I have an audio editing question! My wife got me a "Cameo" video from Paul Eiding for my birthday which was absolutely awesome! It can be viewed here:
There is music playing in the background while Mr. Eiding is speaking. The music is awesome, but I was wondering if it'd be possible for me to remove the music somehow (and leave the voiceover intact) using editing software. (Note: The music was literally playing on a radio in the background as opposed to being added in with editing software).
I have no idea if it's possible to do this or not. Reading about it online, most people want to remove vocals from music, but ironically I want to do the opposite haha and I can't find anything online about that.
Here are the editing programs I have:
Sony Vegas Pro 12
Sony Sound Forge 10
Sony Production Assistant 2.0
ACID Music Studio 9.0
HitFilm 3 Pro
If anyone could provide any guidance, it'd be greatly appreciated!
Thanks so much,
+ Reply to Thread
Results 1 to 20 of 20
Something playing in the background, not mixed digitally, would be one of the hardest kinds to get rid of. It also has the same acoustical character (based on the space) that the person speaking has, so no way to differentiate spectrally. Because it is in the background, it may have a lower level, and there are "noise gate" apps/devices that can somewhat improve that, but at the possible cost of noticeable pumping of the background acoustics and possible changes to the foreground audio as well.
Had it been digitally mixed in post after the speaker's recording, you could just find a copy of the recording and do a mix-subtraction, but that's not possible with environmental/background music.
Some AI/Neural-network stuff might be able to, but it's not cheap, nor easy to find, nor to use.
If I were you, and this were a "commissioned" cameo recording, I'd ask for a re-recording or your money back.
It would also be possible (to an extent, at least) if the music were in stereo. Even though you have two channels in this file, they are identical.
I doubt if this would work effectively on a mono recording.SONY 75" Full array 200Hz LED TV, Yamaha A1070 amp, Zidoo UHD3000, BeyonWiz PVR V2 (Enigma2 clone), Chromecast, Windows 7 Ultimate, QNAP NAS TS851
Take all those apps with a large grain of salt, unless your clip falls under a very narrow range of circumstances.
I only stumbled on this today, i have no personal experience. But iZotope is bleeding edge with this kind of tech
On this page it says it works on mono or stereo
Also there is a free trial good for 10 days
Last edited by davexnet; 21st Jun 2019 at 01:30.
Here's another example, this time removing the vocals. Mono track.
Yeah, but that's not MUCH better than can be done with Phase-Inverted-Stereo center channel extract, which leaves more of the vocal but doesn't have ANY DSP artifacting. (Used the stereo PL for reference).
other in the case where you don't have a discrete channel mix tape.
Is it possible to take a mono track and by any duplication of channels/phase inversion
do anything like this?
You could do an eq kind of thing, similar to the Lch=highs, Rch=lows of some 60s recordings, but it's totally not worth it. Believe me, I have tried! But I've seen possibilities through multichannel, eq optimized, side-channel gain-riding combinations. I would reserve that kind of effort for one-of-a-kind material though, because it is so labor- and time-intensive.
Your example would have one kind of use: I think it would be decent enough for background behind karaoke vocals.
Yes I've heard some of the enthusiasts efforts. Some are quite good. There was a clip on Soundcloud of the Beatles "If I fell".
Somebody had removed McCartneys voice leaving John Lenons vocal + the music. This song was supposed to have been recorded
with the pair of them singing around a single mic. (How was it done? Magic?)
Edit - I found the clip
That one's more simple. Back when they recorded these in the studio, they didn't as yet have true multitrack tape, but they did have 2, 3, or 4 track recorders. Beatles used EMI/Decca studios and they had 3 and 4track ones.
So all stuff was first done on these, and sometimes they would bounce down, but sometimes not.
Then they would mix down to mono for the release master.
When stereo started becoming popular in US, Columbia took the masters and did those weird 60s L/C/R wild separation releases - without Beatles' approval. Wasn't until mid-/late-60s that Multitrack (8, then 16, then 24...) got available, and also around that time the PanPot got refined, so the mixes started finally sounding more naturally blended.
When the Beatles' albums were re-released for CD (mid80s) they this time stipulated that there be no messing with the release mixes, so all the CDs (I have every one) were done like the original british versions...so, much more mono, but tighter mix.
In 2000s, for DVD/SACD/new compilations George Martin, his son, & Geoff Emerick remastered many tracks and so some titles had true stereo mixes for the first time then.
I have a hunch this is a L, or R channel-only mix from one of those old US releases. That would account for the good quality overall but the tinny reverb sound of Paul's vox.
Last edited by Cornucopia; 22nd Jun 2019 at 21:20.
been together. I've never heard a version where the voices were separate in any way.
Let me know if you find such a version!
The wiki pages mention that the intro was double tracked and used on the stereo version;
mono version being single tracked voice. But as I mentioned, the legend has it, that they sang
the rest of the song together, at a single mic, at the same time.
You know what this implies - it's like the Russian dub voice over, with the English in the background -
can't be separated.