A Cappella Generator

[jdeligiannis]

There are two methods that I am aware of which can generate an a cappella.

1) Inverting the instrumental - If you have an instrumental of a song, maybe from a CD single, it is possible, but not probable, that you can make an a cappella out of the main mix. If the two tracks are synchronized, simply invert one of the track's waveform and mix it into the other. Everything that is identical in the two tracks will cancel-out, and the only thing that is identical is the beat.

The problem with this technique is that most of the time the two tracks will have alternate speeds, and will therefore be unable to maintain synchronization. For those who have tried this with Tupac's instrumentals, you know very well that just about every single instrumental from the CDs does not maintain synchronization to the main mixes, foiling the brilliant intentions.

2) CPU Generators - The a cappella of tracks such as “Pain,” “Soldier Like Me,” “Ghetto Star,” “First 2 Bomb,” etc., were not created by use of the above technique, but rather with some kind of generator, which I assume cancels-out certain frequencies. Can anyone post instructions on how to create a CPU a cappella?

Thanks.

 

[Kevsta2004]

You might wanna try looking for some VST plugins

 

[goymz]

I've heard the best VSTs for the job are 'spectral plugins'... visit www.soundhack.com

 

[Dante]

vst plugins are still somewhat noobish, and will never give you satisfactory results. i still don't understand how people think a hollow, panning cpu gen with all kinds of digital artifacts is any good... anyway from my experience there are some hardware options that will trump personal computer oriented plugins any day of the week. mainly industry people use standalone acapella or instrumemntal generators, which look for the variable tones (vocals) among constant ones (instrumental rhythms/loops) and likewise isolate. alternatively there are filter processes that i have only seen in-studio that were pretty cool... the tech called them a cathode filter, but i think he was referring to the physical architecture of it and not the commercial name. anyway, this filter uses a learning curve, and has several preset vocal ranges as well as the ability to learn one (so i assume you load acapellas of an artist and it creates a profile, then compares these harmonics to the intended song and subtracts the variance). i haven't used it, but i have seen the end result and it was pretty damn good.

end of day, there's no real answer. industry cats don't use generators for anything but to kill the instrumental in samples for cuts or breaks that are probably only 2-5 seconds anyway. trying to make a whole song from a gen and have it rival studio quality is typically just a diamond in the ruff, needle in the haystack deal.

one thing that may be of more use than a generator is the timing vst/applications. timing modules are also on hardware mixers (my old roland had one) and basically it looks for snare hits as time markers, and when you layer track on track it will make the minute tempo shifts between snare hits to keep both songs aligned. as john stated in his initial post, cd single and promo instrumentals are usually 1 to 4 percent offtempo from the vocal mixdown tracks and so a module like this may help keep the speed without the mind numbing cutting and pasting.

 

[Hymnz]

if you know what youre doing you can get awesome acapellas.
google will help you with this stuff.
if you have the instrumental you can start with reversing polarity etc, but as dante mentioned, not all beat/song combinations (even if they are released together) have the exact same tempo OR layout.
there are also many other widely known ways to extract vocals, and on top of that, many 'secret' methods employed by people who are really serious about this. some of them arent so much secret as labor intensive .
then there are a couple of n00b methods, among them being the Music Morpher program. i have bought it out of speculation and must say it is a joke. there are like 8 sliders you can move around but nothing can get close to the top manual methods. the 'heuristics' feature is interesting but inaudibly effective.
google around (we all did)...

 

[Dante]

i forgot to mention, there is one other common way to generate acapellas, which i have found to be hands down better than the invert method in many cases. dunno about you peecee windoze people, but on a mac you can sum/diff r and l channels. for this you make stereo tracks from the left channels and right channels of two tracks - lp and instrumental, then sun a sum/diff on it. a sum/diff looks for the differences between the two tracks and adds or subtracts them. obviously this is similar to the invert method which is 2 + -1 = 1, this is 2 - 1 = 1, however the results tend to be a lot better. for example, i was able to generate an extremely convincing dutm stereotypes mix without any of that nasty hollowness and digital distortion people think is acceptable...

i'd never use a cg with all that metallic crap. are people on crack?

 

[Preach]

i'd never use a cg with all that metallic crap. are people on crack?

when the compact disc format was introduced, things took a new turn. i'm not telling you this dante, you already know, i'm telling other people in this forum this because it's something to keep in mind.. the dynamic range of the sound that a record player produces is way bigger than that of a cd. this is what is referred to as the warmth of vinyl which is supposedly what makes vinyl so much better than cd. a matter of taste, i don't really give a shit if i listen to cd or vinyl. i have to say though that playing vinyl sets a better mood with the deck spinning and what not.

anyways, to the point. being that the dynamic range of recorded audio was reduced with the release of the cd, what people now consider a top notch quality recording doesn't have the same wave complexity as an old vinyl record. the definition of top quality was re-defined and people now settle for lower quality. this is what has happened to the tupac scene. due to the degradation of sound quality on original and unreleased tupac songs, and the remixing and og-vibing and og re-doing using cg acapellas, layered remixes and all kinds of bullshit has re-defined what "good quality" means to us. people on the tupac scene simply settle for lesser quality because after having been listening to poor quality shit for years, a tape rip, for instance, might not sound so bad in comparison.

people really need to open their eyes when it comes to the cpu acapellas and remixing of tupac tracks. i NEVER play remixes, og vibes (sorry dante), or anything else using cpu acapellas anywhere but in my computer headphones, and that's just simply to satisfy my own curiosity of what it would sound like. honestly, when whoever it was on this board first got the idea or first discovered how to make cpu acapellas, he fucked it up for all of us. luckily for myself, though, i have since been enlightened. stop with the cpu acapellas lest you doom the tupac scene... FOREVER!

 

[Think_ThuGLiFE]

end of day, there's no real answer. industry cats don't use generators for anything but to kill the instrumental in samples for cuts or breaks that are probably only 2-5 seconds anyway. trying to make a whole song from a gen and have it rival studio quality is typically just a diamond in the ruff, needle in the haystack deal.
.

Very tru



but they used a CPU for biggies Duets album "WAKE UP" feat korn.... WHo would release a song wit a cpu on a retail release. lol when i hurd it , i was shocked they acctually did that.