L.A. Scoring Strings: An Orchestration Review

Film Music Institute > Film Music Magazine (Archives) > Technology (Archive) > L.A. Scoring Strings: An Orchestration Review

L.A. Scoring Strings is the product of the imagination and skill of Andrew Keresztes, a fellow Berklee graduate. It’s now been in the marketplace for coming up to 14 months. While I don’t know the exact number of units in the field, I have a good idea of how many. Based on that, I can say that LASS has been a global phenom in our field, especially considering that one man conceived it and programmed it. It was a horse of a lot of work, but Andrew Keresztes did it!

The results I’m discussing in this review came from my own out-of-the-box orchestration approach to library testing that doesn’t make for great audio demos. Also, I tested with a number of works that were not PD including my string octaves tests with the opening to Air Force One by Jerry Goldsmith. So I have to apologize, I have no audios to share with you at this time.

L.A. Scoring Strings is currently on version 1.5 with the new auto arranger feature. I have not touched on these features for this review, preferring to stay focused with my first level orchestration approach to evaluating the library based on which orchestral devices as a single library it’s capable of doing. More on this shortly.

When LASS was first released, it was detuned a little more than we all expected. Andrew has consistently tightened this to be close but not exactly perfect. Because of this tuning approach, as you’ll read, LASS avoids several troublesome issues that have plagued multipart vertical string harmony within a single library.

Documentation with LASS is better than I’ve seen with other libraries. It’s well laid out and easy to follow. Those who get LASS from Alexander Publishing, also get our LASS Class which speeds up an already quick learning curve.

There’s lots to LASS, more than I cover in a single review. But again, my focal point is one of orchestration: how many orchestral devices within the LASS string section do I have to work with for my electronic scoring?

In approaching any library with the word Hollywood (like Hollywood Studio Brass and Woodwind Collection, Hollywoodwinds), Los Angeles, or some word dealing with film in its product name (cine is another word as in Cinescore, Cinesamples), two expectations are raised.

The first raised expectation comes from the composer and the second by those who may be contracting the composer to create an entire score electronically and expecting it to sound like a live film orchestra or very close to it.

For the composer, the expectation based on demos, forum comments, et al, is that because of the name, this library can/should sound like a live recorded string ensemble, therefore, when the client asks if a real film string sound can be delivered, the composer can confidently answer, “yes,” thus satisfying both sets of expectations.

Put differently, the standard composer’s question is, “Will this finally be the one library I can go to for a majority of my work and use other libraries to fill in the gaps?”

That’s certainly my question, and one I wanted to answer. But after beta testing several libraries and co-producing one for E-MU, I’ve learned that the best way to find out is to first test the library with the standard string writing devices. Then you know what you really have to work with, and conversely, not work with.

However, there’s another, usually unverbalized, expectation from a composer and that’s the business career recognition that the composer is an artist, and that with the right libraries the composer can write, electronically produce, and potentially release new works for sale via MP3, or by providing MP3s as a prelude for score purchasing or getting commissions or having their works performed live.

However, demos are demos and they don’t always point to what a library is really capable of, or where the weaknesses are. Hence, the only way to find out is to test the library, to put it through its musical paces to find out what it can really do. The Greek word for such testing is peirasmos and it’s the picture of an individual refining gold ore to remove the dross to see how pure the gold is.

This is how I approached L.A. Scoring Strings not just to test it, but to learn it. And, it’s the way I approach any string library.

I started out with two main orchestration questions towards LASS based on its main product features.

1. Could it, as advertised, really create an authentic divisi sound with relative time ease? Divisi is part of the writer’s vocabulary, so I wanted to know how much vocabulary I had to work with. There was also a subset to this question which I’ll cover shortly.

Time ease is a very important consideration because with a combination of libraries that were not marketed as divisi, the divisi sound has been achieved, ably demonstrated in demos for the Vienna Symphonic Library by both Jay Bacal (Fantasia On a Theme by Thomas Tallis and The Rite of Spring) and Andy Blaney (Jeux de vagues from La Mer and Jupiter from The Planets.

2. Could LASS, based on how it was recorded, enable a writer to tap electronically the live strings vocabulary of unisons, octaves, lite harmony, and basic divisi such as Vlns 1 div a 2, etc.

