Cinematic Studio Strings - Why It's In My Template

With most of you people voting for some occasional sample reviews on this page and only one person really opposing, I decided to start off this irregular series with the quite well known Cinematic Studio Strings.

You can find information about the background of the maker and the library itself on the usual outlets so I'll save you from that and get straight to the point.

Cinematic Studio Strings (CSS) has over the last year advanced to my main string library. For consistency reasons, I was using Berlin Strings for a long time as the my main string library but never was fully happy with it. It does provide an incredible depth of sampling that CSS cannot match but there was always a certain unemotionality to the legato patches of Berlin Strings. 

The Pros:

CSS has an incredibly consistent programming. It feels like a lot of testing and re-evaluation has gone into the library to iron out as many quirks as possible. There are very few programming faults or sloppy editing issues to be found in this library. So consequentially, you don't need to invest a lot of time working with it to iron out any weird behaviour which is a big plus when you need to be productive

The legato is one of a kind. The library comes with three types of legato transitions which range from relatively short for busier passages to very long for slow legato passages. All these introduce a considerable delay into the legato patch as the actual target note of a transition is reached sometimes 300ms after triggering it. This of course makes it always feel late so you need to trigger new notes earlier. Introducing a general negative track delay will not work as depending on context the lengths of the transitions differ. Many people have complained about this delay and requested the developer to make the library more "playable" but this delay and the time it takes to move from note to note is one of the features that add a lot of realism to the performance.

The library comes with a very clean interface, and only one patch per instrument. Triggering different articulations is done through key switches, CC1 and Velocity and a few other CCs but you can access every articulation in that one patch. I can see the benefits of this approach as it declutters a lot of things but also have my problems with it (see below).

The tone of the library is rather on the dark side but the sustains exude life and passion in the playing which is essential on expressive legato passages. The vibrato on the notes is quite on the strong side, there is a possibility to crossfade to a non-vibrato sustain. As always the "limbo" between such crossfades is as with all libraries that offer it a bit "nondescript" as it of course has nothing to do with how real players would transition from heavy vibrato to no vibrato. But overall the expression and tone of the library is very lively.

The price point. For what you get with this library the price is extremely generous. At just 399 or even lower if you have other libraries by the developer already, you get one of the best sounding string libraries on the market.

In many regards, I consider the performance of CSS as the best that is commercially available. It is one of the few string libraries that I had under my fingers where I really feel like a performance is happening and not just a sequence of triggered samples.

The Cons:

In spite of my praise for this library and the fact that it builds the fundament of my strings there are a few things that I struggle with.

There is a considerable noise floor on the samples especially in the low dynamics. As the library itself comes with a rather dark tone, I usually do quite a bit to push the air in the samples, meaning that I boost the high frequencies of it quite a bit through EQing. This of course also unfortunately boosts the hiss noise in the low dynamic samples. You can counteract this by using denoising, however, it is quite tricky to capture a clean noise profile to use with the denoiser. I use the release tails of the pizzicato notes to capture a usable noise profile. However, denoising too much again reduces the "air" in the sound so it is always a blancing between hiss and air. This problem is non existant in louder or more busy passages but particularly on soft exposed passages it can become quite annoying.

The interface also has a few downsides. While I understand the reasoning of having everything in one patch, I wish there were alternative patches that provide the articulations separately. This one patch approach makes working with it sometimes very cumbersome. There are 4 different types of short notes that differ in length. They can be triggered by using the appropriate keyswitch for shorts, then the different lengths are spread out over the range of the mod wheel with the shortest short at the lowest cc1 values and the longest short at the highest cc1 value. To make things worse, you need to follow a specific trigger order. Placing your modwheel at the right value and then triggering the keyswitch will not give the result you want but you end up with the longest short. So if you are coming from a high dynamic sustain (high modwheel values) and want to switch to a very short note after this, you need to make sure to first trigger the key switch for the shorts and then move the mod wheel all the way down to the lowest values. This gets incredibly annoying if you programm a passage that does a lot of quick articulation switching between shorts and long. 

Theoretically you could use several instances of the patch and only activate one articulation in each of them and purge the not used articulations (which is possible to save on RAM) but then again, this is not the most user friendly way either as still different controllers are needed for the different articulations. Another alternative is to use Midi CC58 which depending on its value toggles through every articulation available. This however requires setting up an articulation management system as programming these switches by hand is cumbersome as well. So after all, I personally feel that the articulation swtiching is not ideal.

For my personal way of writing music, the depth of sampling is not deep enough as it only provides the fundamental articulations. If you are looking for more rare articulations (e.g. sul ponticello which I wouldn't consider to be rare) you need to bring in an alternative library.

A big weakspot at the time of writing is the execution of runs with the library. I feel that particularly fast legato transitions are really not convicing. You can use "ancient" mockup techniques like using trills layered with staccati but there are a few libraries on the market which have dedicated patches for this that work reasonably well. There is an update announced to cover that but I don't know whether this will include newly sampled content or just programming improvements.

The library comes with a few articulations that use layering of samples, for instance the marcato patch has the option to layer in a staccato sample which depending on your velocity simulates a hard attack on the patches. While this works reasonably well in less exposed passages, in my opinion this does not replace a dedicated accented attack sustain. 

A few more things that need to be said about that library:

It has been recorded in a relatively dry space so it is usually needed to give it some reverb to get it to a symphonic feeling, so chosing the right reverb that makes it sound cinematic is more important than with libraries that are wet from the beginning.

However, it comes with 3 different microphones (Close, Tree and Ambient). I usually use a mix between the close and ambient microphones which in my opinion gives a nicer tone than the tree mics. Also, the hiss noise is less prominent on close/ambient than on the tree.

The sections that have been recorded are relatively small, so you should not expect a huge string sound from the library. However, with many "large section" libraries sounding more like a synth patch than strings, I appreciate the small line-up as it provides more detail. Additionally, due to the fact that it crossfades between different dynamic layers and you hardly ever only hear one layer alone, this "small size" as with all libraries is relative.

Due to the darkness of the sound which I usually like to be quite a bit brighter, I use a quite elaborate effect chain on CSS that involves a few EQs and exciter.

Here's the chain that I usually use: 

This is the most elaborate effects chain I have on any of my virtual instruments but I really feel that it is necessary to tame some of the lower mids of the library and bring out some of the high frequency brightness.

Why It's In My Template:

All that being said, CSS is currently the main string library in my template. I was using Berlin Strings as my main library for quite a while but switched it out for CSS due to the superior expressiveness of the library. Especially the legato playing (which is a big part of string playing in general) in my opinion is unsurpassed and the consistency in the recording and programming really makes this library very hassle free. The delay doesn't bother me at all as I hardly ever play in any lines but import MIDI data that I then adjust further in the DAW. I have built myself a quite effective way to quickly program string lines to suite the needs of CSS.

The library especially on expressive writing provides an emotional depth that I didn't find in any other string library yet. This quality alone makes up for the few problems that I have with it and also justifies the additional amount of work that needs to go in to get the timbre to my liking.