Is the Yamaha VL1 still the best physical modeling synth?

[analogholic]

Hello everybody,

 

I was wondering if you know what´s happening in this area?

 

I mean, lots of software synths coming...

 

Is the VL1 still THE shit when it comes to acoustic sounds?

[xmlguy]

No, the VL1 has been long surpassed. The Yamaha technology has been mostly incorporated into other gear in a different form. Instead of using acoustic modeling, they have moved to advanced articulations to give more realism. It has been replaced by Megavoices (Tyros, Motif ES, PSR-3000 etc.), Super Articulation with Tyros2 & PSR-S900), Expanded Articulation(XA - Motif XS), and Super Articulation 2 (upcoming Tyros 3). These newer models have a remarkably improved acoustic sound, very noticable to the average listener.

 

Physical modeling is still being used by Korg, Roland, Alesis, and others, in various forms. Roland has the V-Synth GT. Korg has Oasys with the STR-1 Plucked String synth. Alesis Fusion has physical modeling.

 

I think that Yamaha is taking the better approach for acoustic realism using articulations, rather than physical modeling. Samples of real instruments sound better and more complex than virtual model simulations, so long as articulations are used so that the samples sound dynamic instead of merely playing back the notes in a predictable, fake sounding manner.

 

Where physical modeling gets interesting is when it is used to make completely different instruments that do not actually exist or are more complex than what can be done with articulated samples. There's plenty of room for both technologies and there's quite a bit of overlap. Then there's the wind and reed controllers used for physical modeling in that niche market.

[GovernorSilver]

Isn't Yamaha's articulation technology based on samples? As one increases the number of samples in a multisample instrument, the sound and behavior gets more and more realistic.

[Tusker]

Isn't Yamaha's articulation technology based on samples? As one increases the number of samples in a multisample instrument, the sound and behavior gets more and more realistic.

 

Yes. Personally I find that samples (especially with articulations) work very well for ensembles, and physical modeling works well for solo instruments.

 

Yamaha hasn't updated the sonics of their modeling instruments for awhile. I am hopeful the next generation of Yamaha synths will revisit FM, AN and VL. (FM isn't modeling but it is a Yamaha synthesis asset.)

 

Many other manufacturers are using modeling, ranging from Arturia's brass, to the V-synth GT, to ...Harm Vissers Reaktor ensembles to ... the Oasys plucked string model to name just a few. Each of these has something to offer.

 

On the sampling side ... I find the VSL samples and articulations to be breath taking ... they are a little spendy though.

 

Jerry

 

OT: Where is Burning Busch these days?

[analogholic]

Thanks for the replies.

 

I guess I should have been more specific...

 

I´m mainly after very expressive solo sounds using a breathcontroller, also for live gigs.

Thats why I kind of left out samples...

 

Arturia Brass IMO has some nice trombone sections.

 

I have a Motif ES rack for which I got the VL Plugin (not the very best IMHO).

I also have a Triton rack with MOSS (cool for some stuff)

 

Are there any great solo modeling software plugins out there?

[GovernorSilver]

I've heard good things about LinPlug SaxLab 2 ... if you do not already play real sax....

[Mark Schmieder]

I'm with Tusker on this, which means I'm also with xmlguy. I have just recently moved away from Yamaha VL for most acoustic winds/brass, but still find it far and way the most MUSICAL (as opposed to timbrally realistic) for saxophones and even some members of the oboe family, but that may change soon.

 

I tried SaxLab2 a couple of weeks ago and felt it sounded more like a kazoo than a sax. I tried really hard to get something musical out of it, and do agree they have moved it forward from SaxLab1, but I don't think their approach is likely to yield convincing results ever. More on that in a bit.

 

I purchased The Trumpet a few weeks back, and while it barely runs on my underpowered G4 iMac, I was able to get some tracking of pre-recorded MIDI done with it (but no live playing), and am convinced it will become my go-to for trumpets in all genres. The expressive realism (the original parts were recorded using the WX5) is unsurpassed, and the timbre is eminently tweakable.

