Any advise how to tune field recording samples, which plugins ?

[zen001]

Hi soundlovers,
Let me explain my question :
I would like to incorporate my own sound design samples (source material is mostly field recordings, found sounds, ... sounds whithout a fixed root note, when analysing one sample, the notes are shift continuously between different keys).
So, its very difficult to place them in a musical context without being out of tune constantly due to the shifts.
Any help would be great !

[thecontrolcentre]

What sort of field recordings are we talking about?
Maybe Noises by Audiothing could be useful ... you can load your own audio into it.

[zen001]

Hi, Thanks,
I know about Noises, but would it solve the tuning question ?
I tried (Mia Lab) TS and other tuners, also tried to tune to 440Hz and derivatives.
First method doesn't work, samples are "floating between different keys" too much, would make anyone dizzy
Second method works but I have to cut too many frequancies, so there's not much left from my original sample.
Hope my explanation is clear

[Michael L]

Melodyne Editor is the only way to tune every part of the sample to the same key and keep the transitions smooth. Blobs are better than slices.
Other products lack the detailed control needed for complex field recordings.
Its expensive except when on sale, but you will enjoy it so much more than anything else.
Demo it and see!

[i need Help]

Skip to 3:25 and watch for about 5 mins to see if it’s of interest. Not sure if this is along the lines of what you’ve already tried based on your reply above.

[zen001]

Yes , that's it, this is what I did before. As you can hear, there's not much left from the original sample, which was the reason why I posted my question.
However, I've no clue how this could be achieved in Melodyne
Thanks for answering.

[HAL76]

there are two possible ways to go:

a) extract a tonal stable part and stretch it, loop it, resample it through reverb or use a granular sampler
b) straighten the pitch (possible? result?)

I don´t really know what you want to do, but people did such experimental "instruments" with looped samples decades ago already. In Ambient is has also become usual to resample a reverb tail and loop it. Others use time stretching to make the sounds usable, but the results will differ strongly from the source. And last but not least there are various options you have with granular synthesis today.

And tzo be honest I wouldn´t really think about A.

[zen001]

Hi all,
After experimenting with importing my samples in vst's, choosing a relatively stable tonal part of it (thanks HAL 76 for the growing insight) and then letting the synth find the key of the sample my problem is solved for the biggest part.
It all however came down to the following insight : "every sound has a certain tonal component" or at least "some kind of dominant frequency/frequencies" that can be translated to a certain key.
I hope I put this right ?
Even if a recording doen't "seem" to have any tonality "at first hearing" (f.e. bees buzzing while the leaves are moving and crickets sing in the background while the neigbour mows his lawn ...), there's a part of that recording that has a certain stable part in it (in my case, the longer the part, the better).
This clip has a certain key, and thus can be transposed and mixed in a musical context without sounding "a-tonal".
Thanks to all of you for your contribution and your efforts.
All the best !

[HAL76]

Hi all,
After experimenting with importing my samples in vst's, choosing a relatively stable tonal part of it (thanks HAL 76 for the growing insight) and then letting the synth find the key of the sample my problem is solved for the biggest part.
It all however came down to the following insight : "every sound has a certain tonal component" or at least "some kind of dominant frequency/frequencies" that can be translated to a certain key.
I hope I put this right ?
Even if a recording doen't "seem" to have any tonality "at first hearing" (f.e. bees buzzing while the leaves are moving and crickets sing in the background while the neigbour mows his lawn ...), there's a part of that recording that has a certain stable part in it (in my case, the longer the part, the better).
This clip has a certain key, and thus can be transposed and mixed in a musical context without sounding "a-tonal".
Thanks to all of you for your contribution and your efforts.
All the best !

I don´t know if that is the physically correct way to express it, but "sounds" consist of vibrations that have - more or less - constant frequencys and often one or multiple dominant frequencys (noise is the extreme opposite of "mixed frequencys"). The composition of the frequencys makes sounds seem more or less "harmonic" and "familar" while the harmonic content of many results will be more dishamonic, composed of odd harmonics with weird relations (like 1 - 2.376 - 5.921 - 7.393 ...).

Using samples like field recordings as oscillators is more common than you might think in experimental music (e. g. Aphex Twin Rhubarb). But not every sample matches the requirements and it´s pretty difficult to generate matching combinations as the relations of the harmonics and also noise parts are usually pretty special. You can do a lot with EQing, especially for single samples (set the EQ at high Q to the desired "fundamental" frequency = key) - and good FX chains (Reverb, Chorus, ... filters!) make it easier to handle the sounds and combine them..

You can use e. g. MTuner to determine the "root key".

Me personally - I mostly use grain synthesis (Halion) since the results are of better quality, more flexible and transposable.

Can we heaar something already?

[cron]

It all however came down to the following insight : "every sound has a certain tonal component" or at least "some kind of dominant frequency/frequencies" that can be translated to a certain key.
I hope I put this right ?
Even if a recording doen't "seem" to have any tonality "at first hearing" (f.e. bees buzzing while the leaves are moving and crickets sing in the background while the neigbour mows his lawn ...), there's a part of that recording that has a certain stable part in it (in my case, the longer the part, the better).
This clip has a certain key, and thus can be transposed and mixed in a musical context without sounding "a-tonal".
Thanks to all of you for your contribution and your efforts.
All the best !

I think there are a lot of interesting finds in this area, but it might need a bit more compromise in terms of fidelity to your original sounds.

The opposite of "tonal" in the context of acoustics isn't "atonal": it's "noisy". Most sounds have some sort of tonal component, but that doesn't imply some degree of compatibility with the musical meaning of tonal. What really matters is whether those tonal components are harmonically related. Some bells, for instance, are entirely tonal (in the acoustics sense) and completely stable, but inharmonic and having ambiguous or undiscernible pitch centre that makes them difficult to use as tonal musical elements.

I'd perhaps consider additive strategies if you'd like to hang on to the character of the sources a little more. It might be more natural to add small touches than to severely process what's there - a few quiet sine waves harmonically strengthening an existing tonal component, a subtle blend with an acoustic instrument, or maybe even something like 10% wet through a vocoder or tuned delay - you might be able to get something playable while leaving the source as an almost entirely dry layer on top.

I'd stay open to losing some of the source's character along the way though. Choose your compromises and all. One of my favourite records of the last few years has some amazing field recording / spectral vocoder work and I never really find myself noticing how it's technically put together or feeling like the field recordings have lost something - if anything it feels more like they've been magnified despite them no longer being themselves as such. The bit between around 4:50 and 8:50 is probably the best example on the album.