really really in tune samples -- how about brass?

[gugliel]

My forum friends know me and my tuning bee-in-the-bonnet by now ...

Recently bought the Dan Dean Solo Woodwinds, and they are beautifully tuned. From time to time, the hand of autotune might show itself, but it is such a great pleasure not to have to compensate for bad tuning in supposedly commercial quality samples.

The EWQLSO Gold samples are annoyingly out of tune in places, GOS is out of tune in places, AO is badly out of tune in places, VSL solo strings are a little out of tune, the VSL samples packaged with Gigastudio don't sound that great to me. So my question:

Howzabout brass? Without regard for how good they sound, are the SAM brass packages EXACTLY in tune? Anybody checked with a tuner? Any other brass that are really tuned well?

[JonFairhurst]

Part of the problem with brass is that it depends on the performance. A plucked/struck string will be slightly sharp at first, and will then settle in. A horn player can be all over the map. So the first requirement is awesome performances. My first lib was Bigga Orchestral Brass, and the performances of some of the instruments was so-so. There would be no way to tune them over the length of the sample. On the other hand, a brass sample that is super-humanly on the note will tend to sound without soul.

I guess what I'm trying to say is that the top priority is great, human performances. Tuning is important, but secondary, and can be adjusted in the sampler, if necessary. With a mediocre performance, there is no fix for tuning.

Maybe it's just that I played trumpet and baritone as a kid, and can hardly play a jiggly octave on a trumpet today. But I think that playing brass with solid tone and intonation is more difficult than playing any other family of instruments, with reeds coming in a close second. The whistle/flute, strings and percussion families are much less challenging, IMHO - though top performances and years of experience are still required for a top lib.

-JF

[kitekrazy]

By nature brass instruments do not play perfectly in tune except for trombone. All valve combinations except for 2,3 play sharp.
Brass instruments do not follow equal temperment.

[Mike Auty]

By nature brass instruments do not play perfectly in tune except for trombone. All valve combinations except for 2,3 play sharp.
Brass instruments do not follow equal temperment.

I don't even play a brass instrument but I'm sure you can tune notes with your embouchure...

[gugliel]

All that you say above is very true, and true for that matter for strings too, and even woodwinds to an extent -- but SAMPLES are not performances, we need to add the expression to a string of samples to make the performances. And for this, one must know what the baseline is -- and a recording of a horn all over the map in pitch, while 'realistic', is not very easy to use. And when a horn player plays out of tune, a real one, good ones anyway, the sound never loses the connection to the music, and either gets tuned or gets made 'compatible' despite being out of tune. And I play french horn well enough to know that it is highly tunable with embouchure, trumpet maybe less so but still tunable.

So I'd still like to find the best brass intonation possible in a set of samples.

[Eric G]

What do you want the brass samples to be in tune with? As I'm sure you know, a rich, glorious brass choir chord gets its sound because the members of the chord follow the overtone series based on the root of the chord. If you change the root, you change the overtone series too. It's the old equal-temperament vs. just intonation dilemma.

Or, if I'm playing a concert "D" as the 3rd of a Bb major chord, I'm lipping that note down quite a lot to get it really in tune. If the root of the chord then changes to a G while I continue to hold the D, I'll probably have to lip that pitch up just slightly above ET to be in tune with the new root, all other things being equal. Again, as I'm sure you know, real players make these adjustments constantly. Even good high school players do this to some degree of success.

How will a sample library accommodate that? If I need a big "hero" chord, I'll put the members of the chord on different MIDI channels and adjust pitch, but I can't imagine how a sample library could be in-tune out of the box.

[gugliel]

Eric, you must be a brass player, because tuning down that D is exactly what is needed IN A Bb TRIAD. When the D is in a D major triad, however, it's the F# you'd be tuning. I want the samples to be pure equal temperment, so that I can quickly know how far to change the pitch depending on whether I'm in D or Bb.

Lee, thanks for the qualification -- exactly the kind of thing I was looking for, and it's appreciated.

[Eric G]

I see, Gugiel. I misunderstood what you meant by "perfectly in tune", thinking that you wanted each note to work in just intonation in all keys! Obviously that wouldn't work, but having very accurate equal temperament would be a consistent compromise.

--Eric

[josejherring]

I'm not so concerned with really really intune but really really realistic. I've never used auto-tune on samples. Clever.

When real players play they will adjust the tunning as they hear it. There's a little time lag from when the instrument strikes a note in an ensemble and when that ensemble adjust it's intonation.

My question is this: Does auto-tune let you adjust the time it takes for it to correct a pitch or is in alway instantaneous?

Cheers,

Jose

[George Klos]

I've used autotune on recorded tracks from samples. Autotune lets you adjust both the 'tracking' sensitivity and the 'speed' that it corrects. I tend to loosen both of them from the stock setting that the plugin activates with. For me, the looser settings still sweeten the sound, yet will have fewer tuning artifacts. It just sounds more 'natural', if that can be said of samples.... :)
I use looser settings on vocal corrections as well.

George Klos
Klos Studios

[wes37]

As the majority of us are using equal temperment, what exactly does "in tune" mean? Many musicians with phenomenal ears intentionally tune their instruments ever so slightly sharp for a more "biting" tone (Eric Johnson and Steve Vai, for example).

A real orchestra is always "out of tune". Brass attacks are slightly sharp, strings tend to meander between equal and pure temperment (depending upon the ears of the players), and woodwinds are all over the place. The end result is a big, rich sound, for the most part.

As an aside, I think it's kinda funny when an engineer will obsess over a performance being in tune and then slap some chorus all over it (and yes, I do know the difference between being out of tune and in tune with chorus, but you get the point).

Fodder for the discussion.

[gugliel]

wes, I just mean 'really well tuned to the custom of today' which is equal temperament. Speaking just of my own work, I know there are lots of failings when I begin to tune things. But my current project is up to 150 tracks, 12 minutes, and I think around 100,000 [edit -- currently only 30000 if my counting CAL worked correctly] midi events -- so it gets a little difficult to catch everything. If the samples themselves were perfectly tuned to start with, that would be one thing NOT to have to take care of. As it is, with most of the libraries I use, I've got to listen constantly not just for articulations that jar, but for endings that scratch, notes that go out of tune, etc, etc, and then do it all again for eq and reverb and dynamics levels, and everything else we do.

And, yes, realistic is desirable for me too, not only pure tuning -- but the sound of a 'realistic' french horn includes meaningful tuning, whether in or out of tune, and NOT random tuning just because on the recording session there was some room or instrumental resonance to which to falsely tune!

[wes37]

My points were more philosophical than a critique of your arguement....I know what you mean. Sample libraries should come perfectly edited.

Perhaps users should post the instruments or samples that need re-editing and pressure the developers to reslease updates that correct the problem(s).