 

Those folks (Sample Modeling) are probably working with Nils, the Kontakt Scripting programmer who did all that great work for Thomas Scarbye's electric piano libraries. I'm guessing this since Nils also programmed for Peter Siedlaczek's earlier orchestral libraries, before he teamed up with Giorgio Tommasini:

 

http://www.samplemodeling.com/en/technology.php

 

Many like the WIVI libraries, but the demo version or the program is unconvincing (and also has few tweakable parameters compared to the full program). I personally feel the timbre is unconcincing in Wallander's work and too inconsistent between registers. Siedlaczek and Tommasini seem to be more on the right track (Giorgio previously produced the intriguing but slightly flawed cello and violin solo libraries for Garritan).

 

The bar has been raised significantly in the past two years, and I no longer find the trumpet in the Turbo-charged version of the VL70m to be acceptable as a result. I liked it for its organic character and playability, but had even started substituting some of Chris Hein's Horns recently (before buying The Trumpet). The Trumpet from Sample Modeling has all the realism and note conntectedness of the Scarbee e-piano libraries, and then some. You won't believe how it opens up in reaction to breath techniques, volume, etc. Truly remarkable.

 

I tried the Arturia Brass demos last year, but couldn't do much on my G4 iMac. My recollection is that the trombones were rather fake but sounded OK in ensemble. Trombone remains the most elusive of brass instruments, due to the slide mechanism.

 

I can always fall back on a real sax, but don't like annoying my neighbours and do most of my work past 9pm, so more and more am moving to "silent" sound sources such as modules and libraries.

 

Yamaha has done some remarkable stuff, and it's great live, but doesn't hold up in the studio due to over-compressed samples in hardware-based product. Vienna Instruments, however, with some of its sample-switching, is able to do pretty well with certain instruments and in ensemble (vs. solo instrument) mode. I especially find its trumpet sections useful, and for classically-influenced work, the saxophones library.

 

Don't waste your money, as I did, on the Special Edition and its Extended license, as they don't give you enough (quality-wise or quantity-wise) to work with in a serious project, and do not count towards discounts on other products.

 

FWIW, the Sample Modeling guys will be tackling other instruments besides the trumpet, but each has its own unique attributes and so it isn't a simple matter of taking lessons learned and applying them to samples of other instruments. Rumour is their next releases will be revivals of the canceled Garritan products (Gofriller Cello and Stradivari Violin) using better samples and improved programming (Kontakt scripting). Hopefully they'll offer loyalty discounts to owners of the Garritan versions.

[analogholic]

Thanks Mark for your informative and detailed answer.

 

I just checked out "The Trumpet" and it sounds unbelievably good!

 

From what I can tell from the demo Mp3:s, they seem to have "nailed" the combination of realism AND expression.

 

We can only hope that they will do more instruments (like sax, trombone) in the near future even if it´s very difficult.

 

btw, does the VL-70m and the VL plugin share the same engine?

 

I heard a couple of nice saxes at the Patchman site. (I know he really tweaked those sounds)

 

But in "theory", could I progam those sounds with the VL-plugin in my Motif ES?

 

[Mark Schmieder]

The editor for the VL is very limited, in that it can only work with existing models AFAICR. This is a qualified statement, as the editor is WIndows-only, so I haven't looked at it recently. Actually, maybe even that is incorrect; it might have been Mac OS9-only. At any rate, Yamaha long ago pulled the plug on the VL Expert Editor and only still posts the simplified version.

 

Matt Traum of Patchman did a remarkable job with the saxes in the Turbo chip, which is not available for the PLG card. Although the architecture is more or less the same as the VL70m, the chip implementation is different, and thus he couldn't easily do a version for the PLG card. I briefly owned the PLG card and found that with effects turned off, it sounded identical to the VL70m. But due to it not accepting the Turbo ROM, I sold it (I don't use the BC3a, which I sold, and don't like the hassle of having to hook up my WX5 with a power adaptor and standard MIDI cable vs. the WX cable and/or with the WX cable plus routing to another module/keyboard via standard MIDI).

 

The saxes are the strongest suit of the Turbo-charged VL70m -- especially the tenors. They are miles beyond the factory presets. The trumpets are in comparison not as giant a step forward from the already-reasonably-good factory presets. The mutes are the most realistic.

 

The trombones are useless, as are most of the lower-frequency instruments such as bassoon. This is slightly surprising, as the VL technology makes for a super-phat monophonic synth bass engine. I think it was even the basis for some of the AN1x work.

 

FWIW, the guy that did the VL did the effects modeling engine and the filters for the new Motif XS series. Maybe they'll use him soon for a revisit to some of this older technology, or some of those ideas applied within the context of Super Articulation.

