Repetitive Ostinato Torture Test

[pderbidge]

Here a quick test with Berlin Strings and SWAM Cello + Viola. Sorry but I had a hard day and I didn't manage to do better, I hope to do something in the next days.

[AUDIOPLUS=https://vi-control.net/community/attachments/spiccato-test-1-mp3.14733/][/AUDIOPLUS]

Here with a light pizzicato to accent notes:

[AUDIOPLUS=https://vi-control.net/community/attachments/spiccato-test-2-mp3.14734/][/AUDIOPLUS]

Here with some cello on it:

[AUDIOPLUS=https://vi-control.net/community/attachments/spiccato-test-3-mp3.14735/][/AUDIOPLUS]

Thanks Daniele, glad to see someone else sharing. This is a lot closer to what I was getting with my other string libraries. A much more metallic sound than a string sound. I wonder if that's more from the SWAM instrument than Berlin Strings? Berlin Strings is a very coveted library and yet you can see even as great as it is there are phrases like this that will challenge it. I really like the idea of the Pizzicato accent notes. That's another great way to not only help disguise the machine gun effect but also break up the monotony of the repetitive notes.

I'm still very curious to know how Performance Samples Fluid shorts will sound. Hoping someone who owns it will contribute.

So Far I'm liking CSS and Albion One the most and it has nothing to do with anybody's programming, I think they are just more capable in this scenario.

 

[pderbidge]

FYI: I let the limiter hit too hard in my examples so the dynamics are completely lost. I knew I shouldn't have used one but worried people would think it was too soft. When I get a chance I will upload new examples without the limiter and you'll just have to turn up the volume to get a good level.

 

[pderbidge]

I took off the Limiter so the files should have "some" dynamics restored. I also updated the first post with these files.





Edit: If your new to this discussion. These files are meant to be raw to see how close we can get a sample Library or blend of libraries to perform repetitive Ostinato sounds before the post processing (reverb, saturation, eq, etc..) phase. The whole idea of starting with the best possible sound first before trying to fix in the mix.

 

[pderbidge]

SWAM kinda ruined Berlin beautiful sound

If you have Berlin Strings, please do share. To be fair, I have not found any of my libraries (see my list in the first post) to perform a repetitive phrase like this satisfactorily with exception to blending Albion One with a couple of others. Even Lass, who's generally good with shorts doesn't cut it in this scenario.

 

[jbuhler]

These are all Spitfire: 1. Chamber Strings, Ensemble + Individual sections; 2. Chamber Strings as in 1 + Symphonic Strings, Ensemble + first Vln; 3. SCS + SSS Ensembles. Vla playing eighths, Vc doubling Vla an octave below.

Edit to add: Mics: Ensembles T+A; Sections C.

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-mp3.14740/][/AUDIOPLUS]

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-sss-mp3.14741/][/AUDIOPLUS]

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-sssensemble-mp3.14743/][/AUDIOPLUS]

 

[DANIELE]

SWAM kinda ruined Berlin beautiful sound

As I said this is a quick test so I didn't manage to pay more attention to expression and various controls. With SWAM you could look at bow pressure or bow position to give more randomness to the sound. I used them in my example to give more bite to the sound, I could extend the notes to give less bite and more mellowness, this is the good thing about swam.

I'll try to do some more tests and I'll give you an example with BS alone.

One question: Do we have to use the entire ensemble? Because libraries like Albion One are ensemble libraries, I tend to use very rarely this instruments, I like to go more with the classic way by using sections.
We are talking about different things here.

An ensemble patch has the resonance of all the group of instruments used, a section is different.

In my example I'm in my orchestral template so I have reverb and positioning of the instruments.

 

[shawnsingh]

I've attached an mp3 and a screenshot of some of the MIDI note entry.

