Question about Libraries

geppino12575 wrote:

I wanted to ask something: in this section dedicated to libraries, can anyone post a personal review of a specific library by clearly indicating the brand, or should one stick to generalities?

In my opinion, it is better if Simone/Paolo responds.

Obviously, I share the following reflection:

geppino12575 wrote:

I think that direct comparison between users who use sample libraries could be very useful for those intending to make a purchase, much more so than demos which, being nothing more or less than advertisements, tend to maximize the merits and minimize the defects of every library.

Since this is a discussion forum, as such, it collects the opinions of members based on their own experience. You can quote the name and even the surname of the library without any problem, and if a library has made you regret purchasing it, you can explain your reasons calmly. It is assumed, however, that reviews are written by capable people who have worked extensively with libraries and who know what they are talking about, attempting to write experiences that are not "biased" but still contain objective data. Just to say that an inexperienced person might write a review presenting negative data that actually does not exist or that only appears because their hardware is not powerful enough to allow the perfect functioning of that library. I mean, if we are to analyze the cons, we must ensure they are real and objective and not dictated by our own shortcomings. Another example could be qualifying the sound of samples as "too dark" after hearing them through inadequate monitoring. In this case, the dark sound might not be attributable to the library itself but to a problem with our listening setup.

It is also possible to attach demonstrations in which the library performs well or poorly, which can help support some of the reviewer's opinions.

I say this because, as you well remember, there are libraries that sound divine in some demos, but when we actually go to use them, they show several little flaws, some even serious ones that would induce a buyer not to purchase. I believe that web opinion is a very important thing for the buyer, who can have a point of view that is not that of a commercial or the showcase imposed by the manufacturer.

Recently, I purchased a sampled library with some beautiful sounds, but it presents many problems. I am in direct contact with the engineers who designed it, and if they resolve the issues, they will have created a truly cutting-edge product in terms of libraries; otherwise, it will be a flop, and I will also bring this up on this forum by attaching the objective problems of this library during use.

I would like to give an example regarding the VSL demos (having three libraries available that I have tested and am not using these months): the demos they published on their website are, in my opinion, quite far from the real potential these sounds possess; many of the demos were made without much attention to dynamics, and in some cases, the compositions do not live up to the individual samples (theoretically, a separate discussion could arise here). I have the impression that (and I am not talking about sound quality, but rather the use made of it, more precisely the modifications to be applied to a single articulation so that it results as realistic as possible) many people "settle" for factory sounds without tweaking them, something I feel I can also say for some East West demos—always pompous, with absurd reverbs, pieces with the symphonic Choir performed in a rush (as they say, at high speed), but without caring for the parts, so that the final result does not perform at its best. So far I have owned three different orchestral libraries: East West (Gold edition, Choir, which I no longer have), VSL (Chamber strings I standard and Appassionata Strings, which I still don't use, Oboe d'amore), one from 8Dio (oboe d'amore, which I still don't use), and two from Embertone Friedlander (solo violin and solo cello, the latter with only a single articulation, it is a demo).

Having tried East West for months (in 2005/2007), as many will have discovered for themselves, there are significant flaws regarding certain instruments, foremost among them the horns, with sounds sampled with an evident break at the cut-off point and loop start; some are truly indecent, and one has to find (or had to find, in my case) compromises to use them in pieces.

The East West strings have an attack that is too slow, not to mention the legato, which can be disregarded, but with some adjustments they could be improved, though only slightly (furthermore, I never used the players to compose pieces, so mine is a separate discussion that I will clarify shortly).

VSL undoubtedly has the (let's say) best sounds, but to obtain proper results one still has to work on them; taken as they are, the sounds are too dry, precisely because of how the individual samples were acquired, so unlike East West, it is "sufficient" to add a light reverb and certain situations might benefit, but this certainly does not replace (as Simone already explained in the post about the work engineers must face when working with samples) the natural reverb obtainable in a specific space with a real instrument (the mode in which the East West sounds were sampled).

A few months ago, VSL introduced a new library: this latest one allows you to position the instruments of a section (whether strings, brass, or woodwinds—it seems to me they have yet to release the latter) as you wish, also choosing the number of instruments to use; we have a similar function with the Embertone solo violin library, but the latter differs from VSL in that it does not allow you to position the violins (8 to be precise) wherever you want, though it still allows you to decide how many to use. It should be noted that the interface presents the 8 violins positioned in an arc, from left to right, so depending on which ones are chosen, you will have a "sound" that is more or less open—forgive me, that term is not quite appropriate.

