Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
thats what i thought, the video claims it was done from the OG song, not session tapes or something
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
I'm sayin'. that is incredibly f*cking accurate. If they're somehow manually isolating the waveform is the fact that it's a synth line make it easier to do that? That synth wave is pretty much a single line with no harmonics or anything. I can't see doing the same with a horn section or a guitar.
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
I'm sayin'. that is incredibly f*cking accurate. If they're somehow manually isolating the waveform is the fact that it's a synth line make it easier to do that? That synth wave is pretty much a single line with no harmonics or anything. I can't see doing the same with a horn section or a guitar.
I was thinking the same thing, that's a very flat/pure tone. I can't see something like this being possible on instruments with dynamic sound/tone, it's just science.
sad to say, but theres no multis available on westbound yet. i know for a few labels there will be none at all.
if you listen close , you can still hear the section being solo'd faintly in the track.
alot of hip hop acca's as well as a few of the michael jackson accas were created this way. there are multis for rock with you and dont stop til you get enough.....the rest im unaware of.
Wow, the developer seems like an enormous douchebag. This looks like just an FFT / inverse FFT with a little UI tacked on. It's the most basic signal processing math there is. Definitely not "revolutionary", "innovative", or "unique".
I will say that his color choices for the interface are really pretty. Much better than Metasynth, which is more powerful but has always been deliberately foul-looking.
theres video of him "drawing" drum/drumkits using photoshop but that shit sounds JUST like an 808 kit
Haha, I'm not sure if it was intended as a compliment, but I'll take it as one. I wasn't very impressed with my own drawn drums to be honest...
That's pretty wild. Is this program is available for Mac?
Working on it right now, although to be honest it's more like I'm struggling with setting up my dev environment.
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
Hahahaha, this thread is epic. Yeah, you're right, I'm the only guy on Earth who has the tape of the original recording of Funky Worm with only the synth on it ;-) how lucky is that? Someone call Biz Markie, I'm about to become rich!
There's a Mac program called Metasynth that's been around for years that does the same thing.
Yeah, it does the very exact same thing, mmmh except maybe for the fact that it doesn't process recorded sounds, let alone using images.
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
I'm sayin'. that is incredibly f*cking accurate. If they're somehow manually isolating the waveform is the fact that it's a synth line make it easier to do that? That synth wave is pretty much a single line with no harmonics or anything. I can't see doing the same with a horn section or a guitar.
No harmonics? Oh if only it was true. Look at the video, you see all these parallel lines? They're harmonics. I had to keep up to 20 of them. If it had no harmonics it'd sound like a sine wave. The fact that it was a synth helped knowing where to look for things, but it doesn't help that much, the fact that it had a shitload of harmonics that were getting lost into the noise of drums didn't help a bit, neither did the fact that it's really full of curves.
Horns and guitar have different challenges to them, horns being the fact that they too have lots of harmonics that you need to keep, and on top of that their lines aren't smooth. Guitars have extra noise that you'd better make sure to keep if you don't want the result to sound like a cheap synth, but I don't think they would be that hard, that is if we're talking about plain old acoustic guitars.
But I get that a lot, that's why next thing I'll do will be an easy tutorial pertaining to removing vocals.
Wow, the developer seems like an enormous douchebag. This looks like just an FFT / inverse FFT with a little UI tacked on. It's the most basic signal processing math there is. Definitely not "revolutionary", "innovative", or "unique".
I will say that his color choices for the interface are really pretty. Much better than Metasynth, which is more powerful but has always been deliberately foul-looking.
Hahaha, an enormous douchebag, right on. Too bad you're pretty much entirely wrong about it being just an FFT/IFFT (although you really mean a STFT/ISTFT). If it was based on STFT, then the time resolution would be fixed, which means that in the lower end of images things would get blurry vertically, not horizontally. It actually uses something better adapted which is basically just frequency domain windowing with modulation by masking envelopes, which allows to do anything you want concerning frequency/time resolution depending on the frequency, but also on the frequency scale. That's how here you can get natively a logarithmic frequency scale (or pitch scale if you prefer), whereas STFTs are strictly linear, which is pretty impractical.
