LA Scoring Strings vs Hollywood Strings

[aka21stCentury]

Developers are probably tight lipped about tuning for the following reasons.

Kontack, because of the alternate tuning possibilities. One can duplicate samples, rename and detune them, then layer the sets on top of each other.

One big problem discovered early on in sampling: problems with air resonance. In each instrument this air resonance is a different frequency as we all know. Fairlight (Harvey Holme's grad students) would sample the strings (via piezos) and then compare the result to a measurement mic. These differences were what Harvey called a transfer function. Obviously if your resonance is near 3rd octave C then if you are playng 2nd Oct C or G you'll get more energy from the air column. This is why extremely sophisticated computers are required to accurately play back samples.

In the early days we had the Diamond String synth, basically just two tuning controls, master and fine tuner. You can hear this used on Bowies 'Heroes' album played by Eno on the flip side. The beating is just one reason you had to have fairly good ears and been around orchestras to know what a string section sounds like. These were oscillators (banks) and not samples. However, I've been around sampling from early on, and perhaps, some think that tricks (some of which are described below) are proprietary. Most of it is known to those in the biz, Fairlight or NED & Emu Systems.

Also if you consider one violin sample duplicating both 1st and 2nd parts in an assemble, this will not sound similar to a true orchestra section. The fundamental air resonance is unique for each violin, yet the notes themselves will be slightly out of tune from the other players, their air resonances again will be slightly different, so re-tuning a violin and air resonance together is a pointless task in attempting to recreating an accurate string section.

What gives a piano its signature sound is this beating, similar to that of a string section.

Also digital tends to be inaccurate in its analyzing of sine waves with saw tooth waves included as part of the waveform (bowed violins and violas) despite the Fourier math 'describing' sine waves as true curves, and not an etch-a-sketch, but in fact they are as waveforms come in and out of digital as square waves, but often analogue filters are used on the D-A part which will soften the harshness produced by an over abundance of square waves.

It is not rocket science this part. What is a bit more complicated is the math in summing or attempting to put humpty dumpty back together again.

Patent examiners apparently do not understand this either and have been assigning patents for years with a similar misunderstanding of the true nature of digital recording (the etch-a-sketch model). A good set of samples would use a variety of violins and allow the air resonance to remain tuned when detuning individual notes as a matter of realism.

So what really comes in via A-D are square waves that must be correctly matched to a sine wave or sawtooth wave, even a square wave at times with bit mapping is essential a data base or look up table to improve the chances that 'the guess' at the true waveform is closer to its actual waveform. Just one reason Dave Rossum slaved away for over ten years working on artificial intelligence to improve sampling.

Sampling, and the ability to know in advance the transfer function of each instrument, has the potential of improved accuracy over digitally recording the same instrument, particularly a violin or snare (both with sine and sawtooth waves in their waveforms. However these transfer functions as noted above are dynamic and here is where things get complicated fast.

Snare waveforms even contain square waves so the Denon Technology, for example, of filtering square waves on the output would alter a snare sample's accuracy). This is particularly difficult with most dumb A-D, D-A converters. With bit mapping one hears some improvement, but we are far from AI and truly accurate sampling... if anything Emu Systems may have the lead because of my previous statement regarding Dave Rossum's research.

For accurate string ensembles I personally would chose a multitude of varying instrument samples and de-tune each randomly, if possible retaining the formants --- and in an acoustic instrument this will be the air resonance or phase cancellation at the neck. Thus each fundamental should be ever so slightly out of tune (in hundredths of a cent) but without altering the air resonance tuning...

With held out chords: string sections, or even horn and reed sections eventually 'morph' into a perfect fifth or fourth with a Just third or sixth, again that is the idea behind LaMonte Young's composition '1960'.

'1960' if played with orchestra as intended, one realizes how long it takes for the musicians on stage to attempt to play a perfect fifth with no beats as an orchestra.

Also realize that orchestras will naturally aim for just intonation and not a tempered scale... same with blues guitarists. Most guitarists would re-tune prior to changing into certain keys... even heavy metal guitarists (besides leaving the third out of chord voicing) will tune to just tuning.

So, the ability to de-tune notes, leaving untouched the air resonance, phase cancellation of the neck, regardless of the pitch of the note, and do this in a slightly random manner for each violin on a per note basis and then an ability to save the patch, then layer each of these altered violin samples to create a string section 1st, 2nd, 3rd, etc. is the ideal sampler and probably why little response is heard from developers.

[Stig Eliassen]

O...M...G, 21st. You lost me again.

What I need is a sample library that works out of the box. I don't want to spend time re-tuning patches - the developers should get that part right from the start. Their lack of response and acknowledgement is plain stupidity.

Secondly, EWQL use the Play software, which is fine by me. Kontakt is a rather massive sampler, and I don't wanna mess around in it to "fix" patches before I can use them.

