From legato to Staccato and back with problems

[GarretGraves]

First the need answer I need most at the moment:

How to go from legato strings to staccato strings and back in unison using EWQLSO Platinum Plus whilst sounding realistic. Can't seem to do this right. Samples don't connect naturally.

Now the story (Skip if you want. I'm only telling this to show what I've done to solve this so far):

I've recently upgraded my orchestra library to EWQLSO Platinum Plus and got rid of that old Kompakt version I had. The PLAY engine is giving me some nice control over things now. But I'm still having an issue in the realism department. I've looked on the internet for some tutorials on this matter but there were only a few and only 1 of them was only sort of helpful (and somewhat poorly put together I might add.) Just to give you an idea:

This guy gave more of a tour on Cubase 5 than how to use the SO.

This guy did a better job in regards to sustained notes and DXF's but didnt mention what samples he used nor did he mention stacatto at all. 48 minutes of sustained lines?! sheesh

So far that's all I've found as far as tutorials. I've read the manual but it only tells you what the samples are and what they're suppose to do but they don't describe HOW you're suppose to do it. I've been on the forums of East West and tried to get some answers there but since I'm a new member there it takes until the forum moderator approves my posts that people can see them. (Or maybe no one is taking the time to respond )

Which as led me back here. I only mentioned the above to show I've made the effort to learn on my own and you know where I've looked. Does ANYONE know where I can learn how to use this properly or at least teach me how to go from legato strings to staccato strings and back and making it sound natural?

[Moseph]

Can you post an audio example of what you're working on?

Depending on the context, it might sound more natural to use a marcato, portato, or detache articulation rather than a staccato, which would give you a bit more sustain on the note and might help it better blend into a legato passage. You sometime have to play around with the articulations, and the one you end up using may not always be the obvious choice.

EDIT: I'm not sure specifically what other articulations EWQL SO has to choose from since I don't use it; I just know it has a lot.

[GarretGraves]

https://soundcloud.com/soleparadigm/orchestra-test-1

Sorry for the low-volume mix. Didnt have time to add a limiter or anything.

My mixing is also an issue. I lack a certain warmth you'd expect from an orchestra.

[Moseph]

For most of the staccato notes, I'd try to find a slightly longer articulation to use instead. VSL has an articulation called short detache that I would use in this situation; EWQL SO probably has something similar but it may have a different name.

The fifth, sixth, ninth and tenth staccato notes (that is, the really fast ones at 0:12-0:13) I would leave as staccato, or if EWQL has spiccato articulations I might also try that.

[dannthr]

You should be using dxf or xfd patches with your sustains and sculpting them by modulating dynamics using a midi continuous controller.

[GarretGraves]

My midi controller is busted. I use "edit events" in the piano roll to automate. Little bit harder but it gets the job done. That part I understand and need to toy with. Its just that there's SO MANY samples I don't know what the right one is for what I'm doing. Whatever will make it sound natural.

As for the mixing goes, I'm kind of having a hard time in that area too. I'm listening to some more orchestra music like the 25th anniversary Zelda disc that came with skyward sword. So far all I could match closely is the panning for the strings. (Violins and violas to the left, cellos and basses to the right) But my mix feels a bit cold and somewhat lifeless compared the to disc. Is it possible to match the quality of a real orchestra with EWQLSO? I know it can get awesome. But I've become such an audiophile in recent years that I badly want that level of authenticity out of my gear.

[dannthr]

I understand and appreciate your frustration.

Please understand that while EWQLSO is a brilliantly engineered library (Dennis Sands' preliminary engineering on that prior to noise reduction post processing is brilliant), the library is very old and sampled with a philosophy that layering and stitching together different patches is what is required to get a really believable performance.

This is what we call the Old Skool.

Amazing things have been produced and created in the old skool, but it requires a great deal of effort from the programmer to get it right.

Believe me when I say, my first gig in games was programming a famous composer's Symphonic Choirs composition. It's work, no one wants to do it, but it can be done.