But go ahead and reproduce the results with any other program, since it's such basic processing, and definitely nothing innovative ;-).
Pretty easy to do with the original session tapes. I doubt this was done any other way, it's too clean. If I'm wrong that is going to be a game changer.
Hahahaha, this thread is epic. Yeah, you're right, I'm the only guy on Earth who has the tape of the original recording of Funky Worm with only the synth on it ;-) how lucky is that? Someone call Biz Markie, I'm about to become rich!
An instrumental of this song was released (bootleg) so it's not exactly crazy to assume that the master tapes are available to certain privileged folk. As to whether or not the song was multi-tracked is beyond my knowledge though. Let us know when you can isolate and guitar or bass riff as clean as this, that would be really impressive.
the whole thing is a fix due to pulling out a synth. that's not a complex waveform.
redo the demo and prove us wrong. i can't remember the track but there's a RAMP joint where Roy Ayers ruins it trying to sneak a fart. remove that and i'm a believer.
lol, not this time, someone on the-breaks.com pointed me to the insanity of this thread
Seriously... this shit is crazy... I am very interested once there is a MAC version available!
PM me your e-mail address, I'll add you to the Mac mailing list
Let us know when you can isolate and guitar or bass riff as clean as this, that would be really impressive.
Looks like I'm going to have to do it since so many people think that isolating synths is just the easiest thing on Earth and that anything made by plucking actual strings would be impossible... Do you have any particular sample in mind? i.e. anything that people would have loved to sample?
the whole thing is a fix due to pulling out a synth. that's not a complex waveform.
redo the demo and prove us wrong. i can't remember the track but there's a RAMP joint where Roy Ayers ruins it trying to sneak a fart. remove that and i'm a believer.
Yeah sure, isolating synths is just so easy, even when they have like 20 overtones you need to keep, that drums get in the way and are curvy as F*ck. Next example is vocal removal anyways, your fart removal will have to wait, although farts are simple, unless it's one of those that last 20 seconds.
So can you isolate acapella's with this? Or is a voice too complex a sound? Can this give me the elusive "Tweeter and the Monkey Man" acapella?
Sure you can do that. Vocal removing tutorial is coming.
somebody please explain the reverse phasing thing that people are creating accapellas with. where the accas sound phased.
most of them are done using cd singles, or digital versions of a song. You need the instrumental and the regular version. You need the digital (cd single) and not a vinyl version because you need them to be exact and locked up for the duration of the song. Vinyl has minor shifts and doesn't stay locked.
Anyways, you take the regular version and the instrumental. Put them into pro tools or some other program and you invert one of the files. That basically flips everything, not reversed, but more in terms of where the sounds are in the mix. Basically the idea is that when you invert one of them it ends up phasing out all of the identical parts. So, the vocals remain because they are the only thing that isn't the same on both versions.
somebody please explain the reverse phasing thing that people are creating accapellas with. where the accas sound phased.
most of them are done using cd singles, or digital versions of a song. You need the instrumental and the regular version. You need the digital (cd single) and not a vinyl version because you need them to be exact and locked up for the duration of the song. Vinyl has minor shifts and doesn't stay locked.
Anyways, you take the regular version and the instrumental. Put them into pro tools or some other program and you invert one of the files. That basically flips everything, not reversed, but more in terms of where the sounds are in the mix. Basically the idea is that when you invert one of them it ends up phasing out all of the identical parts. So, the vocals remain because they are the only thing that isn't the same on both versions.
Word - I've heard of this system before too but have never seen it actually done. Sounds pretty slick. Is the acapella "perfect" or are there still artifacts from the original instrumental still attached?
I think there are artifacts. that's why they sound phased. I think it generally works out ok if you put something under it, but as an acapella alone you'll hear the phasing.
I'm not an expert at all this fancy mad scientist stuff. I tried doing it with a cd single I had of Cut Chemist's song with Edan and Mr Lif. I had it all lined up and inverted it, but for some reason it phased out parts of it and then other parts came back in, full instrumental. I think he might have done slightly different mixes of the songs. I also tried it with a cd single for a Raekwon song off of Cuban Linx. I forget the song now and can't find the cd, but it was one that didn't have an acapella. It must have been recoded on tape because the two files, although taken from a cd single, were way off. The instrumental was like 30 seconds shorter than the vocal version. It was a different speed too. Subtle, but enough to make the files not lock up at all.
