hmmm... I wonder if they could pick it up if you made it a half step up and a meter or two faster ... and after mixing... bring it back down? That is... if the process of the transpose would alter their info. Or write it as such and just change the instrumentation to kazoo, garbage cans, actual toy instruments, etc... Yeah, I to am afraid of such... but moreover... inadvertently ripping off someone else's score...
EDIT: deleted, I was obviously too jerky, due to a lack of sleep, to get the joke. Last edited: Dec 12, 2018
You are right, but if the track will be released and the developers hear my work then I can get sued.
Yes, you record the analog signal (= line out or AUX) from your mix or submix (stem) on to a free track (line in), in sync with and parallel to the digital source track(s). You often do this anyway, for instance to process the mix or certain channels "outboard", analog and outside the DAW's domain. Think of how you would a hardware tube compressor, or a reverb, or whatever, on a single track and feed its output back to the DAW. Line out, Line in. [EDIT:] The above pertains to stripping metadata, not audio-watermarking! Refer to bottom of this page for better explanation of the difference and their application. Last edited: Dec 10, 2018
Are you sure about that? From Wikipedia "Watermarking is the process of embedding information into a signal (e.g. audio, video or pictures) in a way that is difficult to remove. If the signal is copied, then the information is also carried in the copy. Watermarking has become increasingly important to enable copyright protection and ownership verification." From Fraunhofer "Our audio watermarking solution is a robust technique based on the Fourier transform. The watermark is inaudible and withstands mp3 encoding, time stretching & pitch shifting, changing the sound volume, dynamics compression, trimming and DA/AD conversion including microphone recording." From audiowatermarking "...surviving low bitrate transcoding with lossy coders, mixing, effect processing and acoustic coupling (transducing via air)."
Can you guys explain me what do you mean by this? I mean using watermarks on samples can ruin the sound, it would sound realistic that way. Do sample library developers really do something like this?
First of all, if all you're doing is pitching a demo and hoping for your stuff to be featured in a film, I wouldn't get in expecting to get a pay in 6 months. Second, everything from EastWest is $20/month. If you're serious about doing this, http://www.soundsonline.com/december-super-sale
Watermarks on sample libraries don't mean actual watermarks in the samples. It means buyer info embedded in the NKI files, so they can trace leaked libraries online to who originally bought it.
The most important is whether they would realise it or not. If a developer won't be sure whether I used their samples or not there won't be any problem. I just used only some legato and spiccato articulations. Also modified those with reverb, eq, multiband compressor. I can't hear too much difference between the libraries, but maybe the companies would hear it. Hopefully not.
Hi @No Avenger , yes I'm aware of Fraunhofer protection as very professional and highly reliable... on complete, finished works. I know how it works but that technique of audio-watermarking is hardly applicable to the individual samples making up any library. The technique's effectiveness is, partly, audio spectrum dependent. Also, the longer the protected work's time base, the better the protection (or rather, the more reliable detection). Mixing together two or more individually watermarked audio sources or sample files will (in most cases) render either of the constituting sources' watermarking non-responsive, thus useless. The thing to remember is, infringement detection is done by comparing non-watermarked audio with the (intercepted) watermarked stream, much as by phase-inversion & cancelling, although the process is really graphical, not auditive, in nature. But principle's the same. The resultant noise, against a precise time-base, and regardless of, for example, pitch change, can be analysed for the known encoded "fingerprint". The phase-inversion will no longer work if a second program, meaning another sound, is mixed in with the watermarked stream. Or rather, it will work but then the mixed-in second audiostream will mask off the watermark, resisting detection. But again, this applies to finished products/demos, not single note samples. Sample libraries are generally protected in metadata in the files, rather than by audio-watermarking proper. As long as a label, or any distributor of your recorded work, does not require express statement that all used samples/sample libraries are licensed to the owner of the copyright applicable to that recorded work, I doubt there is anything they can fruitfully undertake, legally, against the recording artist, upon -discovered- infringement. Well, unless they happen to own the rights to the original library themselves!
don't worry just do it... Download metropoolis ark 1 2 and 3, get la scoring strings, cinematic strings 2, get some Spitfire libraries, everything heavyocity and Output, Evolution World percussion series 2.0 and whatever else you think you may need. Figure out the ones you REALLY like and will be using, and buy a few as the checks start rolling in. I've bought so many libraries and don't even use half of half of them lol. Plus No one will know unless you tell them. Don't tell on yourself lol. And don't let lack of money stop you from pursuing your dreams. Find a way and then capitalize on it and set things right as time goes by. There's too much unlived potential because of lack of resources. So i say do it... these are moments that could redefine your career and happiness. Good luck Bro!!
I think (if mem serves) he was more concerned w/ producers sending in near identical tracks (in one case the actual music used on another one of their trailers (!) ) and ppl using "loops" etcs. I don't think anyone would complain about original music being made from (ie) OTB / Spitfire libs. Or would notice. I think it's more that everyone & their pet dog was using Epic Trailer Guitars #54 and they were getting a bit fed up w/ it? :D
You are right. But this was his second concern.. I'm not moralizing or anything here. The thing that bothers me is that the final product is (possibly!) compromised for eternity ;) And that is prob a nightmare for a representative of WB for US.. (imagine arranging an financing all the video editing, advertising, artwork.. and then have to take it all down because of some "sample developer" bothering you.) These questions are like: will I get hit by a bus tomorrow, and expecting the no way, be free, run through the streets answer. The best I can give is - probably, I hope.. not. It is about how much risk are you wiling to take. Sure OTOH without taking any risk you won't get very far.. Here you have the orig post, leave out the middle ethics blabla.. because as I see it, somebody started to give him headaches and that's all. In addition to what you already pointed out.. A suggestion for the OP: make your composition with.. whatever, and then find someone /w licensed libs that will pass your file through (as a service - you still own the copyright in any case). Totally legal. Or take, as already mentioned, a 9$/mo cloud plan for this year..
Thanks for useful pieces of information. You are right that the label will accept my work whether the samles are licensed or not, but as Taxidriver mentioned the library developers can take action if they can detect their samples in my track. I also read the post of Agus and I'm not sure what the develovers could have detected. (SFX / real instruments) The sound effects are usually so unique that they can be easilly detected, but if there isn't any watermark in real instruments then everything should be fine. I'd send the instruments in context. Not the raw samples. Luckily I have to send only the stems this way: sustain brass, spiccato strings, legato strings. In this case the frequency range would be a lot different than in the watermarked raw samples. I will read more about this Fraunhofer protection, but it sounds impossible for me to detect the samples if there are more instruments playing, since the frequencies are mixed up and the samples are also processed by me.
I don't know how much this gonna work out LAYER TWO/THREE DIFFERENT LIBRARY SAMPLES TOGETHER AND MAKE A FRESH TONE ! All the best !
