Discussion:Sequencing Techniques For Strings,Strings Samples

[zenguan88]

well i feel that such a thread should be created so that each of us could trade tips to get that realism in strings sampling.

first of all this thread only discusses about software plug-ins or in any case hardware strings or string pads. be it the famous jupiter strings pads to the sound library of EWSQL etc. (no recording, only samples) : )

i shall post a few question first, as i get to know from someone ( he might be wrong ) but what i got news of is that in EWSQL Platinum ,the vibrato articulations could not be edited, is there a way to go about doing that say by mod wheel or something? that is because i feel the vibrato is an important element in creating realistic strings sounds, what i mean of editing the vibrato is the rate at which the 'player' uses the technique.

also another of my question is those that had tried samples library before such as EWSQL all the way to VSL, GPO, etc etc would you guys be able to list the stuffs that you like about the library and what are the features you find its useful and what is troublesome this way.

and of course feel free to add in more question so that we each could discuss them, hopefully this thread will be successful and the brass and woodwinds can be followed up : )

 

[Cheez]

Hi Zenguan,

I think the question you are asking is not exactly about sampling techniques for strings. And furthermore, techniques for sequencing for strings has been discussed before in this thread:

http://soft.com.sg/web/modules.php?name=Forums&file=viewtopic&t=2550&highlight=strings

We've also discussed about winds and drums techniques.

Want to edit your title? Perhaps this should be discussion mainly on samples? For techniques, it will be easier to add to the other thread so people who wants to know about techniques can read all the posts in one nice place.

 

[teraslasch]

lol zenguan, sry i din tell u, but there are many different kinds of articulations patches, as well as the mod wheel patches. mod wheel patches allows you to control wad is stated there that will be controlled by the mod wheel. There are keyswitches patches, that let u switch patches with a very low low note.
but basically, if the mod wheel patch that allows me to control vibrato doesn't sound nice , i wouldn't use it. In ewqlso, you can't really change the individual articulations (except if ure in a modwheel patch or keyswitch patch), so many people use layering.

 

[Cheez]

There are 2 elements in vibrato. One is the speed (rate) and the other is the depth. For synths, LFOs are used to control these 2 parameters and it is relatively easy to program.

However, using LFOs in sampling creates unrealistic effects. You need to know how samples are created. They record a string section for however long they want the note to hold. The depth/rate of vibrato is therefore fixed. To change the depth or rate, sample developers will have to record different instances of vibrato at different depths and rates (for each note) and somehow crossfade them smoothly. This means lots of sample recording, creating huge libraries taking up lots of memory.

Having said that, changing vibrato rate and depth in large string sections are usually not called for. In large sections, the changes in rate and depth are less noticeable and seldom written for. However, it is useful in solo instruments, which will add realism to the seqeunce.

So....although EWQLSO, VSL and other large libraries cannot do that, Garritan's Stradivari violin and Gofriller cello can do it!!!! This is because Garritan used a different technology one level up - some kind of morphing technique. And it makes perfect sense, since the Stradivari and Gofriller are SOLO instruments! For large sections, not necessary.

 

[zenguan88]

i couldnt see the sample engine of the EWSQL therefore i dont really understand how it would had works and fare, but from what i know VSL seems pretty cool in activating different types of articulation quickly and efficiently. (demo in their website)

hmms alright i understand about the vibrato now, apperently the demo i heard was suppose to be a solo cello and a solo violin i guess, correct me if i am wrong tera. so this question came upon me on controlling the rate of the vibrato. hmmms okay i got the vibrato problem address . i do agree most of the time you will only want it on the solo strings.

i have another issue though which is, are we able to call for repetive upbow playing or downbow playing for a single phrase? and what are the factors we have to take note of for sequencing a ensemble of strings. do you guys prefer to have like say a 4 section strings and you sequence each section one by one, meaning violin viola cello double bass a track by a track ( multitrack ) or just a single track? for me i do sequence them by multi track as i can go back to editing it later, and it gives me a cleaner and clearer workflow to work upon.

 

[Cheez]

Up and downbows are nowadays automated by the softsampler engines. If they have that programming, you don't have to blink an eye or worry about choosing up or down bows. When you say "repetitive up and downbows", there are 2 different situations. One is alternate bowing of a phrase, which as I said, in many cases is automated. The other situation is repetition of one note (but not fast and long enough to be called tremelo - which is usually a different set of articulation in itself). This is where the "machine-gun" effect can come in when one note keeps repeating itself. Nowadays, most samplers does round-robin, which basically means that the same note is sampled many times, and the sampler randomly trigger different sample of the same note when you play repetition notes. That way, you'll not get the same sample being triggered again and again (ie "machine-gun effect") but rather each note is slightly different from the other.