Again, both of these questions revolve around a single orchestration consideration: orchestration vocabulary.

Divisi is the source of sumptuous colors in the strings, open and muted. Yes, divisi can be simply used to add an extra harmony part, which is certainly useful, but its potential rests in creating beautiful ensembles and textures within the larger string ensemble along with both solo and creative doublings with woodwinds and brass to create an even more brilliant tonal palette.

So being able to record with divisi strings is a big deal for the broad number of colors that previously were extremely difficult to achieve with sampled string libraries without hours of work. With LASS, mini-ensembles were organized in an extremely thoughtful manner.

Divisi is also used to divide the section for polyphonic writing which can be seen in works by Bach, Richard Strauss, Richard Wagner, Ralph Vaughan Williams, Claude Debussy, Maurice Ravel, Hector Berlioz, and many others.

Listen carefully, and you’ll hear that most developer divisi demos are along this latter usage of divisi, as LASS’s have been.

But the library that says it can do divisi does bring up a tactical writing/production question: how many vertical harmony parts can you have in a stack without hearing beading or the dreaded organ/accordion sound?

Testing via orchestration devices reveals that answer.

Unison/Octave Combinations
Based on how LASS was recorded, how many of the 65 string combinations I researched for Professional Orchestration 2A: Orchestrating the Melody Within the String Section could LASS do? These 65 combinations aren’t just vocabulary, they’re the core vocabulary for coloristic writing whether for film or some other genre.

So, with LASS, or any string library, how much vocabulary do I have to work with out of the box where the result is a sound like recorded strings as opposed to a big synth sound with indistinguishable string colors?

As I learned from John Williams’ orchestrator, the late Herb Spenser, to be successful in Hollywood, you need to know about a thousand devices (combinations, doubles, etc.).

So to know which string devices and how many you have available, is to know what you can produce with a specific library. The end result, hopefully, is a higher level of coloristic electronic string scoring previously unavailable.

The career end result is having a set of colors available that set you apart from everyone else.

So yes, it’s an artistic issue, but it’s also a competitive one, too.

Other LASS Features
LASS has many great features which I’ll only briefly touch on because my main thrust was in testing divisi and the unison/octave string combinations.

Again, the number of these that work and sound like recorded strings, the greater my vocabulary and the more I can produce with greater ease and speed.

LASS comes with First Chair solo instruments that are recorded in the same space, in the first chair positions. This means that you can write for a soloist, and the soloist will match the ensemble, and for MIDI mockups, you can also layer the solo strings on top of the larger section. You also have the option of writing for either string quartet or string quintet.

Not often mentioned with LASS are the recorded muted strings for sustains, to be supplemented later in 2010 (or early 2011) by a separate recorded muted strings legato library. Though they are sustains, I pushed them to do the opening of Vaughan Williams’ The Lark Ascending and they sounded fantastic.

For each mini-ensemble, LASS has a full complement of legato and staccato bowings, pizzicato, a special repeated notes feature to avoid the machine gun sound that often accompanies repeated sequenced notes, genuine spiccato, and the new Auto Arrange feature. I could easily write another 1000 words just describing those features, but I won’t because my single emphasis was in determining how many string devices (to use Herb Spensers’s term) I had to work with.

Testing Divisi – #1
Since one picture is worth several paragraphs in a review, CLICK HERE and look at the Details Tab and you’ll see two diagrams I created that explain the ensemble approach Andrew Keresztes used to record LASS.

When looking at these graphics, keep in mind that every articulation was recorded with each of these mini-ensembles. I mention this, because as you get into LASS, you see the kind of meticulous detail that Andrew as a single individual, doing this with no corporate support, put into both recording and programming it.

I only had to do one initial divisi test to get the first wave of questions answered. Thanks to my friend David O’Rourke and his contacts at the Nelson Riddle estate, I was able to get a copy of Nelson Riddle’s arrangement of Unforgettable, which at one point has nine (9) vertical harmony parts.

If LASS was going to bead or sound like an organ, it would do it right here. From this one test, I’d essentially know what LASS could or couldn’t do.

So the first thing I need to report, which shouldn’t be too much of a surprise, is that while LASS is recorded on the onstage position so you don’t have to pan it, you do have to set the volume levels for each section to have a balanced sound.