 

The modeling algorithms are from the 70's and earlier. Korg uses them too, but took a different approach. They are Public Domain, and very little has been done with them since then. Most people doing new research into pure modeling (vs. hybrid technologies involving both sampling and modeling, or subtractive and/or additive synthesis or even FM in conjunction with other stock technologies), are focusing on the acoustic guitar (especially nylon/classical/flamenco and strumming or advanmced picking techniques), and acoustic piano.

[JeffLearman]

Just keep in mind that what we call "physical modeling" today will be called "incredibly simplistic math models that barely scratch the surface" tomorrow. There isn't a phyiscal modeling synth or program that gets very close to modeling the actual physics, other than to say things like "a harmonic series of sine waves models a vibrating string" ... ok, a little better than that but not a whole hell of a lot.

 

This is the area that's really going to take off in the next couple of decades as we have more raw CPU power for modeling purposes.

 

We've come pretty far with samples, and I don't expect any breathtaking improvements in that area. (Big improvements over ROMplers, which are still a bit short on memory, but not on things like Ivory.)

 

[Tusker]

But in "theory", could I progam those sounds with the VL-plugin in my Motif ES?

 

I wouldn't use the PLG-VL model for recordings ... except for some synthetic (abstract) expressions. It's a trifle grainy. There are some recordings made in the 80's with the original VL1 (which had more resolution). Some of these recordings hold up pretty well, even today. I am using three models (live) on the PLG VL these days with some regularity ... Mizu Horn, Floboe and Lyricon. They are not emulations of common instruments so you can concentrate on the expressiveness which (even today) surpasses most digital synths. I find that they respond better to breath than my patches on the nord modular ... the modular allows for non-linearities but things can get uncertain. These three VL models are quite well behaved, and easy to use live.

 

Jerry

[analogholic]

Thanks Jerry and Jeff for your answers.

 

Yeah, The PLG-VL is bit grainy, and think you right about using it for patches that are very expressive but maybe not emulations (so one won´t sound so fake )

 

Btw, this may sound silly...

 

Is there different resolutions in different keyboards handling breathcontroller data?

 

I´m using the halfcrappy CME UF6 at the moment (as it´s my only ax with BC input)

 

The CC resolution is very crappy, and pitchbend is not very good either...but BC data???

 

It sounds pretty "coarse" with pretty much anything I try to control...

 

[xmlguy]

I purchased The Trumpet a few weeks back...

Thanks for the info on The Trumpet. That's a very interesting development in sample modeling.

[Mark Schmieder]

CME gear is notorious for producing "jumpy", "imprecise", and "arbitrary" MIDI CC's, so it's not surprising its BC input is inadequate in response.

 

Kurzweil synths still accept BC input, and you could also buy a MIDI Solutions converter box, which has a standard MIDI output at the other end and can be MIDI merged if memory serves.

[analogholic]

Eeeh...should have thought about the old saying when I got the CME for a "good" price...

 

"You get what you pay for"

 

I know about the Midi-Solutions stuff, but I´ll try to fix a wrecked DX7II I have at home first.

[xmlguy]

I've seen a demo model CME 88 weighted key for the low $200 range at a local store, but with a broken pitchwheel. They might have a hard time giving it away.

[marino]

Just to clarify, the VL1/VL1m sounds several times better than the VL70m or the plugin. The 'big brother' has many more modeling parameters, more resolution, and much better effects. I'm using it to this day for reeds, flutes and saxes. Trumpets and trombones are ok depending on the type of music and the programming, but for those I prefer to layer the VL1 with (good) samples. Strings, and other things like analog modeling, aren't really so developed.

It can also do ethnic reeds and fantasy sounds with fantastic quality.

 

I've heard Arturia brass and another modeling program which did mostly woodwinds IIRC - and neither impressed me so much.

I'd be glad to hear a modeling program which works as well as the VL1, for reeds and saxes at least. I'll check The Trumpet for sure.

 

Now, why Yamaha abandoned the VP1 project is beyond me....

 

 

 

[Mark Schmieder]

If you remind me what VP1 stands for, I might be able to answer that question. I don't always remember all the short-form abbreviations for unreleased technologies :-).

