Simulating Dynamics

[johnmarkpainter]

In the midst of rethinking my sequencing techniques....tips and tricks anyone?


When we are emulating Dynamics with Samples, we aren't trying to duplicate the Live instrument in a room with us...we are trying to emulate the RECORDING of a live instrument in a room.

Recordings are all Compressed/Limited to limit the Dynamic range to something that can be reproduced (Radio/TV Broadcasts and Pop records are HEAVILY compressed as are most Films)

Varying the Timbre and Vibrato/Intensity is more crucial than varying the volume. Often, the Dead giveaway of a Sample library is in an awkward Fade that is attempting to sound like a Dynamic performance.

In GPO or the various Solo libraries, I THINK would like to be able to tweak the minimum Volume for an Instrument while keeping the Timbre shifts (and maybe making them more drastic?)

I have experimented with running Compressors on each instance of GPO and it is closer but it doesn't respond in quite the way that I mean.

I have found that in recording situations (with Live Players) I have to write the Strings one dynamic level Louder than what I really need to make sure they are "speaking".

I write the Brass one Dynamic Level SOFTER than what I really need because the Brass are capable of making more 'noise' and are generally not capable of playing as quietly. Also, Trumpet players get that excited rush of Testosterone when they see accents and ff (I am a Trumpet player so I know...)

[wst3ae]

When we are emulating Dynamics with Samples, we aren't trying to duplicate the Live instrument in a room with us...we are trying to emulate the RECORDING of a live instrument in a room.

That might be the single most important tip that any body can receive, except I'd extend it to be more general than just dynamics<G>!

And actually, there are a couple of schools of thought on this, though they generally apply to recording live players.

School 1 attempts to capture the performance in it's entirety, room sound, everything. It's the capture or document a performance approach, and there is nothing wrong with it at all.

However, when you are working with samples you are working with sounds that have already traversed the entire recording chain at least once... so the strict document a performance approach might be a wee bit tricky!

School 2 says that there are no rules, and you can create some wonderful soundscape this way. Put one reverb on the violin and a completely different reverb on the cello, for example. It isn't something that could happen in real time in real life, but it can be very cool sounding.

Recordings are all Compressed/Limited to limit the Dynamic range to something that can be reproduced (Radio/TV Broadcasts and Pop records are HEAVILY compressed as are most Films)

While this is certainly true I'm going to stick my head back in the sand and continue to try to capture a dynamic range that appeals to me! Funny, I can have a much wider dynamic range with samples than I can with live players.

Varying the Timbre and Vibrato/Intensity is more crucial than varying the volume. Often, the Dead giveaway of a Sample library is in an awkward Fade that is attempting to sound like a Dynamic performance.

Yup... all aspects of each instrument ought to be varied, but that becomes quite a large undertaking! Eventualy we'll have tools that help us do this, think Midi Maestro or iMIDI!

In GPO or the various Solo libraries, I THINK would like to be able to tweak the minimum Volume for an Instrument while keeping the Timbre shifts (and maybe making them more drastic?)

Interesting... very interesting! That makes so much sense I sorta gotta slap my forehead!

I have experimented with running Compressors on each instance of GPO and it is closer but it doesn't respond in quite the way that I mean.

Compressors, even the C1 from Waves, are imperfect, and emulations of old hardware are imperfect on two levels<G>!

I think I'd prefer to use volume as my control over dynamics, but dang, it gets really involved as soon as you add the second instrument!

I have found that in recording situations (with Live Players) I have to write the Strings one dynamic level Louder than what I really need to make sure they are "speaking".

I've found this to be completely variable... some players play louder and compress the dynamics more than I intended, some do the opposite.

The only exception is the electric guitar player... he always turns down as soon as I put music on the stand!!!

(Yes, I am a guitar player)

I write the Brass one Dynamic Level SOFTER than what I really need because the Brass are capable of making more 'noise' and are generally not capable of playing as quietly. Also, Trumpet players get that excited rush of Testosterone when they see accents and ff (I am a Trumpet player so I know...)

Interesting again. My experience recording brass and woodwinds is pretty much limited to "horn bands", and in those styles everything is pretty much ff already!!!

I have enjoyed the opportunity to play with trumpet players, drummers, and yes, even guitar players who knew how to manage their levels... but I'll stipulate they aren't common. The one that completely floored me was an 18 year old kid, I told him he was a wee bit too loud and he kept ratcheting it down till everyone was happy. I never expected him to do that at the performance, but he did!

I had to forgo drummer jokes for a week!!!!

I think this is an excellent topic, and expect to learn a lot.

[JonFairhurst]

...Often, the Dead giveaway of a Sample library is in an awkward Fade that is attempting to sound like a Dynamic performance.

As you note, GPO already has filtering along with the volume control. I've been using GigaStudio's Dynamic Expression Filter recently in the GS3 domain, and it works quite well. It's not as versatile with setting up different curves/limits on volume and timbre as you mention, though that might be possible for those who generate custom filter coefficients for the DEF.

The critical thing is that changing the timbre rather than just volume is one of the most musical things you can do with samples. Dynamic vibrato (as heard with the Strad's Sonic Morphing), and realistic legato are the other key components. Given all that and some nice ambience, the only other thing one needs is artistry!

[johnmarkpainter]

After more random playing and experimentation....

From my Mellotron days (yes, I had a real one!) I was programmed to do Volume Pedal Fades from Zero. This another nasty habit I need to curb.

Allowing a soft Legato attack instead of a pure Fade In adds up...especially when done on many different instruments at once.

jmp

[DPDAN]

Recordings are all Compressed/Limited to limit the Dynamic range to something that can be reproduced (Radio/TV Broadcasts and Pop records are HEAVILY compressed as are most Films)

That is not true, most really good quality recordings of classical music are not compressed. However, once the finished audio soundtrack for a movie leaves the engineers hands, there is little he/she can do to make sure that it is not "messed with". All too often it is, compressed and or limited as you say, but generally speaking good recordings are not compressed. You are right about pop recordings, they are almost always compressed, and usually way too much IMO.
Just thought I would chime in... *ding*
Dan

[johnmarkpainter]

Classical recordings are often compressed/limited. Just hopefully not in an obvious manner.

Our Ears are very much like our eyes. We can view a scene with bright lights, dark shadows, sun light, lamp light etc... our eyes compensate for it. You can't take a picture of it though.

The recording process is similar. Recordings and playback equipment can't (yet) duplicate the experience of sitting in a room listening to an acoustic performance. Your ears can pick out extremely subtle sounds in a room.

Compression isn't just a signal processor. It can be a manual Fader ride, (riding a fader up in quiet passages is Expansion) It is also an artifact of microphones, preamps, analog media (Tape and LP's).

Ribbon mics compress quite a bit which is partially why they are often used in Orchestral (and particularly Filmscore) recordings. Tube mics do as well but in a more subtle way...and of course Tube Preamps.

All of these things add up to help "squeeze" the sound of an Orchestra in a room so that it can be played back on a pair of speakers.

Again...the Timbre change of the recording hits your ear as much as the actual volume changes to add up to a percieved Dynamic range that seems wider than it often really is.