Because they’re less memory intensive, I used the Leg_L (legato) for all the large and mini-ensembles.

Thanks to how LASS was recorded, I could set up 8 Violins 1, 8 Violins 2, 6 Violas, 6 Cellos, and an acoustic pizz bass which LASS has. I could then divide these sections into 4 & 4, 4 & 4, 3 & 3, 3 & 3, and pizz bass.

Right here I need to point out, today, you can only do the above setup with LASS. Another point. This setup, while used for Nat Cole, is also a setup that some kid composer named Wolfgang Amadeus Mozart also used when budgets permitted. For a smaller, more “authentic” Baroque sound, I could even write with 4, 4, 3, 3, and 2. Again, no other single string library on the market today enables this but LASS.

Test Results – No beading, no organ/accordion sound.

Testing Divisi – #2
The one test with Nelson Riddle showed me that LASS could be used for dense vertical harmony so that 6-part or more vertical voicings used by Debussy, Ravel, and others are achievable if I wanted to write it.

But what about other four-part harmony ensembles?

To answer this question I used Ralph Vaughan Williams Norfolk Rhapsody #1 in E minor to test with. This work for full orchestra has a number of exposed lines, div a 3 with Violins 1, smaller ensembles within the larger ensemble, and in one section, what some would define as a big band brass voicing in the strings.

Testing Divisi – #3
The first combination I tested was a four-part harmony section with Violas div a 2 over Cellos div a 2. This is a very standard combination used in the repertoire, by Mancini, Goldsmith, lots! Jerry Goldsmith also used it in Air Force One.

LASS replicated the recorded version quite well without beading or the organ/accordion sound.

Testing Divisi – #4
Next in the same score was 8-part vertical harmony with divided Violins 1, 2, Violas and Cellos. Unlike Nelson Riddle where it was more appropriate to use the smaller ensembles, here it was div a 2 with 8 players for Violins 1a and 8 Players for Violins 1b, etc. The voicing was a four-part triad in the upper register (violins) doubled an octave down with the violas and cellos.

Again, no beading, no organ/accordion. The sound achieved was comparable to the recording.

So whether I used the larger ensemble div a 2 or a smaller one, neither beading nor the organ/accordion effect appeared.

So the questions are:

1. Will Violins 1 + Violins 2 (in LASS in any library) sound like 32 violins or just a bigger sound, like putting two synths together in a layer (unison combination)?

2. Do the unison combinations sound like or comparable to their live recorded counterparts?

Part of what’s behind Question #1 is that all the libraries with strings recorded in their seated positions are left to right Violins 1, Violins 2, Violas, and Cellos. However, there’s also what’s called the European seating plan which is Violins 1, Violas, Cellos, Violins 2.

This second arrangement is really ideal for electronic scoring because for a unison sound for the violin section, you’re getting the same line out of two audio monitors. And that tends to create a lift, and a bigger sound, giving the listener the sense there are more musicians playing.

So while I think LASS did a good job on this combination, I still preferred panning Violins 1 a little left and the Violins 2 a little right.

Here I need to make a scoring comment.

Regardless of the genre, Violins 1 + Violins 2 is one of the standard devices. In the low register it creates this amazingly rich sound. In the high to very high registers, it adds excitement, romance, etc.

That I’m aware of, while several libraries have recorded individual Violins 1 and 2 sections (L.A. Scoring Strings has Violins 2 derived from and programmed from Violins 1), none of them have recorded the sound of Violins 1 + 2 in unison.

So, in my view, this is a hole for all the libraries.

Violins + Violas
You can do Violins 1 + Violas or Violins 2 + Violas. In LASS, either creates a slightly different sound since Violins 1 are a little edgier and Violins 2 are a little softer.

Violas + Cellos
Very impressive, especially in the upper register. Sounded realistic not like a big synth.

The Most Used Octave Combinations
These are Violins 1 – Violins 2, Violins – Violas, and Violins 1 – Violins 2 – Violas. There are more, but these are the simplest and most frequently used, and I tested them with the full ensembles.

All of these worked. No beading, no organ sound, etc. Lower octaves can predominate at times so that requires some editing of velocity or CC11 depending on the line. I also observed that to get the realistic sound, you have to play in each line. Copying and pasting, sliding tracks, etc., still sound stiff to me. Keying in each line individually makes the difference.

