simpsons-ish

[kealle]

I could definitely use some thoughts about this mix. I'm not sure what I'm doing in this regard, just trial and error.

http://soundcloud.com/keithleslie/simpsons-ish-remixt" onclick="window.open(this.href);return false;

thanks in advance

[Frodo]

Hey kealle:

Compositionally, it really captures Elfman's quirks nicely- and the cue is very well written. Congratulations!

What VIs are you using? I have other questions and thoughts, but I want to know more about what you've used, how you've used it, and (most importantly) where you feel it falls short.

[kealle]

Thanks for listening Frodo, and for the compliments.

The samples are from Kontakt2, which has a very modest amount of orchestral samples from the VSL. With the exception of the legato string samples, everything is "ensemble" size (10 to 14 players) and with only a handful of articulations. I'm not complaining, but I'd have done some things differently in terms of orchestration if I had more to work with. Hopefully soon that will change.

As for the mix, I had done a rough mix using the convolution reverb in Kontakt and then some minor mixing in DP, but decided to freeze all the tracks separately and do a fresh mix only in DP, which is the one you heard.

On the final mix, I grouped each instrument section together and tried applying reverbs, compression and EQ that way. On the strings, I also split the mixing between violin/viola and cello/basses. So each section was worked on separately, and then I tried balancing them with each other.

So I put compression on the instrument sections and used the leveler on the master track. I think some limiting on the instruments as well, and on the master fader. EQ's all over the place as well.

Seems like a bit much and maybe I got a little carried away. On the final mix that I posted, I can hear the compression, and I'm not very happy about that. And too much reverb all around. It just seemed that as I tried to remove or adjust anything, it never seemed to make it any better, or worse. It just stayed in this perpetual state of "not-quite right".

As for where I think it falls short, it's hard to say exactly. My intention is for it not to sound like real instruments, but rather bright and cartoon-ish, bouncy and playful. I'm afraid that my sample library, or lack their of, is the big culprit, but it's the mix as well.

thoughts?

Hey kealle:

Compositionally, it really captures Elfman's quirks nicely- and the cue is very well written. Congratulations!

What VIs are you using? I have other questions and thoughts, but I want to know more about what you've used, how you've used it, and (most importantly) where you feel it falls short.

[Frodo]

Hey kealle:

What I sense most of all right now is the need for a greater variety of articulations for each instrument. This is more an issue of sequencing rather than mixing. The convenience bundles which VSL released for Kontakt and MachFive only scratch the surface of what that library can really do.

I would first try alternating among 3-4 articulations each of long and short notes for each instrument

Check out a youtube link or pick up a CD and import a track of an Elfman-esque cue-- such as the opening to the Simpsons or the "Men In Black". Listen carefully to how the strings sound and compare that to what you've sequenced. Multiple legatos and multiple staccatos with a virtual instrument really go far to avoid that "machine gun" effect. The ear almost subconsciously picks up on those moments when a single sample is played long or short as opposed to using as many different samples as one's machine can handle to the same effect or better.

If you like the VSL samples, I would strongly recommend getting the Special Edition of VSL at the very least. There are other options out there, such as EWQLSO as well as the full (and expensive!!!!!!) version of VSL.

But before money becomes an issue, I'd like to ask how much automation have you employed for the sake of expression?

Additional articulations and automation are the first steps (and quite extensive ones) to make. What follows are matters of engineering the mix. We can get into the finer points of mixing a virtual orchestra in due time.

I will stress that your ability to compose and to visualize is most admirable. Bravo!!

[kealle]

Frodo,

One of the reasons that I froze all the tracks was to have more control over the automation aspect of the mix; a seemingly impossible task to perform on the MIDI data itself. So yes, I did do a fair amount, but was it enough? Or effective? I'm leaning towards no.

Prior to freezing the MIDI tracks, I spent a fair amount of time with articulations, trying to make each melodic line have more character. Even so, it still lacks. Having only ensemble sections, everything sounds heavier than it should, at least from how I hear it in my head. This in regards to density, dynamics, and orchestration.

This most recent mix I think I realized, and in agreement with you, that mixing must be more than just reverb and compression. Which is why I decided to spend energy on automation and more minute details in the MIDI data.

I've looked at VSL special edition and will eventually invest in it, when I'm able. It's beyond necessity to do so, clearly.

I do need to listen to Elfman more. This was done more as a recollection of rather than a study/emulation of his music, which I love.

By multiple legatos and staccatos do you mean to double the part/track?

I intend to keep working on this one, and have been for a while, until I get it right. So again, thanks for the feedback and compliments.