A good solo performance might require as many as 4 or 5 different patches, stitched together in sequence to serve a single performance.

[GarretGraves]

I understand and appreciate your frustration.

Please understand that while EWQLSO is a brilliantly engineered library (Dennis Sands' preliminary engineering on that prior to noise reduction post processing is brilliant), the library is very old and sampled with a philosophy that layering and stitching together different patches is what is required to get a really believable performance.

This is what we call the Old Skool.

Amazing things have been produced and created in the old skool, but it requires a great deal of effort from the programmer to get it right.

Believe me when I say, my first gig in games was programming a famous composer's Symphonic Choirs composition. It's work, no one wants to do it, but it can be done.

A good solo performance might require as many as 4 or 5 different patches, stitched together in sequence to serve a single performance.

I figured as much and I'm willing to do the work. No question. But I'm surprised that almost no one has posted a tutorial or anything in regards to an in depth look at the library and how to use it. One guy so far as been attempting to make a series of videos to help people out but he's been to busy to make them. Its funny. You'd think someone out there after all this time would have gone and done something like that. I've seen videos for basically everything else like Massive and Absynth and FM8. It would seriously be helpful to laymens like me to learn some as initially complex as this.

Best I can do now is bug forums with questions. and attempt things myself of course.

EDIT: Now when you say old, are you telling me there are better libraries out there that simplify the work?

[Moseph]

As far as getting a properly warm orchestra sound, I've found these things helpful:

1) Idiomatic arrangement. For example, don't use a four french horn unison patch for your (single) horn line and leave it at that. Write four individual horn parts in harmony for four individual horns. Use two of each woodwind. And double your parts liberally. Try using the basses to double the cellos an octave lower. Try doubling the violin 1 part in the viola. Try doubling the violin 1 part in all of the woodwinds. Things like that.

2) Read up on how to use a multiband compressor (or even single-band) on the master channel. When done right, this can help squish everything together so it sounds more cohesive and less like a bunch of separate instruments off in their own worlds.

3) Running everything through an analog tape sim can help with warmth (I use Nomad Factory's Magnetic II for this). I've also started using EQs that deliberately color the sound in an analog-y way rather than using more transparent EQs.

4) Reverb settings are important (though I'm not sure the best way to manage a wet library like EWQL SO since I use mostly very dry orchestra samples). Using a reverb with a lot of pronounced early reflections can increase fullness and warmth.

5)This is not so much a warmth issue as general programming, but I use a two-axis virtual CC controller (AnalogX MIDI Mouse Mod) routed through a virtual interface (LoopBe1) to program velocity crossfades. I have vel crossfade mapped to the y-axis and expression (CC 11) mapped to the x-axis. This lets me fine-tune levels more than using only crossfade would. For example, if there's an unpleasant break between two velocity layers that I want to avoid, I can run the crossfade up just below the break and then boost expression instead. I find that recording all of this live is faster and produces better results than drawing curves, and you don't need any hardware to do it.

[dannthr]

I figured as much and I'm willing to do the work. No question. But I'm surprised that almost no one has posted a tutorial or anything in regards to an in depth look at the library and how to use it. One guy so far as been attempting to make a series of videos to help people out but he's been to busy to make them. Its funny. You'd think someone out there after all this time would have gone and done something like that. I've seen videos for basically everything else like Massive and Absynth and FM8. It would seriously be helpful to laymens like me to learn some as initially complex as this.

Best I can do now is bug forums with questions. and attempt things myself of course.

EDIT: Now when you say old, are you telling me there are better libraries out there that simplify the work?

Yes.

And yes to idiomatic writing.

And don't say I never did anything for ya!

BAM!

[GarretGraves]

Ah! CLICK!

But it was as I feared. I'm gonna need more than one instance of Play to fulfil my needs. Was hoping to avoid that but I guess it was inevitable.

Thanks Dan! And thanks for the honorable mention. haha

[dannthr]

Ah! CLICK!