Problem is that a lot of instrumental versions aren't just the vocal mix sans vocals.. I've got plenty records where they add bars to the intro, cut intro samples, etc. If you can't rely on them being identical it won't work.
Yeah sure, isolating synths is just so easy, even when they have like 20 overtones you need to keep, that drums get in the way and are curvy as F*ck.
while I think what you do is pretty cool, you gotta admit that the synth you removed is a pretty simple sound. I mean the enveloppe for all its overtones is basically flat, that's why it looks easier to remove with a "visual" program like yours than a plucked instrument or a bell type sound.
Yeah sure, isolating synths is just so easy, even when they have like 20 overtones you need to keep, that drums get in the way and are curvy as F*ck.
while I think what you do is pretty cool, you gotta admit that the synth you removed is a pretty simple sound. I mean the enveloppe for all its overtones is basically flat, that's why it looks easier to remove with a "visual" program like yours than a plucked instrument or a bell type sound.
Arguably, that type of sound has upsides and downsides. But it's definitely not easy, nor is it as easier than other types of sounds as people like to assume.
As for that phasing thing, when I was a kid I would plug the - wire of a speaker into the other speaker's + slot. The - of both speakers being ground together it'd get me the difference between both channels on that one speaker. Am I the only one who did it? Surely is simpler than using a computer program ;-). By the way in my experience it removes the vocals more than anything. Except a few stereoified vocals in choruses. And usally the whole thing sounds like it was recorded under a bridge.
I wouldn't argue that isolating any sound would be easier I just haven't seen a proof of concept in full hey A_SN, could you do a screen cast with the next example? from start to finish, narrate that shit too, thanks
Grandfather
2,303 Posts
Comments
im a bit apprehensive, anyone here used it?
I'm sayin'. that is incredibly f*cking accurate. If they're somehow manually isolating the waveform is the fact that it's a synth line make it easier to do that? That synth wave is pretty much a single line with no harmonics or anything. I can't see doing the same with a horn section or a guitar.
For more information on how he does it read - http://photosounder.com/blog/2009/03/instrument-isolation-funky-worm.html It's pretty amazing, he does it using photoshop!!
I was thinking the same thing, that's a very flat/pure tone. I can't see something like this being possible on instruments with dynamic sound/tone, it's just science.
if you listen close , you can still hear the section being solo'd faintly in the track.
alot of hip hop acca's as well as a few of the michael jackson accas were created this way. there are multis for rock with you and dont stop til you get enough.....the rest im unaware of.
http://en.wikipedia.org/wiki/Fast_Fourier_transform
I will say that his color choices for the interface are really pretty. Much better than Metasynth, which is more powerful but has always been deliberately foul-looking.
Haha, I'm not sure if it was intended as a compliment, but I'll take it as one. I wasn't very impressed with my own drawn drums to be honest...
Working on it right now, although to be honest it's more like I'm struggling with setting up my dev environment.
Hahahaha, this thread is epic. Yeah, you're right, I'm the only guy on Earth who has the tape of the original recording of Funky Worm with only the synth on it ;-) how lucky is that? Someone call Biz Markie, I'm about to become rich!
Yeah, it does the very exact same thing, mmmh except maybe for the fact that it doesn't process recorded sounds, let alone using images.
No harmonics? Oh if only it was true. Look at the video, you see all these parallel lines? They're harmonics. I had to keep up to 20 of them. If it had no harmonics it'd sound like a sine wave. The fact that it was a synth helped knowing where to look for things, but it doesn't help that much, the fact that it had a shitload of harmonics that were getting lost into the noise of drums didn't help a bit, neither did the fact that it's really full of curves.
Horns and guitar have different challenges to them, horns being the fact that they too have lots of harmonics that you need to keep, and on top of that their lines aren't smooth. Guitars have extra noise that you'd better make sure to keep if you don't want the result to sound like a cheap synth, but I don't think they would be that hard, that is if we're talking about plain old acoustic guitars.