I have a few general mix questions, in no particular order, although perhaps I'm jumping the gun:

Should limiting be applied to the master track during the mix, or is this more for the mastering phase? I'm never sure what decibel level I should be shooting for on either the master fader or the instrument tracks. I've been putting the limiter on the master, using the "final mix" preset and using that to keep the levels in check. This seems a bit generic to me, but I know that it pushes everything toward the edge without going over, but I think I'm losing something in the sound.

Is the leveler effective when used during mixing (and on MIDI in particular) or is this used more for live recording of audio?

Reverb, EQ, Compression: Best on group tracks, individual tracks, master fader? Any and all? Or is it a case-by-case basis?

Quantizing? This version is pretty strict, time-wise, and therefore a bit lifeless.

feel free to not address any of these questions if you wish. Or in due time.

thanks,
kealle

Hey kealle:

What I sense most of all right now is the need for a greater variety of articulations for each instrument. This is more an issue of sequencing rather than mixing. The convenience bundles which VSL released for Kontakt and MachFive only scratch the surface of what that library can really do.

I would first try alternating among 3-4 articulations each of long and short notes for each instrument

Check out a youtube link or pick up a CD and import a track of an Elfman-esque cue-- such as the opening to the Simpsons or the "Men In Black". Listen carefully to how the strings sound and compare that to what you've sequenced. Multiple legatos and multiple staccatos with a virtual instrument really go far to avoid that "machine gun" effect. The ear almost subconsciously picks up on those moments when a single sample is played long or short as opposed to using as many different samples as one's machine can handle to the same effect or better.

If you like the VSL samples, I would strongly recommend getting the Special Edition of VSL at the very least. There are other options out there, such as EWQLSO as well as the full (and expensive!!!!!!) version of VSL.

But before money becomes an issue, I'd like to ask how much automation have you employed for the sake of expression?

Additional articulations and automation are the first steps (and quite extensive ones) to make. What follows are matters of engineering the mix. We can get into the finer points of mixing a virtual orchestra in due time.

I will stress that your ability to compose and to visualize is most admirable. Bravo!!

[Phil O]

As for the mix, I had done a rough mix using the convolution reverb in Kontakt and then some minor mixing in DP...

First off, listen to Mr. Frodo. He's got his poop in a group when it comes to this stuff.

As far as the mix is concerned, my first reaction was that the stereo image was sounding a little smeared. Could be the reverb you're using - not sure. You might experiment with the width of the reverb, or perhaps more than one instance with some creative positioning. I know that Shooshie swears by Altiverb, but I don't know if there's room in your budget for more toys. In any case that was my initial impression. Other than that, it's sounding pretty good to me.

Phil

[kealle]

I anxiously await to hear more from Mr. Frodo.

As for my budget, Altiverb is on the list, but right behind VSL Special Edition.

I'm just using the DP Convolution reverb, and regrettably, I know nothing about "stereo image", but "smeared" sounds pretty accurate.

Thanks for the feedback.

As for the mix, I had done a rough mix using the convolution reverb in Kontakt and then some minor mixing in DP...

First off, listen to Mr. Frodo. He's got his poop in a group when it comes to this stuff.

As far as the mix is concerned, my first reaction was that the stereo image was sounding a little smeared. Could be the reverb you're using - not sure. You might experiment with the width of the reverb, or perhaps more than one instance with some creative positioning. I know that Shooshie swears by Altiverb, but I don't know if there's room in your budget for more toys. In any case that was my initial impression. Other than that, it's sounding pretty good to me.

Phil

[Frodo]

Frodo,

One of the reasons that I froze all the tracks was to have more control over the automation aspect of the mix; a seemingly impossible task to perform on the MIDI data itself. So yes, I did do a fair amount, but was it enough? Or effective? I'm leaning towards no.

When doing an orchestral project even a finished mix will never have "enough" automation. Tweaking can go on forever. Eventually, it becomes a matter of just letting it go while balancing that sense with that of a feeling that it has gone "far enough". I always feel as if my VI automation could use "more", no matter how much time I've spent on it. That's because I'm always learning different and better ways to accomplish the same tasks even as I work on a single sequence. There's a beauty in that somewhere even as it requires tearing one's hair out by the roots at times!!

Prior to freezing the MIDI tracks, I spent a fair amount of time with articulations, trying to make each melodic line have more character. Even so, it still lacks. Having only ensemble sections, everything sounds heavier than it should, at least from how I hear it in my head. This in regards to density, dynamics, and orchestration.

This most recent mix I think I realized, and in agreement with you, that mixing must be more than just reverb and compression. Which is why I decided to spend energy on automation and more minute details in the MIDI data.

I've looked at VSL special edition and will eventually invest in it, when I'm able. It's beyond necessity to do so, clearly.

I do need to listen to Elfman more. This was done more as a recollection of rather than a study/emulation of his music, which I love.