I've started to get deeper into the EWQLSO platinum library, and will review my needs again soon.

Thanks.

[John_Toolbox]

You might want to consider this since you're already running VE pro with a slave:

http://www.audioimpressions.com/


The price has just been reduced to $999

[Stig Eliassen]

Originally Posted by John_Toolbox

You might want to consider this since you're already running VE pro with a slave:

http://www.audioimpressions.com/


The price has just been reduced to $999

Oh no?! Another one to consider?!

Thanks, John.

[DarylG]

Originally Posted by John_Toolbox

You might want to consider this since you're already running VE pro with a slave:

http://www.audioimpressions.com/


The price has just been reduced to $999

I don't think this is a very good product. The ideas and technology behind it are very interesting, but I don't think it sounds as good as VSL, LASS or even Hollywood Strings. It is better in tune than HS though.

D

[Rectifried]

posted on the VI forum
but most there it seems dont use PT...composers I guess dont... and alot seem to be PC not mac

I finally ditched my PCI HD3 for a HDN PT9 and mac 8core 2.66 14gig ram... so VIs are back on my radar...
but would rather not do separate machine... did that with giga for years

the EW DAW help section doesnt even mention PT....EW lurker jayasher said dont even try w/o VEP

So Im back at my old DUC


is anyone using HS/play in PT without VEP?,PBidue

I like the idea of HS with the 2 for 1 now I would do the EWQLSO, HW


I know rtas is weird with VIs and I know PT should go 64 bit...

but who is scoring here with PT..
what works...

[aka21stCentury]

With 12 gigs you could run REAPER in rewire slave mode. I agree on the overall sound of the strings.

But I liked the performance realism. Over saturated in a hall reverb IMO. On 18th Century strings that last sustained chord reminded me of an Emulator II sample.

We never liked sampled strings in the old days (pre VIs). Additive synths yes. Sampled strings no. Particularly for adagio strings or legato cello.

My favorite adagio strings were on a Kawai (K-5 rack) that I foolishly sold (it was a NAMM loaner with white noise problem, so Kawai gave it to me, but a gate would have made it at least functional). My foolishness never ends.

Audio Impressions runs on Kontack. Tuning was good to my ears, but one could re-tune for alternate scales in Kontack 4.1 (but not Kontack Player) and utilize Just or Pythagorean tuning as in Omnisphere, MetaSynth, or the rest of Komplete 7's VIs (excluding Kore 2).

I am wondering if an exciter may not help with the overall sound. $999 is a good price. 5 MIDI tracks. That was impressive. With SVP I would run 60-80 tracks on a Quadra. Remember those?

[John_Toolbox]

Originally Posted by aka21stCentury

With 12 gigs you could run REAPER in rewire slave mode. I agree on the overall sound of the strings.

But I liked the performance realism. Over saturated in a hall reverb IMO. On 18th Century strings that last sustained chord reminded me of an Emulator II sample.

We never liked sampled strings in the old days (pre VIs). Additive synths yes. Sampled strings no. Particularly for adagio strings or legato cello.

My favorite adagio strings were on a Kawai (K-5 rack) that I foolishly sold (it was a NAMM loaner with white noise problem, so Kawai gave it to me, but a gate would have made it at least functional). My foolishness never ends.

Audio Impressions runs on Kontack. Tuning was good to my ears, but one could re-tune for alternate scales in Kontack 4.1 (but not Kontack Player) and utilize Just or Pythagorean tuning as in Omnisphere, MetaSynth, or the rest of Komplete 7's VIs (excluding Kore 2).

I am wondering if an exciter may not help with the overall sound. $999 is a good price. 5 MIDI tracks. That was impressive. With SVP I would run 60-80 tracks on a Quadra. Remember those?

You need not go into kontakt for tuning in different scales. Pretty much anything you'd ever need to tweak is controlled from the DVZ core engine GUI. I switched to this from EWQLSO and the IK miroslav philharmonik and haven't once wanted to go back. My old template had over 400 tracks, this is a much, much cleaner way of working once you get used to it. The only catch is your slave PC does need to be booted in windows, the DVZ GUI is a .exe file.

[John_Toolbox]

Originally Posted by Rectifried

but who is scoring here with PT..
what works...

Hosting the instruments outside of Pro Tools(Ve pro), or being stuck with < 4GB of RAM available for loading VIs. Until they release a 64 bit PT, Vepro is the way to go.

[DarylG]

Originally Posted by John_Toolbox

Hosting the instruments outside of Pro Tools(Ve pro), or being stuck with < 4GB of RAM available for loading VIs. Until they release a 64 bit PT, Vepro is the way to go.

64bit won't really help that much unless the VI performance of RTAS improves dramatically. Currently it's an embarrassment.

D