Various thoughts:

• The mix I used: Hollywood Strings gold violin 1, loose spiccato patch + Orchestral Tools solo Violin 1 spiccato. I added a light amount of convolution reverb (concert hall style) to the Hollywood Strings, and used only the tree mic for the solo Violin. [Edit: It might be possible to get a bit more spiccato crunch on the attacks by adding more close mic position, but I didn't try that here]
• Set the level of the solo violin so that it hides under the Hollywood Strings. In this scenario, the solo violin adds a fuller body to the tone where the hollywood strings at medium-lower velocities was slightly raspy by itself. I think this "fuller body" is not as common a purpose for layering solo strings, it's usually more a purpose of adding the slight clarity or piercing quality, but not in this case. Layering the OT solo violin can also help add a crisp attack (like when layering with Metropolis Ark 1), though that wasn't actually the case in my clip here, the hollywood strings already had enough of the rosin crunchy sound.
• OT solo violin 2 (I did not use here) has a very different, choppy spiccato style, compared to solo violin 1's more rounded style. That could be a useful different option to layer.
• Hollywood Strings spiccato patch responded well to shortening the MIDI notes. I wouldn't be surprised if some of these other libraries you all are trying could have a similar effect. Many patches probably won't, but some will. I know VSL especially worked really well this way since they were dry samples. So one staccato patch from VSL dry libraries was flexible enough to provide a broad spectrum of short note lengths (too bad they didn't have enough round robins and velocity layers - that in combination with ability to change note length would have been a superb combo)
• I do agree with what others have already said about humanizing velocities and note timing, but I think there is a method to the madness - actually random humanizing might not be as helpful as targeted tricks. In the screenshot I attached, the long MIDI note is one of the downbeats that starts another group of 4 measures. In the measure before that, I moved some of the notes away from that downbeat, and the note after the downbeat, I pushed in the other direction, also away from the downbeat. This emulates the psychological emphasis that a musician puts on the accent and trying to "time" exactly the downbeat - they naturally add a bit of space by human error, but it's not really random. Similar approach I find works well to make string runs sound a bit more natural - performers tend to rush notes slightly so that they have room for error to arrive at a beat with the correct note on time.
• Same idea with note velocities - when trying to keep a rhythm like this, musicians will try to mentally keep track of groupings of measures, e.g. maybe group of 4 measures. They will naturally very slightly accent the first note of each group for musical effect. Also, when string players try to do accents on fast passages, often they affect nearby notes too, because of how they try to manage bow speed and bow position to hit the accent. So nearby notes can get slightly louder around an accent when fast notes are happening.
• To be clear, the accents you might obviously hear at around 0:18 and 0:27 in my version - those are NOT the accents I'm referring to in the point above. The accents i'm referring to are more subtle and occur every 4 measures (e.g. around 0:05, 0:09, etc).

Cheers!



[AUDIOPLUS=https://vi-control.net/community/attachments/spiccato-example-mp3.14745/][/AUDIOPLUS]

 

[Lassi Tani]

I did a test with Berlin Strings: Violins 1 Spiccato layered with Violins 1 Doubles (reduced channel volume), which have that blurry sound of strings playing double notes. It also has violas doing the same pattern without the double notes for accenting.

I think in such passages that @pderbidge posted, strings wouldn't have time to play each note clearly, and the double notes will sound blurry.

 

[procreative]

I think though its the constant two note riff that contributes to the sense of the mechanical. Its a bit like those poor backing singers in Lionel Ritchie's All Night Long singing "all night" ad infinitum!

 

[DANIELE]

I think though its the constant two note riff that contributes to the sense of the mechanical. Its a bit like those poor backing singers in Lionel Ritchie's All Night Long singing "all night" ad infinitum!

I think the same. The pattern and bpm together tend to give a very mechanical result in terms of sound. I think you'll hear some of it in a real enviroment with real players too.

Layering and accenting is the right way to achieve a more organic sound.

 

[Ben]

I layered several VSL libraries (Solo, Chamber, Orchestra, Appassionata con Sordino) and added some reverb (MIRx + MIRacle), compression + eq:

[AUDIOPLUS=https://vi-control.net/community/attachments/vsl_ostinatotest-mp3.14750/][/AUDIOPLUS]

Just the pattern without reverb and any FX:
[AUDIOPLUS=https://vi-control.net/community/attachments/vsl_ostinatotest_nofx-mp3.14752/][/AUDIOPLUS]

Cubase project attached.

 

[Parsifal666]

The Albion One shorts sound a lot better when you use the close mic alone imo. As they are in your example, they are washing out everything.

I could be wrong...it's just that I've heard that A1 room sound so many times, especially back when I was trying to find some way of integrating that library into drier samples.

Though I'm not familiar with Chris Hein Ensemble Strings, I have his other libraries and I believe in this case you'd happily forgo the A1 in favor of drier stuff (EWHS is also great if you don't mind getting your hands dirty).