[Cthulhu Fhtagn]

Not that his has anything to do with anything -

 

I normally post on the Lowdown Low Down forum. I play an upright string base in a marginally dysfunctional jazz trio. Without going into a long, tedious rant as to why my band should be tied up, covered in honey and left to the fire ants...

 

The issue is a keyboard ... a Yamaha Tyros to be precise. I have no additional information other than it looks like Michael Dorn should be standing behind it saying "Captain Picard, the shields will not withstand another direct hit from their distruptors". The argument is a drummer. The key say no because a drummer would not be able to keep up with her (???). The issue is the programing of the "arranger" and how it either will or won't loan itself to allowing for a live drummer in the band. The band leader/vocals is livid. She thinks the keys try to keep up with the drummer, not vice versa and that the arranger is holding us back.

 

Unfortunately, in this battle, I get to be the tie breaker. For some crazy reason, I would like to make an informed decision based on facts, not emotion.

 

So I defer to you, oh wise and grizzled veterans of the keys to help educate me on how this arranger works and its advantages and limitiations.

 

P.S. the sticky threads were a riot. Please note I am using an existing (though probably unrelated) thread.

[xmlguy]

Think of an arranger as a programmable backing track that gives you some control over how it plays, including chord changes, fills/breaks, intros, and endings, all with variations. However, it still depends on the skill of the musician controlling it, and for most Jazz forms, that's the problem with an arranger: any good musician will easily sound better than a great arranger for live music. A good Jazz musician will sound better than a bad one, regardless of the gear.

 

Another limitation of arrangers, and specifically the Tyros, is that you have much less control over the rhythm parts during live play than other sequencers, such as grooveboxes/MPCs with pads, step sequencing, part mutes, rolls, flams, fills, one shots, and live effects, as with the Electribes, SP-606/555/404, MPC1000/500. To effectively control the rhythm requires a musician who specializes in that part of the music, even if he/she is hitting velocity pads with fingers instead of drum sticks to do it.

 

Arrangers are better for the solo and duo performers, keyboardist/vocalist scenarios, to get a better sound than a recorded backing track.

 

If you don't use the Arranger features, the Tyros has a pretty good set of the Yamaha ROMpler sounds, but it isn't nearly as good as the Tyros 2 or PSR-S900 with the Super Articulation voices. It's also rather bulky for a 61 key with synth action, and I think most keyboardist would much prefer an XS8 for gigging than any arranger to get the weighted 88 and top of the line live performance features for a keyboard of that size.

 

With Jazz, if you can't "keep up", you probably don't belong on stage, regardless of what part you play. It sounds like your band leader is right. The keyboardist is too dependent on using the arranger styles and that would hold you back a lot for Jazz improv, unless she can stop using the arranger features and follow a real drummer or rhythm specialist.

[Cthulhu Fhtagn]

Thanks. I really needed this info. We had the drummer discussion about three months ago and the keys had made some objections (strenuous) that made absolutely no sense. Drummer not keeping up, sound quality of the instrument, the fact there were two musicians (one solo instrument and the keys).

 

The keys is taking lessons from a local celabrity who either solos or plays with a singer. Runs the same gear. Explains the reliance on the arranger.

 

The ROMpler explination says why she felt she needed to spend $1700 on another keyboard if we were to ditch the arranger.

 

And it explains the whole musician/arranger thing. I guess it boils down to our keys is OK. The arranger proabably allows her to sound better than OK a lot of the time (there are still some problems with intros and outros with the arranger) - with a full time bassist (me) and a full time live drummer, it kinda forces the keys to play out there simply on their own merit and talent. If the keys is mearly OK, then some problems we've had that we didn't know we had could become evident.

 

Am I on the right track?

 

Like I said, I'm trying to take the emotion out of the discussion.

[marino]

If you remind me what VP1 stands for, I might be able to answer that question. I don't always remember all the short-form abbreviations for unreleased technologies :-).

 

I'm not sure, but I think VP stands for "Virtual Polyphonic", while VL means "Virtual Lead", an elegant term to say mono. (well, the VL1 is actually duophonic)

 

I seem to remember that Yamaha built exactly *two* VP1s. The new modeling flagship was polyphonic, it sported new models (good strings in particular, but many others as well) which were never made available for the VL1, had an absurd price, and sounded fantastic. There are plenty of demos on the internet, and the sheer sound quality of that thing was stunning.