The key to realism - always sequence 4 sections separately. In fact, I sequence 5 sections separately - Violins 1, Violins 2, Violas, Cellos, Doublebass. My library (and most nowadays) has separate violins 1 and 2 sections sampled so that it actually sound different. Also, because of different articulations, it's normal to have multiple tracks for one instrument. For example - I have no less than 5-6 tracks just for violins 1, 5-6 for violins etc. I can easily have 20-30 tracks just for strings alone.

But we are getting into techniques again - which zenguan, should really be on the other thread. But no matter. The other thread is old. Forget what I said - carry on with the discussion...

 

[zenguan88]

cheez hmms what i meant by up and down bow is like for example the phrase i am playing is C G E F D , but during this phrase i do not want a legato neither do i want a downbow effect, i want to create a staccato effects thats all up bow, as this is an example lets say that C G E F D are all crotchet notes. so as the the player moves up the bow, the bow become shorter and shorter making it harder for the player to excute the note, is it possible to trigger such a performance.

the example that i had given is an overkill, usually i saw such stuff happening in an introduction to a phrase etc, and usually it only lasted 2 notes having an up bow ( down bow i seen little of that happening ).

So another example would be the notes C D E , and i want my player to play UP UP DOWN. from what i know the up and downbows are automated in a way that the first note you play would be up (or down) and the next note would be the reverse of the note you just played.

sorry for creating a new thread do let me know if i could continue this in the previous thread that had been created : )

 

[Cheez]

Ah ha. You are hitting the limitations of current samples! I know what you mean and it's something I've always wanted to do as well. It's not impossible - but we need to wait for sample developers to come up with that. So far, the legato samples out there are mostly downbow. If they start to develop upbow legato samples, then what you said is possible. We will only have to play upbow legato patch, then key-switch to downbow when we want the downbow. The switch will give you the initial "attack" of the downbow. And if the patch is subsequently downbow legato, you get the continual legato downbow. Done!

But for now, we can only use legato and non-legato samples to "simulate" this effect, which is what I usually do. The samples are either key-switched or cross-faded with mod wheel for easy sequencing.

But what you said about "staccato up-bow" is easily done by just using upbow samples. Up and down bow can be keyswitched easily so you can switch between up and down bow whenever you want. It doesn't have to be automated at all. The difficult part is what I said above.

Don't worry about going back to the other thread. It's too buried in the past that it's getting difficult for people to search anyway. We can continue on this thread.

 

[zenguan88]

yea hahas!

okays right now i am looking for a suitable string lib preferbably with their own sampler, a particularly wide range of sounds and variety and also a good set of articulations to play along with.

i am not really into looking for very classical sounding kinda strings, but what you hear on the radio on most days.

cheez you mention about kirk hunter, since you use it right? could you give me some idea on how is it working for you and what about that you like to that library, how does it compare to the rest?

 

[Cheez]

KHSO is a low cost but VERY high quality sample set. It is in the league of EWQLSO (some finds the strings of EWQLSO better than EWALSO).

It used to come in 2 forms - the smaller Sapphire and the larger Emerald. It appears that Sapphire is discontinued. They are producing the new version - Ruby (now called "Virtuoso Pro" series). The Virtusuo Pro is going to be expensive - almost around the cost of VSL. It's going to be modular and the strings alone will cost about 500-600 USD depending whether you are upgrading or not. My curnent entire set cost only 300+USD.

The strings are the strength of KHSO who has been producing string libraries since the early days of sampling. This is Kirk's strength. The strings give the warmth you need in many situations. The solo strings are incredibly realistic that beats many other larger libraries (just listen to their demos - or mine once I get it done....). Kirk designed it for applicability in mind. So you get amazing articulations and control that covers almost any type of playing/writing. Eg: Soft sustained strings (mod wheel ->) increasing louder and crossfading into tremelo; hard attack marcato (mod wheel ->) legato; marcato (mod wheel ->) spiccato; mod wheel control the hall reverb etc. Too many articulations to list down. The control makes sequencing almost any kind of arrangements possible and therefore more realistic. The crossfade is also very smooth. EWQLSO users have complained that one of the biggest problem of EWQLSO is crossfading which is not smooth (hopefully to be fixed in their update). Smooth transition is very important for reaslism. And best of it - they have first and second violins sets. So if you have violins 1 and 2 playing unison, you get different sounds making the layering thicker.

My only problem - Kirk decided not to develop Giga anymore. This is annoying to me and made me feel I got into a bad investment of a Giga library. He will continue to develop for Kontakt and EXS, but he did hinted on another forum that he's going to come up with his own sampler like the rest. The other problem with the strings in Giga is that the legato is not smooth enough. It gives me all the articulations I'll ever need, but this part fall slightly short. I'm thinking of covering that part with VSL (probably Appassionata strings). VSL has the best legato samples that nobody can beat. Anyway, I'm really considering moving away from Giga seeing that Tascam made a mess out of it.

KHSO is not standalone - yet. For standalone strings, you probably have 2 choices - EW or VSL. There are strengths and weaknesses in both libraries.

