I suspect I’m not the only one here who likes to add vocal samples, either sung or spoken, to my tracks.
I wish I had more confidence in my voice but alas am still working on that. In lieu of having a singer to collab with, I turn to vocal samples.
The problem is that the vast majority of said samples are just…fucking awful. Jump on Splice, for example, and there’s no shortage of vocal packs. The problem is that the vast majority are either, a.) featuring insanely cheesy lyrics about love and relationships, or b.) ridiculous low-pitched “DROP THE BASS” or “HOUSE…HOUSE MUSIC” samples. You know what I’m talking about. The same is true of pretty much all the sample/loop production companies.
I’ve managed to find some good stuff over the years…chanting in different languages, sampling anime movies, random other sources from freesound or reddit or whatever. But man - it’s always a struggle.
Where do you all like to turn to for this? I’m especially looking for sung vocals with non-cheesy or even dark lyrics. Maybe even in a punky style. Something different. Ugh it’s all so cookie-cutter.
Don‘t want to be that guy, but to mention the obvious: find some people who want to sing. Don‘t expect professionals, but be kind and let them experiment. Enjoying an evening like this is more fun than browsing the internet.
Not what you wanted to read, but my perspective over the years.
It will depend on the kinds of vocal chops you like and are able to use in your music, but my go-to is the Prelinger Archive.
Lot’s of great content that’s all public domain. A lot of it can be isolated easily, some can be found clean without music/other audio in the background and there’s always Stemroller if you want to separate it out with a bit more ease.
Many of the records me and @aarb420 pick for our sampling challenges have vocals (search here for ‘sampling extravaganza’ to get to them), and I’ve (only very recently) started to include Stemroller exports in the threads to make it easier to grab the vocals. As it happens the latest pick doesn’t have vocals but that’s the way the cookie crumbles. We try to pick stuff that’s a bit less well trodden so might provide some interesting inspiration.
Otherwise I’m going to hang around because I’m always interested in this too.
…apart from the fact that everybody can “sing” if u “only” take the “effort” to find UR voice…
it’s sooooooo much satisfaction, to involve ur very own thoughts and voice into ur personal sonic creative process, u really better keep on trying…
nothing beats that individual little xtra, if any vocal snippets got their original source from URSELF…
and if singing is too much to ask from urself for now, any spoken, whispered words will do totally fine…we got so much tools these days to take any piece of breath and turn it into something special…stop the hunt for cheesy substitutes, many others already used and abused in one way or the other before u, become ur own original and just record ur very own words…and if it’s only uuhs and aaahs at first…just do it…
I’m a big late 80s /90s movie and video game fan, there’s a shit load of useable dialogue in them, I usually don’t have a specific idea or theme in mind when I start hunting but I’ve found the quotes feature on IMDb pretty useful, Ive found some great nuggets that way especially in sci-fi movies…
It may be obvious, but search for “acapella” on YT. Tons of stuff comes up. Some of it is just amateurs singing to a camera, but some are actual studio recordings and outakes.
If you don’t want to do it yourself I support the idea of talking someone into helping you out. Years ago I had a friend with a great voice and I gave her some, uh…stuff to record whatever she wanted for me to chop up for samples. I handed her a cassette recorder and the next day she gave me a 90 minute tape of pure gold. That was almost 15 years ago and I still pull samples from it now and then.
Yeah I feel that. I’ve been going through this exact thing for a bit. I should say that for me I’m not a fan of just grabbing stuff randomly, as much as I know people can and do this without any comeback. I prefer to keep it simple & clean on that score.
So sticking with the commercial end, I tried Splice & Arcade, but found the sounds a bit overprocessed. Loopcloud is a tad better because they have a lot of darker sounds on there, so finding decent sounds is relatively simple. Then you have Tracklib. This is my personal favourite - however, there is a fee if you want to release anything. Since this is purely a hobby for me, that kinda ruins the Tracklib thing for me.
Actually, for the sort of music I make, I’m currently having the most joy with vocal Kontakt libraries, but in honesty just having the right samples would be just as good - I just think this a route that ticks a few boxes for me.
How much vocal processing have you tried? I would say experiment with that and process your own voice… find something that works as kind of a signature processing, then maybe dial it back over time if you you get better at singing. You can do a lot these days with voice processing.
My significant other has a great voice, so I have them record things from time to time. Once we generated nonsense words on the internet another time it was vowel sounds. Etc.
The most fun was when I took screenshots of YouTube’s terrible auto-generated subtitles of last year’s Eurovision. Then my partner read them into my recorder, and I chopped them up into short phrases (I also used Shortcuts to automatically name them using voice transcription).
So, I guess my answer is, hook up with someone who has a great voice and accent, and is into experimenting.
Probably not to all tastes but librivox.org has a lot of material in several languages, all fully public-domain and usually decent recording quality for spoken-word samples from public-domain books.
I feel this, can’t stand all those cheesey vocal samples, even worse are those sci-fi vocal snips from some film in the 50’s or the YouTube equivalent of that (that got old in the 90’s). You couldn’t pay me to spend 5 mins with a real singer either. I am going to go all in on AI for vocals, it’s getting really crazy what can be done in that space now…
For the spoken stuff, the secret weapon is film archives and lecture recordings. The nice thing about lectures is, if you have a desk job, you can stick them on as background noise and learn stuff (ok, not well) while you work, and then just grab samples whenever you hear something you’d like to… recontextualize.
If you do this with shows, pick something that doesn’t have music playing in every scene. Just something to look out for before you commit to it. You might get a few eps in before you realize you want to sample something, only to discover it’s got a bunch of music in it. Newer shows are usually bad for that, and even older ones can be. Same goes for movies. Actually, they’re pretty awful for it. Of course there may be better ways to clean it up than what I’m used to. I haven’t tried any machine learning tools.
I go for physic or astronomy lessons or alien movies from the 50ties… i discoerd edward tolle latley in a track lately, pick your subject and find guys talking about it.
This is awesome! Do you two have any particular approach you use when it comes to searching a giant endless library like Prelinger? Start with a particular vibe, time period, or genre? Or just random flip through the vast “crate” until you land on something that grabs you?