 

Then - complete silence. The production was simply stopped. Not even a bunch of them for the chosen few, like the GX-1 in the seventies. I guess nowadays, it could be made for less money... but they preferred to recycle the VL70m technology (for a while), which is the lowest step of the scale. Oh well.

 

 

[Mark Schmieder]

OK, that's what I thought it was, and Phil Clendennin stated a few years ago that they couldn't pull it in for a reasonable budget so dropped it (sort of like how the GX-1 was custom-built to order).

 

He also hinted that should that equation change, Yamaha might consider releasing a polyphonic PM engine in the future. But that was a few years ago by now, and they may have decided they are better served via articulation and phrasing tricks than via physical modeling solutions.

[orangefunk]

I was a big fan of the VL1/7 keyboards but sold my VL7 a few months back. I felt I just wasn't using it much and the effort required to learn how to play the thing was quite a hurdle. Its not just about blowing into a BC and playing, theres also significant foot pedal and mod wheel technique to learn as well as playing notes themselves.

 

I've posted this track many times here... but its the first thing I ever did on the VL (back in 2004/2005 I guess) and I've never been able to get close again...

 

Travels on VL7

[Mark Schmieder]

Well, the VL7 was severely crippled vs. the VL1 or even the VL70m, and I'm not sure how successful I would be forgoing the WX5 input mechanism. It was also set up for guitar, but I never tried that (though it's still possible, if I ever buy another hex pickup).

[orangefunk]

Sorry Mark, the VL7 is not a "severely crippled" VL1.. it is a monophonic VL1 (which was duophonic).

 

The patch data is virtually identical for both if you only use one element... the VL70 is nowhere near a VL1 or 7...

 

I have exchanged patches with VL1 owners in the past and there was no difference. The amount of patches which didn't work (because they used two elements) was miniscule..

 

If someone wants the VL1 then I suggest picking up a VL7 because for 95% of purposes its identical...

 

The VL70/EX5/PLG-VL were all different entities from the VL7 and VL1 IMHO.

 

 

[Mark Schmieder]

But the VL7 never got the ROM update like the VL1 and VL1m, correct? Everyone I know who owns a VL1 says the Yamaha ROM update improved the quality of the synth (basic timbre, core algorithms, effects, etc.) by practically a factor of four. Matt Traum of Patchman Music also wrote me a lengthy description of the differences, which I can locate in a printed pile if necessary but probably no longer have in electronic form.

 

I passed by an opportunity to buy a VL7 based on a lot of pro feedback. All of the VL1 and VL1m auctions were a bit tinged by scam artists and shill bidding, so I gave up -- though it's still on my list.

[orangefunk]

Yes it did. Mine was version 2. Like I said, I even loaded in patches from VL1 disks and they all worked... the only ones which did not were the ones which used two elements... for example some electric guitar sounds and sound fx.. but every horn and sax patch was exactly the same.

 

The VL7 was something like $3000+ compared to the VL1 $5000 or whatever it was... so its not really comparable to the lower end VLs which came later I don't think...

 

 

[Mark Schmieder]

OK, well that settles it then. I think the scarcity of the replacement ROM's may have led to some misinformation getting spread around the web.

 

Matt does say he still thinks the VL1 is better than the VL70m even with his Turbo chip, but that he did the best he could to get it as close as possible to that quality, and given the more direct comparison to the strictly monophonic VL7, feels that for many of the sounds it is now pretty close.

 

The effects in the VL70m are bloody awful; I bypass them for the most part. Although the Motif's effects are less specific than the VL70m's and thus not always as appropriate characteristically, the PLG150-VL sounds way better than the standalone module, IF using the effects. When comparing dry without effects, the VL70m wins hands-down due to the unique availability of Matt's Turbo chip.

[orangefunk]

There was a thread on here a year or two back where someone was arguing that the VL1 was way better than the VL7 along the same lines you referred. However, many VL1 people commented that their main patches only used single elements. Indeed the VL programmer (Japanese?) for Yamaha also talked about his main intention behind his original programs for the VL1 and how he planned certain kinds of sounds that took advantage of both "elements" (a voice element that is) but was not convinced about the advantages hence why most VL1 patches are single element and thus compatible with VL7.

 

I wish I could find the thread... I remember Carbon contributed to it... we had an argument about it IIRC...