EW:
Pro - recorded in a hall and reverb is natural (don't need any reverb or EQ), strings are warm
Cons - programming not so good, reverb tail makes crossfading problems sometimes, legato not realistic enough

VSL:
Pro - legato very realistic, powerful standalone player
Cons - recorded in a very dry room and therefore strings may sound slightly harsh (so need good reverb and EQ).

People will argue which is better and it will always be split. My personal preference is VSL because I feel that no library can create that kind of realism I'm hearing from them. But there are those who feel the dry samples are too dry.

 

[zenguan88]

thanks for the infomative reply.

for me i feel VSL dry strings could be in a PRO instead of a CON situation this is because, most people might want to generate their own feel of reverb. This makes VSL patches more versitile.

oh oh oh yep another thing to add i love the chamber strings in VSL more : )

maybe the only downside of VSL is its going to be costly.

and yes i do hope kirk would be able to come out with their own sampler! : )

 

[teraslasch]

hehe.. i still find my ewqlso to be bettter than the other libraries out there, maybe its just personal preference. Judge yourself by listening to the demos, then decide .

 

[Cheez]

Yes, tera. It all comes to personal preferences. Some who will swear by EWQLSO, some will swear by VSL. People who like one will not like the other.

The truth is, there is never a perfect library, just like there will never be a perfect keyboard. To get the best effect, it's always mix and match between libraries that are complementary. At the moment, I mix and match 2 orchestral libraries - hopefully it will be 3 soon.

The problem with the VSL site is that they have too many demos. Some are done not quite well, some are done very well. To search for the good ones, you need time. Soundsonline pick only the best demos - so they have few demos which unfortunately may not be representative enough.

Example - one of my favorite young composers who wrote this with VSL some years ago - http://vsl.co.at/Player2.aspx?Lang=13&DemoId=4753

You can hear that it's hard for other libraries to beat the realism of the legato he used extensively.

 

[zenguan88]

had been looking for a strings software for now. well i decided to go to the VSL way, as in this way i can have the option of adding my own reverb. therefore control more parameters.

well my options now are these 3 from VSL, its either the chamber, the appasicato, or the SOLO Strings.

well my favourtie would still be the chamber as it gives more articualutions. i am jealous for the appasicato lush sounds, but right now i can make do without it.

however i got a pressing question about the solo strings, do you think its feasible to buy the solo strings? as in its like i can then control how many violins can be put together or viola put together. yes i do know the tone of the sounds is different, for eg, recording a 8 strings together is different compare to recording 1 strings at a time. but i am curious if this method works.

if it does then i might consider getting the solo strings, as then i can control my string sections, have more articulations to play with

do let me know your thoughts!

 

[Cheez]

Do note that with chamber strings, you will still have difficulty creating large string sections.

As to your questions on solo strings, you cannot do that unless the library has a few different solo strings. They usually don't. Adding the same solo strings over and over again just gives you the same sound. The same goes for chamber strings. Libraries like KHSO has violins 1 and 2. That makes things easier - you can make them unison and thereby creating thicker sound.

If you want solo strings, my suggestion is to consider Garritan's stradivarius (violin) and gofriller (cello). These are simply the most playable strings.

Another technique sometimes used is to layer solo strings with string sections.

Lastly, reverb may not be good enough nowadays. You probably want to consider convolution impulses. If you're using Mac, the best is Altiverb. Otherwise GS3 has Gigaimpulse.

 

[zenguan88]

yep i got one convulation reverb here in logic call the space designer( i think its a new plug in for LP7.2 ) i plan to get altiverb.

alright i shall state the way i am working, for strings sections , i usually use samples in my fantom first to give like a rough guide to whats going on, after i had done the song, or most parts of it, or when i feel its enough i will bounce the track(without the strings) and then i would create a whole new song just for the strings and then bounce the strings out, therefore saving CPU consumption.

also i do know altiverb is taxing therefore the need to do by this method, however i would only add one altiverb instances to the string mix. thats bcos in real life, strings are usually recorded in only one place, therefore theres no need to call up more than one reverb instances.

alright i get the meaning of the solo strings being impossible to create to a large string sections, hmmms what i thought was that there will be different preset of the 1st violins in the solo strings therefore allowing me to edit the sound a little and yet still pass it off when i combine them together. seems like its a little impossible.

hmmm i do not need such a large section but more of nice sounding articulations and dynamics!

i guess chamber string is the way to go then

 

[Cheez]

Chamber or large strings will depend on the music you write. But eventually, you'll need the entire template - solo, chamber and large sections. And you'll find out that sometimes it's better to mix strings from different libraries. Each library has it's own strengths and weaknesses.

Since I already have 2 libraries with solo, chamber and large sections, I feel my next call will be Appasionata.

 

[zenguan88]

oh yes yes yes hahas.

hmms i now only had the budget for one strings section so got to take in mind which i would need more

AHHA thanks for your advice anyway