Listening to Elfman's scores is important, but only part of the process. Flying in a track into DP along side your project for constant A/B comparison is quite another matter entirely. Having a frequency analayzer to take snapshots of your mix as well as the source reference track is also critical.

By multiple legatos and staccatos do you mean to double the part/track?

While doubling parts and tracks with different, similar samples is a great technique for addressing matters of texture, in this instance having multiple legatos and staccatos used linearly is what was intended. When creating a long, melodic violin line it's good to alternate among 3-4-5 different legato patches as often as possible even if it means keyswitching every 2-3 notes. With shorter notes and fast runs, the same is true. Many orchestral libraries address the problem of "the machine gun effect" by including a "round robin" patch where the samples change with every strike of the keyboard. It creates less predictability and less ear fatigue as the ear quickly recognizes that the exact same articulation might have been used in repeated succession.

I intend to keep working on this one, and have been for a while, until I get it right. So again, thanks for the feedback and compliments.

I have a few general mix questions, in no particular order, although perhaps I'm jumping the gun:

Should limiting be applied to the master track during the mix, or is this more for the mastering phase? I'm never sure what decibel level I should be shooting for on either the master fader or the instrument tracks. I've been putting the limiter on the master, using the "final mix" preset and using that to keep the levels in check. This seems a bit generic to me, but I know that it pushes everything toward the edge without going over, but I think I'm losing something in the sound.

I tend not to use limiting or compression until the very end. My reasons for this embrace the hard work the programmers of the library have done already in creating the library with high end gear. For the DAW user, it should be largely a matter of MIDI velocity and CC7 or CC11 automation where dB levels are concerned-- at least at first.

Use effects only as necessary-- and be sure you recognize that necessity before applying that effect. The MIDI balance should sound "good" as is if it is well balanced. However, using a source track (such as the Elfman example) should reveal more specific needs for EQ, compression, and limiting--- as well as application of reverb.

To the issue of reverb, I would recommend an IR such as DPs ProVerb or, if affordable, the immaculate Altiverb. Check out this link:

http://www.audioease.com/Pages/Altiverb ... tions.html" onclick="window.open(this.href);return false;

It deals with psycho-acoustic aspects of applying ambience to orchestral mixes and a bit of a "how to". Regardless of what reverb one chooses, the principles are the same.

But--- be aware that this is part of the engineering phase of the process and not part of the MIDI programming phase. You wear multiple "hats" when creating mock-ups. You are a MIDI programmer first--- then you become a mixing engineer. After that point, you become a mastering engineer. These are VERY different thought processes that most DAW users take for granted simply because a DAW enables one person to do it all.

Is the leveler effective when used during mixing (and on MIDI in particular) or is this used more for live recording of audio?

Again, I would focus on leveling the volume and expression automation first and save as much as possible while saving compression and limiting for the mastering phase.

Reverb, EQ, Compression: Best on group tracks, individual tracks, master fader? Any and all? Or is it a case-by-case basis?

That depends on how one wants to work. I've made groups: woodwinds, brass, percussion, and strings with some sub-groups as needed-- but only in the mastering phase. The timpani, for example, may need a little more punch (being in the back of the orchestra), so its compressor setting might be different than that of other instrumental groups.

Running automation on effects should not be overlooked, either. There are times when a particular passage of music requires more twiddling than others. One setting rarely suits the entirety of a single track or project.

Quantizing? This version is pretty strict, time-wise, and therefore a bit lifeless.

Dude-- quantizing is NOT your friend. When doing a mock-up, it's better to slow the tempo down--- WAAAAAY down--- and to play the part in real time. There ought to be very few instances when quantizing should be used.

Even if you copy-paste a part, experiment with turning off the grid in the MIDI editor (aka: GE or Graphic Editor) and tap the right or left arrows a few times to offset the timing a bit. Orchestral programming is largely about the imperfections. Not even the best orchestras play "perfectly".

feel free to not address any of these questions if you wish. Or in due time.

thanks,
kealle

One of the reasons why I wanted to respond at all is because you post has been just sitting here for many days feeling lonely. You've raised some important questions that rarely get answered adequately--- and my hope is that others will chime in with constructive tips and techniques that will also help you.

So many DAW users struggle with the composition itself. This is not your problem. You've written a great snippet for which I again commend you.

The issue now is not about being a composer. It's about being a MIDI programmer, then about thinking as an instrumentalist for each instrument in your virtual orchestra. After that, it's about being a tracking engineer. Beyond that, it's about being a mixing engineer and later a mastering engineer. Understanding each phase of that process is the key to your success!!

[Frodo]

First off, listen to Mr. Frodo. He's got his poop in a group when it comes to this stuff.

LOL!! "Poop in a group". Hobbits have been known to be hoarders of fertilizer!!