Yes, in testing with real scores including one cue, it’s definitely L.A. Scoring Strings and you have a huge flexibility in using this library with the opportunity to recreate literally dozens of the standard scoring techniques for divisi, unison, and octave combinations.

To achieve the non-organ sound and other issues, LASS has been programmed in a detuned way that can be unsettling to some at first, but I’ve noticed that this is mostly in Violins 1. There is also a bit of edginess with Violins 1 you don’t hear with Violins 2. You can deal with this by applying EQ in the 2K to 3K range for Violins 1. I also used the Spectrotone Chart and comparing the EQ between Violins 1 and Violins 2, made additional changes in Violins 1 lower register so that both Violins 1 and 2 sounded the same.

When you make this EQ change with Violins 1, you can get a very good sustained pad sound across Violins 1 – Violins 2 – Violas – Cellos with Basses playing an alternate part or being tacit.

I confirmed this by testing the pad found in Professional Orchestration Volume 1, Flute with Dvorak’s New World Symphony (#9) which I compared to the recorded YouTube version I found with Herbert Von Karajan conducting. Once the EQ is added to Violins 1, you get close to Von Karajan.

One of Jerry Goldsmith’s favorite low string pad voicings used, in order, Root – Fifth – Third with Basses – Cellos – Violas, or Cellos B – Cellos A – Violas (non-div). Either of these are achievable with LASS and are beautiful sounding, regardless of which ensemble you select to use.

Every sampled string library needs reverb, and some ship with the reverb on so that on first hearing out of the box, the library sounds big.

LASS is no exception. It needs reverb, but when you first hear it, it’s dry, so you have to add ‘verb. There’s reverb inside the Kontakt player and LASS also comes with convolution impulses created by Numerical Sound’s Ernest Cholakis which are worth your going through. I use Logic so I had no problem bringing them into Space Designer. But Ernest’s work can be imported into any convolution reverb.

A lot of LASS owners, though I can’t give you a statistically accurate number, use Altiverb. Still others use the Vienna Suite with the Forti/Sorti Impulse Responses Ernest Cholakis did.

A very few have a Lexicon 960 and record their tracks at 192, which is where I understand Shawn Murphy also records (I was told this by a friend who interviewed Mr. Murphy – so this is strictly anecdotal).

Yes, but…there are still other articulations and runs you’ll want in which case Symphobia 1 and 2, and a number of our colleagues are getting a great sound by blending LASS with QLSO and other libraries. Some also like the string runs in VSL to supplement with LASS.

I’m now running LASS on an older G5 dual 2.7GHz with 4GB of RAM and Logic 8.x.

For a new system, I’d look to get (for the PC) an i7 930 with 24GB of RAM, and Western Digital Caviar Black 1TB drives with 60MB cache. With 24GB of RAM, you can load a rather complete string template into RAM and record in one pass. For this you’ll need Kontakt 4.1 which reads all the RAM in the system.

As I’ve written in other places, on the PC, look at the Vienna Ensemble PRO, particularly if you’re using the PC as a farm system. If you’re using the PC for your sequencing machine, both Sonar and Cubase are 64-bit native. This is not an endorsement, just a statement of specs.

On the Mac, if you’re going new, I’d go for a Mac Pro 8-Core, and Logic 9.1.1 since Logic is native 64bit, too. Same rec, 24GB of RAM but you can get it a LOT cheaper from an independent Apple dealer. Because of the name Apple, RAM is over priced when you buy it and they install it.

At this point, this is as far as I’ve gotten. I’ve answered my main questions. What I’m satisfied with is that I can do divisi, that the standard divisi vertical harmony combinations work, that I do NOT get the organ/accordion sound with a dense vertical harmony stack, that depending on the creative situation I have the option of writing with varying sized ensembles.

Here I offer a word of caution regarding the word “chamber” as applied to strings and in particular, LASS. A full-sized orchestra is about 80 musicians and up. A chamber orchestra is about 50 musicians and less.

LASS is organized as a Mahler-sized/large film string orchestra, but you can also create smaller large orchestras.

LASS is setup for 16 Violins 1. That’s nearly as large as the New York Philharmonic string section. You can create a beautiful sounding 12 Violins 1 which is still a pretty standard sized symphony section. 8 Violins 1 is a small symphony section, which you can also create with LASS.