 

[pderbidge]

These are all Spitfire: 1. Chamber Strings, Ensemble + Individual sections; 2. Chamber Strings as in 1 + Symphonic Strings, Ensemble + first Vln; 3. SCS + SSS Ensembles. Vla playing eighths, Vc doubling Vla an octave below.

Edit to add: Mics: Ensembles T+A; Sections C.

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-mp3.14740/][/AUDIOPLUS]

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-sss-mp3.14741/][/AUDIOPLUS]

[AUDIOPLUS=https://vi-control.net/community/attachments/vampscs-sssensemble-mp3.14743/][/AUDIOPLUS]

Thanks for stepping in jbuhler. I can see where the longer bass notes can help with the shorter upper note pattern. I would probably consider that as probably a something almost not heard but still felt. Could be a nice trick.

 

[pderbidge]

I've attached an mp3 and a screenshot of some of the MIDI note entry.

Various thoughts:

• The mix I used: Hollywood Strings gold violin 1, loose spiccato patch + Orchestral Tools solo Violin 1 spiccato. I added a light amount of convolution reverb (concert hall style) to the Hollywood Strings, and used only the tree mic for the solo Violin. [Edit: It might be possible to get a bit more spiccato crunch on the attacks by adding more close mic position, but I didn't try that here]
• Set the level of the solo violin so that it hides under the Hollywood Strings. In this scenario, the solo violin adds a fuller body to the tone where the hollywood strings at medium-lower velocities was slightly raspy by itself. I think this "fuller body" is not as common a purpose for layering solo strings, it's usually more a purpose of adding the slight clarity or piercing quality, but not in this case. Layering the OT solo violin can also help add a crisp attack (like when layering with Metropolis Ark 1), though that wasn't actually the case in my clip here, the hollywood strings already had enough of the rosin crunchy sound.
• OT solo violin 2 (I did not use here) has a very different, choppy spiccato style, compared to solo violin 1's more rounded style. That could be a useful different option to layer.
• Hollywood Strings spiccato patch responded well to shortening the MIDI notes. I wouldn't be surprised if some of these other libraries you all are trying could have a similar effect. Many patches probably won't, but some will. I know VSL especially worked really well this way since they were dry samples. So one staccato patch from VSL dry libraries was flexible enough to provide a broad spectrum of short note lengths (too bad they didn't have enough round robins and velocity layers - that in combination with ability to change note length would have been a superb combo)
• I do agree with what others have already said about humanizing velocities and note timing, but I think there is a method to the madness - actually random humanizing might not be as helpful as targeted tricks. In the screenshot I attached, the long MIDI note is one of the downbeats that starts another group of 4 measures. In the measure before that, I moved some of the notes away from that downbeat, and the note after the downbeat, I pushed in the other direction, also away from the downbeat. This emulates the psychological emphasis that a musician puts on the accent and trying to "time" exactly the downbeat - they naturally add a bit of space by human error, but it's not really random. Similar approach I find works well to make string runs sound a bit more natural - performers tend to rush notes slightly so that they have room for error to arrive at a beat with the correct note on time.
• Same idea with note velocities - when trying to keep a rhythm like this, musicians will try to mentally keep track of groupings of measures, e.g. maybe group of 4 measures. They will naturally very slightly accent the first note of each group for musical effect. Also, when string players try to do accents on fast passages, often they affect nearby notes too, because of how they try to manage bow speed and bow position to hit the accent. So nearby notes can get slightly louder around an accent when fast notes are happening.
• To be clear, the accents you might obviously hear at around 0:18 and 0:27 in my version - those are NOT the accents I'm referring to in the point above. The accents i'm referring to are more subtle and occur every 4 measures (e.g. around 0:05, 0:09, etc).

Cheers!



[AUDIOPLUS=https://vi-control.net/community/attachments/spiccato-example-mp3.14745/][/AUDIOPLUS]

This is fantastic. You made me realize I failed to mention Hollywood Strings and EWQL in my list of String Libraries, and I did try them as well. I remember spending the most time with the x9 round robin sections and I did like the shorts in Hollywood Strings but I didn't get results like this. Good job. I thinkg your methodology makes perfect sense. I was thinking of eventually doing the same as you as regards to a more targeted randomization instead of a random randomization and plus your idea of the accents every few measures is brilliant. Do you find OTS to be adding a lot to the sound? Unfortunately that is one lib I don't own. Also, I wonder how much Play's built in Convo affects this sound? I would prefer a little dryer until I had enough instruments in my composition to decide just how much Convolution verb works with the rest of the composition.