So far, 8Dio is the library that has "impressed" me most regarding legato; besides VSL, I have not yet heard a string and woodwind library that allows for such fluid passages.

It is true that Embertone has produced excellent virtual instruments; the solo violin I bought months ago (on sale, only 70 euros) is excellent; it even has the simulation of the bow hair on the string and, in addition to this, allows for the simulation of the typical timbral change depending on whether the bow is passed near the bridge or not—a truly good solution.

I will end here this brief exposition of my impressions, which are based primarily on the use of adjustments in order to make passages realistic, and I am not basing them on pure sound quality; in this case, I am speaking specifically about the solo violin and the violin section (not the complete strings), which taken (anyway) individually from East West, VSL, 8Dio, Embertone, in my opinion sees VSL as the leader in terms of quality (on which one must work anyway to obtain very defined passages), followed by Embertone, then 8Dio (but only for one reason: here the solo violin is excellent, but the library in question seems to be accompanied by 4000 ready-made musical phrases, and I don't know exactly how many articulations are available and, if I haven't understood correctly, I haven't looked into it much, whether the solo violin is"}

present in the library of the violin section they sell), anyway between Embertone and 8Dio they are neck and neck, finally there is East West which sells (orchestral library aside) a solo violin sound package that is truly pathetic.

Since Piccinesco has entered the discussion quite forcefully (just kidding...), I feel like giving my two cents on the various libraries, sticking to strings for the moment.

I have had the opportunity to try and use the following libraries: Garritan Personal Orchestra 4 (GPO4), East West Quantum Leap Symphonic Orchestra (EW), on the Kontakt engine, not the subsequent version on the Play engine, Cinestrings, 8Dio Adagio Strings, Vienna Chamber Strings (solo strings), Vienna Horizon Opus 1, LASS 2.0, Berlin Strings, Embertone Friedlander Violin and Blakus Cello (among other things, I am in contact with the person who developed it), Symphobia 1, 2 and 3 (Lumina), Cinematic Strings (not to be confused with Cinestrings).

Unfortunately, I am missing the East West Hollywood series, which I haven't had the chance to try and therefore cannot comment on.

Given that expressing an opinion on all these libraries would be very long and perhaps boring, I will try to be concise by highlighting pros and cons based on my personal taste and some objective elements, so take what I say with a grain of salt.

Currently, among all these, my choice falls on LASS 2.0 regarding sections, while for solo strings the Vienna Chamber strings are certainly remarkable even if they are a bit dated, but 8Dio is also a good choice for solo strings.

LASS has the advantage of having a very fluid legato, which works well even for relatively fast passages; furthermore, the dynamic control with the mod wheel is very good and, if desired, can be edited as you wish. There is the possibility to create divisi in a realistic way by loading three different samples for each single articulation (e.g., leg A (4 players), leg B (4 players), leg C (8 players) total: section of 16 first violins) and if a three-note chord is played, the software automatically distributes one of the three notes to each sample; the final result is very good. The three samples are slightly different in terms of intonation, color, and response to controllers, which greatly benefits realism. According to some, the LASS sound is a bit "rough," "coarse," not soft and round like the EW, but for me it is more realistic. LASS is dry, so you need to apply good reverb to it. The samples are already positioned stereophonically, but that doesn't mean these settings cannot be modified. There are also FULL MIX samples, which are already a mix of the three individual samples, but, since space is available, it is preferable to use the three separate samples for each articulation.

LASS solo strings are not quite as good.

Lately, Berlin Strings has become very fashionable, the latest born as a string library. It's excellent, but I'm not 100% convinced. It has a unique "spiccato" sample that is truly remarkable, but the legato is a bit artificial, with a sound tail cut that is quite unnatural in the transition from one note to another. Furthermore, the legato sample alone weighs 4 GB, which makes it usable only with very powerful software. It has many articulations, perhaps too many; some seem like repetitions or unusable to me. Berlin Strings is recorded in the Teldex hall in Berlin, so it is already quite wet.

