. Depending on difference it’s sometimes best to phase invert one to find the quietest position, then merge (and disable phase inv obviously). Ive done this a lot with various masters where there’s been some clicks, pops even audible tape slicing or tape damage etc and replaced with a section from elsewhere in the song.
Thank you for your advice. I did think after I posted the 3rd screen shot I was thinking that I could delete some time at the start of the second layer to nudge that layer sideways until both layers align/match (was hoping I could drag my second layer like an object). It turns out that del time affects all layers.
Have just discovered there is a Time shift function under the Transform Tool that lets me slide my selection around in its specific layer.
Great to get some advice from someone with a bit of experience of having done this before … both files are the same sample rate, though different volume amplitude as you can see in the wav view at the top of the screen.
Thanks again for the tips, I’ll keep them in mind …… you mentioned some things I hadn’t thought of yet
Cool yes, I typically just zoom right in and use the Transform mouse tool instead of timeshift but same result, just I like the control of the mouse.
When you EQ Match it should match amplitude too. Essentially an EQ is just frequency isolated volume faders no different than on a mixer, it’s just the mixer is a full-band EQ on one fader essentially.
Whenever I’ve historically had a mono, two-track, or 24trk tape with a dropout…in the op case, a tenth of a second… and have to deal with it…
I’ve always transferred the tape into Cubase/Nuendo as is, create a second empty track, use my ears to locate a snip section to copy, copy/paste to the empty track, use my ears and maybe a bit of crossfade, get everything nice & tidy, then export the new file.
Which for me, takes about five minutes or so.
And then load that into SL for all my own particular further surgery.
It’d drive me a little nuts to try and do that as fast within sl.
I had thought too of the traditional way of pasting the good section using Cubase etc, I finally got my layers aligned in Spectralayers at a common point in the song. I did have a nagging suspicion that when I had originally digitised the tape I had used a different tape deck to the one I used years later. My suspicions are confirmed as there is a slight speed difference between the two recordings.
I guess if I can determine the note difference in cents between the two recordings, I can adjust until I’m happy with it (the drop out is short enough that I could probably get away with it). But as there a backing vocal that comes in just at the point of the drop out, I’d probably have to paste in larger section from the other file and this is where the difference becomes more noticeable.
There is a time strech / pitch shift function under Transform that I’ll play around with.
Just discovered you can time stretch one layer as both layers are playing in real time by shrinking / expanding your selection – already got close, just from a few tries – I’m exited now.