I think I need to put HS back on my list of possibilities and consider your methodology of how to better approach the midi manipulation.

 

[pderbidge]

I did a test with Berlin Strings: Violins 1 Spiccato layered with Violins 1 Doubles (reduced channel volume), which have that blurry sound of strings playing double notes. It also has violas doing the same pattern without the double notes for accenting.

I think in such passages that @pderbidge posted, strings wouldn't have time to play each note clearly, and the double notes will sound blurry.

I think you're right. Blur is definitely one of the keys here.

 

[pderbidge]

I think though its the constant two note riff that contributes to the sense of the mechanical. Its a bit like those poor backing singers in Lionel Ritchie's All Night Long singing "all night" ad infinitum!

I can't disagree. When I hear this type of two note phrase, usually in a tense action piece type of arrangement, it is used as an underscore so it sounds less monotonous as it gives the composition some drive but the other strings or instruments are the main show. In my piece I did use it as an underscore, however there were those two bars at the beginning where it is naked and exposed which is made worse by the fact that it's the start of the piece. I've been mulling over ideas to remedy that. Having said that I think a real string section would still not be nearly as mechanical sounding as the Sampled Strings I've tried so far. I do wonder how much less work this would be with Fluid Shorts. Anyone with Fluid shorts willing to give it a try?

This video on Fluid Shorts is what's got me convinced so far

 

[pderbidge]

I layered several VSL libraries (Solo, Chamber, Orchestra, Appassionata con Sordino) and added some reverb (MIRx + MIRacle), compression + eq:

[AUDIOPLUS=https://vi-control.net/community/attachments/vsl_ostinatotest-mp3.14750/][/AUDIOPLUS]

Just the pattern without reverb and any FX:
[AUDIOPLUS=https://vi-control.net/community/attachments/vsl_ostinatotest_nofx-mp3.14752/][/AUDIOPLUS]

Cubase project attached.

Thanks Ben, good to see VSL in this test. What articulations are you using?

 

[pderbidge]

The Albion One shorts sound a lot better when you use the close mic alone imo. As they are in your example, they are washing out everything.

I could be wrong...it's just that I've heard that A1 room sound so many times, especially back when I was trying to find some way of integrating that library into drier samples.

Though I'm not familiar with Chris Hein Ensemble Strings, I have his other libraries and I believe in this case you'd happily forgo the A1 in favor of drier stuff (EWHS is also great if you don't mind getting your hands dirty).

I tend to agree with you on this. In the beginning the lack of space might sound less appealing but in the mixing phase when gelling all the instruments it should all come together. It's a tricky balance to train ones ear to hear when a sample is "good enough" to where the Mix will fix the rest. I understand the idea of not always trying to fix in the mix but there are times where mixing is the right phase for gelling things together.

In my example, however I think Signor Paganini is actually adding more room sound than Albion One so I think it would also be necessary to turn down the reverb there as well with it's built in reverb knob.

 

[pderbidge]

Thank you all so far for your contributions. The one thing I've learned that I should have paid more attention to is "Accent, Accent, Accent" I think shawnsingh really drove this home to me although I can see many others making this point as well. I think I was more focused on randomizing, which is still important but didn't consider key accent points. Even with a possibly more capable library like Fluid shorts all the tips and tricks made in this thread should still be adhered to. You guys are awesome thanks. Feel free to keep sharing.

 

[Ben]

Thanks Ben, good to see VSL in this test. What articulations are you using?

Vio1: Solo, Cham, Orch "performance repitition spicc." + Appassionata con sord. "perf. rep. portato" to add breath
Vio2: Solo: "perf. rep. stacc." + Cham, App: "perf. rep. spicc" + Orch: "perf. rep. harsh" -> use different patches from Vio1 and add more color
Via: all "perf. rep. spicc"
Vc: same as Vio1

I also activated the humanizing in Vienna Instruments for timing and tuning. But I did not tweak these settings for such a fast pattern. I do not own the Dimension Strings, but I think they would work perfectly in these scenarios.