But it was as I feared. I'm gonna need more than one instance of Play to fulfil my needs. Was hoping to avoid that but I guess it was inevitable.

Thanks Dan! And thanks for the honorable mention. haha

Naw, you don't have to, I was just being lazy.

You can use a Rack Instrument and route multiple MIDI channels to it for each articulation.

PLAY supports 16 articulations in a single instance--more than enough for my example.

EDIT: I put the same setup together at work, here's a screen shot of PLAY as a Multi-Timbral instrument placed on Cubase's Instrument Rack with 5 MIDI tracks routed to 5 of 16 MIDI channels available in PLAY, with all of the articulations on a different MIDI channel.

http://www.dannthr.com/temp/ewqlsoplayss_multitimbral.png

[GarretGraves]

Naw, you don't have to, I was just being lazy.

You can use a Rack Instrument and route multiple MIDI channels to it for each articulation.

PLAY supports 16 articulations in a single instance--more than enough for my example.

EDIT: I put the same setup together at work, here's a screen shot of PLAY as a Multi-Timbral instrument placed on Cubase's Instrument Rack with 5 MIDI tracks routed to 5 of 16 MIDI channels available in PLAY, with all of the articulations on a different MIDI channel.

http://www.dannthr.com/temp/ewqlsoplayss_multitimbral.png

I think this is what I'm doing now actually. I got Play set to a port and then a handful of MIDI outs routed in FL Studio for each articulation. That the same thing? Sounds like it.

What I meant was Im gonna need more than one instance of Play because I think I'm gonna need more than 16 channels for the articulations for the other string instruments.

[dannthr]

I think this is what I'm doing now actually. I got Play set to a port and then a handful of MIDI outs routed in FL Studio for each articulation. That the same thing? Sounds like it.

What I meant was Im gonna need more than one instance of Play because I think I'm gonna need more than 16 channels for the articulations for the other string instruments.

Yeah, in my current orchestral templates, I have an instance of Kontakt for each String group (so 5 total), I have an instance for High Brass and Low Brass each, and so on and so forth.

[GarretGraves]

Yeah, in my current orchestral templates, I have an instance of Kontakt for each String group (so 5 total), I have an instance for High Brass and Low Brass each, and so on and so forth.

Yeah this scares me a little cause I'm worried my machine won't be able to take that much at once. Got an Intel i5 processor with 16gigs of RAM and all my drives are 7200rpm. Sounds like an up to date machine but i still freak out when I start loading all kinds of shit.

[GarretGraves]

https://soundcloud.com/soleparadigm/tr-strings-demo

How's this sound? I feel the difference in the stitching together.

[dannthr]

https://soundcloud.com/soleparadigm/tr-strings-demo

How's this sound? I feel the difference in the stitching together.

Nice, sounds better. Try using DXF patches on the long sustains, so you can sculpt and shape it--think of the players arching their bows across the strings and the energy of that motion translating to dynamic changes.

Let yourself detach the strings slightly between phrases (this means finding the beginning and ending of individual phrases, and treating those like a light arc, with a climax and everything).

[GarretGraves]

Nice, sounds better. Try using DXF patches on the long sustains, so you can sculpt and shape it--think of the players arching their bows across the strings and the energy of that motion translating to dynamic changes.

Let yourself detach the strings slightly between phrases (this means finding the beginning and ending of individual phrases, and treating those like a light arc, with a climax and everything).

I've done that. But I'm curious as to how low to begin and end. I don't want it too quiet cause this is going over a rock track and if it gets too low it'll sound like it disappears.

[dannthr]

I've done that. But I'm curious as to how low to begin and end. I don't want it too quiet cause this is going over a rock track and if it gets too low it'll sound like it disappears.

You can probably go lower than you think--dynamic range is what makes something feel big.

It's an issue of contrast.

If you have a picture with white and black, then the black will look blacker than if the picture had gray and black.

Contrast.

Quiet stuff surrounding big stuff is what makes the big stuff sound big.