Any of the LASS string ensemble setups you use, balance with a large number of woodwind/brass sections, just as they would with a live ensemble.

So again, there’s lots you can do with LASS for years.

For more information, please go to STUDY HALL where you’ll find a series of LASS class videos I produced.

Knowing what I have to work with, my next action is going through the staccatos and seeing what’s there by type and length. After that I’m looking forward to some serious production work!

Peter Lawrence Alexander is the founding music technology editor of Film Music Magazine and the author of the multi-volume Professional Orchestration Series and How Ravel Orchestrated: Mother Goose Suite. He’s a BMI writer and under contract to write library music.


  • David
    October 22, 2010 @ 11:56 am

    Hi Peter,

    You made the statement, “I also observed that to get the realistic sound, you have to play in each line. Copying and pasting, sliding tracks, etc., still sound stiff to me. Keying in each line individually makes the difference.” I assume you mean that you would play these lines into the computer in real time, as opposed to entering them in step time, is that correct?


  • October 22, 2010 @ 5:41 pm

    By play in, I mean either live or step time by playing the keyboard one pitch at a time. However, some step time the whole thing from a QWERTY keyboard, and others do their work in Sibelius and import it into their sequencer. However you work. I just found that copying and pasting for me didn’t have the sound I wanted.

  • Tomasz
    October 24, 2010 @ 9:25 am

    Welcome Back!!? Peter
    It sure is great to read from you again! YOu have cleared many doubts with your review on LASS.
    Thank You.

    P.S. we miss your valuable opinions on the Tech page.. (here)

  • David
    October 25, 2010 @ 4:31 am

    Thanks for your reply. I do all my work step time in Finale and was trying to imagine how that would have any effect on the sound (compared to copying and pasting) since step time doesn’t retain velocity information like playing it in real time, as with the hyperscribe function. Am I wrong about this or is there something I’m missing? I guess I’m trying to understand how just step time entry can make a difference in this case.

  • October 25, 2010 @ 4:46 am

    Very interesting and thorough review, many thanks!

    If you don’t mind me correcting you on one tiny detail, I believe -actually I’m quite certain, being Greek- that the word is “peiramatismos” (πειραματισμός, experimentation) and not “peirasmos” (πειρασμός, temptation).

    Have a good day,

  • October 25, 2010 @ 2:22 pm

    @Chris – you can correct my Greek tenses. I plan on taking New Testament Greek in 2011.

  • October 25, 2010 @ 2:23 pm

    @Tomasz – thank you!

  • October 25, 2010 @ 2:25 pm

    @David – I use Sibelius where you can adjust the velocity. That said I do all my work in Logic where you can also step time.

  • David
    October 26, 2010 @ 5:28 am

    Thanks for clearing that up, Peter. I purchased LASS several months ago and have not had the opportunity to experiment with it like I want. I also have Sibelius besides Finale, and I have noticed a difference in how certain passages sound different in the two programs. I understand AudioBro has come out with a soundset for Sibelius, which should automate the process greatly. Are you using that soundset in Sibelius, and if so, how is that working for you?


  • Dave
    November 18, 2010 @ 9:11 am

    I’d like to provide an alternative perspective on the PC required.

    You don’t need 24 gigs of RAM, because you don’t need to load the whole set into RAM. All you need is a SSD (example: OCZ Vertex 2 120GB). In fact if you use the 16 bit versions, turn the DFD buffer size down to 6kb, I dare say you’ll clock in at less than 2GB. For the complete divisi section library including the overhead of all the Kontakt instances. Combine with a percussion library of your choice and WIVI, you could have it all running smoothly on an i7 laptop! Probably even run something like Omnisphere as well if you go for 8Gb of RAM.

  • June 11, 2011 @ 8:11 am

    Without even considering the integrity of the samples and learning curve, the truth of the matter is LASS will work in 64 bit mode on a Mac , Play will not. I would like to purchase HS especially with the East West incentive 2 for 1 but I don’t see any change for Play to work in 64bit in the near future. Am I correct in assuming that?


  • June 14, 2011 @ 8:54 am

    You are INcorrect. Play 3.0 iis now in public beta, 64bit, and expected for release soon. Please see my pre- review printed here.

Comments are closed.