8Dio Adagio is based on a different concept: they focused heavily on various types of legato to make the library "emotional" rather than realistic. In the version I have, some samples have evident sampling errors, sonic artifacts, and strange noises that make the samples unusable (surely in subsequent updates they will have resolved this problem which was very obvious); moreover, the bad thing is that it's the user who has to adapt to what the library does best; if you try to do something else, the result is not good. If you listen to the demos, they will probably sound like the best ones out there, but actually playing a phrase with those samples is not easy. LASS might be less "emotionally engaging," but you can play whatever phrase you want with it; then it's up to you to make it musical.

Strangely, in 8Dio, solo strings are better than sections; they are much more playable and the timbre is not bad at all.

In my judgment, and not mine alone, the top in terms of realism is achieved with Vienna. Given that the Vienna libraries I have are a bit dated, but they have a very particular charm, a "seriousness" that allows you to feel safe; you don't have to worry that a certain phrase with that library might not work—if you use Vienna and a phrase doesn't work, it's because the phrase isn't good, because there will certainly be an articulation that allows you to bring out the phrase you have in mind. Vienna is a guarantee in terms of realism, even if lately the trend is that one can do without realism in virtue of the "epicness" of the piece. Lately, demos must have 24 (!) horns playing in unison, choirs doing things that no choir could do in reality, etc... well, Vienna shuns all this; Vienna is

for the sake of realism, and I like this thing very much, because it is useless to develop credible libraries if then one realizes things that are "incredible" (in the negative sense of the term).

For the moment I will stop here so as not to weigh down the post too much; I promise myself to complete it in a subsequent post talking about Cinematic strings (quite good), Cinestrings (doesn't convince me much), Embertone (I advise against it for various reasons), EW (honest but now outdated), Garritan (a toy compared to the others, good for Finale and nothing else)......

I see there are many opinions on different libraries. I would like to read them in dedicated discussions of the series: "LASS: Pros and cons" and so on for the others. If we want to talk about libraries by saying what we like and what we don't like, we should also attach some examples in various contexts. For example: "I don't like EW in legato for this reason" and here you could attach sample audio files that show the flaw you have encountered.

In response to Maurizio and partly also to Giuseppe.

Vienna is a library used for professional purposes. It costs a fortune (at least triple any other library) and promises excellent results. If put in the hands of someone who knows how to use it, I would find it very difficult to say whether what is playing is a real orchestra or Vienna. For a novice or simple music lover, it is a real orchestra, there is no doubt about that. Other libraries offer more or less beautiful sounds but they are very 'exemplified,' meaning they are usable across a wide spectrum. All this is simply based on a market choice. Vienna produces for a few experts and professionals in the sector, charging a lot (the user base is also smaller); others produce for composers and music lovers, including self-taught ones who want to produce a credible audio track without getting too deep into the engineering topics underlying mixing and mastering, paying less (but the user base is larger). Many libraries are already "ready"; just a few adjustments and the sound is already decent. Vienna must be prepared in every aspect, and this requires knowing in great depth not only the basic rudiments of musical instruments but also studio recording and mixing techniques. One must have the necessary tools to prepare a virtual environment with reverb machinery and know how to perform various calculations to make the sound added by these environments as credible as possible. This explains why I used to argue so much with Giuseppe during those Skype lessons, trying to make him understand that using 'packaged' sounds is not the same as starting from a dry mono sound to position it at our will in space and create ourselves the sound that instrument must have in an orchestra; but then, as you see, experimenting and gaining experience clears up any initial doubt. Recording a dry MIDI track created via Vienna is equivalent to receiving a mono stem (or stereo depending on the instrument) from a recording studio (there are obviously some unsolvable issues dictated by the fact that we are receiving the sum of samples and not the whole, as if to say that the whole is not equal to the sum of its parts). We will have to, at a later stage, recreate the stereo field and all the typical environmental effects, i.e., perform the tasks normally carried out within a recording studio when an orchestra is recorded, but the final result will be very credible.

Practical example: a clarinet in an orchestra cannot be stereo, yet in many libraries the samples are stereo. In reality, no one would ever dream of recording a clarinet in an orchestra with two microphones, and having to record it with a single microphone, the recorded sound can only be monophonic. In a second phase, during Mixing, it will be the sound engineer who recreates the stereo field by positioning it in the space where it belongs. This is why I say that starting from a ready-made package is not good. We must be free to create our own stereophonic space and, in the end, the mix will benefit in terms of spatial width. The mix of an orchestra must always have a very wide stereophony; just think of the image of a full orchestra occupying the entire width of the stage. When I listen to these pre-packaged mixes, I always hear everything excessively in the center -> ergo -> the mix is wrong!