But I get that a lot, that's why next thing I'll do will be an easy tutorial pertaining to removing vocals.
Hahaha, an enormous douchebag, right on. Too bad you're pretty much entirely wrong about it being just an FFT/IFFT (although you really mean a STFT/ISTFT). If it was based on STFT, then the time resolution would be fixed, which means that in the lower end of images things would get blurry vertically, not horizontally. It actually uses something better adapted which is basically just frequency domain windowing with modulation by masking envelopes, which allows to do anything you want concerning frequency/time resolution depending on the frequency, but also on the frequency scale. That's how here you can get natively a logarithmic frequency scale (or pitch scale if you prefer), whereas STFTs are strictly linear, which is pretty impractical.
But go ahead and reproduce the results with any other program, since it's such basic processing, and definitely nothing innovative ;-).
If this is legit, its pretty dam
Good job, welcome to the strut.
Whatever. He just
Seriously... this shit is crazy... I am very interested once there is a MAC version available!
An instrumental of this song was released (bootleg) so it's not exactly crazy to assume that the master tapes are available to certain privileged folk. As to whether or not the song was multi-tracked is beyond my knowledge though. Let us know when you can isolate and guitar or bass riff as clean as this, that would be really impressive.
redo the demo and prove us wrong. i can't remember the track but there's a RAMP joint where Roy Ayers ruins it trying to sneak a fart. remove that and i'm a believer.
PM me your e-mail address, I'll add you to the Mac mailing list
Looks like I'm going to have to do it since so many people think that isolating synths is just the easiest thing on Earth and that anything made by plucking actual strings would be impossible... Do you have any particular sample in mind? i.e. anything that people would have loved to sample?
Yeah sure, isolating synths is just so easy, even when they have like 20 overtones you need to keep, that drums get in the way and are curvy as F*ck. Next example is vocal removal anyways, your fart removal will have to wait, although farts are simple, unless it's one of those that last 20 seconds.
Sure you can do that. Vocal removing tutorial is coming.
most of them are done using cd singles, or digital versions of a song. You need the instrumental and the regular version. You need the digital (cd single) and not a vinyl version because you need them to be exact and locked up for the duration of the song. Vinyl has minor shifts and doesn't stay locked.
Anyways, you take the regular version and the instrumental. Put them into pro tools or some other program and you invert one of the files. That basically flips everything, not reversed, but more in terms of where the sounds are in the mix. Basically the idea is that when you invert one of them it ends up phasing out all of the identical parts. So, the vocals remain because they are the only thing that isn't the same on both versions.
Word - I've heard of this system before too but have never seen it actually done. Sounds pretty slick. Is the acapella "perfect" or are there still artifacts from the original instrumental still attached?
I'm not an expert at all this fancy mad scientist stuff. I tried doing it with a cd single I had of Cut Chemist's song with Edan and Mr Lif. I had it all lined up and inverted it, but for some reason it phased out parts of it and then other parts came back in, full instrumental. I think he might have done slightly different mixes of the songs. I also tried it with a cd single for a Raekwon song off of Cuban Linx. I forget the song now and can't find the cd, but it was one that didn't have an acapella. It must have been recoded on tape because the two files, although taken from a cd single, were way off. The instrumental was like 30 seconds shorter than the vocal version. It was a different speed too. Subtle, but enough to make the files not lock up at all.
while I think what you do is pretty cool, you gotta admit that the synth you removed is a pretty simple sound. I mean the enveloppe for all its overtones is basically flat, that's why it looks easier to remove with a "visual" program like yours than a plucked instrument or a bell type sound.
As for that phasing thing, when I was a kid I would plug the - wire of a speaker into the other speaker's + slot. The - of both speakers being ground together it'd get me the difference between both channels on that one speaker. Am I the only one who did it? Surely is simpler than using a computer program ;-). By the way in my experience it removes the vocals more than anything. Except a few stereoified vocals in choruses. And usally the whole thing sounds like it was recorded under a bridge.
I just haven't seen a proof of concept in full
hey A_SN, could you do a screen cast with the next example?
from start to finish, narrate that shit too, thanks