makers of Tassman make a horrible guitar sim

[braincell]

Red, I enjoyed your post but I disagree with something you wrote. You said an emulation can never sound analog. What does an analog synth sound like? An Arp sounds nothing like a Moog for instance. They have two *totally* different sounds. I think you are being a snob. What we mean by analog synth sound is just a warm sound and several of them have it and like I said, listeners don't care.

Gary, I disagree with your laws. Use what you want to use and let listeners decide if they like it. I've recorded guitarists before and it worked out well but it is logistically more difficult. Setting up the amp and mic and preamp is tedious plus you have to worry about the room sound. I don't think there is a good guitar simulator but I expect something reasonable in the future. Personally I don't need a guitar for what I'm currently working on. I have nothing against simulators in theory. This seems like a religious issue to some. This is why you see so many guitar bands with no keyboard players. It's a rigid way of thinking and it's holding back music in my opinion.

[kensuguro]

I think digital has much to offer, and indeed if physical modeling would be treated more as "waveguides" - conceptually ditch the need to imitate perfectly good cheap and abundant instruments - and concentrate on the synthesis capabilities itself, we could see much more interesting progress than we do today (thanx for the idea btw)

Well, that's exactly the issue I think. The words "physical modeling" already puts too much emphasis on trying to sound like the real thing. And usually when it fails to do so, the escape route is "well, listen to all the other sounds it can make!". That's a useful way to use the synthesis engine, but not necessarily in a "physical modeling" sort of way, since you start to completely ingnore whatever the physical properties it was you tried to model. (and failed at) Of course, I'm all for using these synthesis basics as a different approach to synths in general. Frankly, I don't think anyone's succeeded in doing so. Atleast not to the point where I would consider it a creative tool.

For ex, a kaplus strong algo, even with its many variations and using different waveforms for initial impulses, etc., still sounds like a broken guitar that downright sounds like a bad karplus strong implementation... or just a strange feedback loop that reminds me of early granular synthesis. (at high delay values) But that's just me, perhaps there are other ways to use the delay chain, or other chain configurations that can produce more interesting results (almost like FM).. It will be interesting to see a karplus strong be used, and not advertised as another "mimicks plucked string instruments!!" type synth. If it is, though, it better frickin' get the plucked strings emu right before it claims to do anything else. That's just my way of viewing emus... do what you say you'll do, or don't claim it.

An example of a successful (even by chance) synth is the klatt synth... or more easily recognizable as DR. Sbaitso. It's a type of vocaloid synth that I believe is embedded in everyone's brains by now.. It's become the staple of retro-future sounds, that it's used in words, speaking, bits and pieces of incomprehensible sounds, etc.. used as themes, foreground material, background percussive material.. That's a good example of a synthesis technique used for something other than what it was intended for, but retained enough mystery (because of its imperfection) to be perceived as just an instrument, rather than a "fake voice". But I think that worked because the emulation was so bad, that it wasn't believable at all in the first place. And people took advantage of that. So, I do hope that bad physmod can be used in this way some time in the future, just that it hasn't happened yet. It is full of potential, but again, not as a full physmod with believable emus, but just as a synth in its own right. I wished I had the time and resources to look into this further.

I did try the karplus strong with Mod III... just that I didn't have any real way of converting MIDI to frequencies, to tune the delay line. Easy to do in max/msp... but requires a very hack-ish approach on the mod. And I don't think the usual delay line would work, it requires strange configurations, variable configurations... stuff you wouldn't get in a pre-fab karplus-strong implemenation. Bleh, sometimes I don't know if I want to spend the time building stuff like this, or spend time writing stuff with sounds and synths and getting the results I plan for.

[ReD_MuZe]

brain:
i never said analog is better than digital. its a different animal.

i generally prefer digital actually. but you cant beat the original.
minimoog sounds better than any emulation and even much better than its own child the voyager (there is realy nothing like a well maintained minimoog in action in the hands of a pro minimoog player).
nothing to do with analog vs digital voodo. it has to do about emulation vs. imitation vs. innovation.

there is an inherent problem with emulations. the original sets a bar, that can never be passed, and often can never be reached. so instead of progress we get the regression.

the reason you have guitar emulators today, is because for centuries people have fine tuned this instrument to perfection and still are. better guitars are being made every day. (wish i could say that about celli)

by the way. not big news to me about being a snob. some people mistake artistic principles as snobism and thats ok with me. its just a name for the same thing - liking specific things. its called taste i heard your music, read your posts. you are just as guilty as i am

and gary i cant agree more, the nice thing about it, is often once you get your music preformed better, you get to sell more of it. 100% of nothing is less than 20% of something . its our job to create the industry. we are the industry.

ken:
the reason it usually sounds bad is actually the quality of the interpolation, but i'm not going to step on anyones toes here... with the propper tools, even the simplest karplus strong algo can sound beautiful, but realy a waveguide is a much more complex structure than a karplus stong. just for something to be counted as a waveguide it needs feed-forward and feedbackward... so here is a small recipe for a very cool waveguide module (perhaps ill implement it on 3.5 aswell)

take two delays and chain them serially, and feed them back (try all kinds of phase settings and also cross feedback is nice. you can mix those two delays and everything. so whats special about it? whats the secret? simple. when you calculate delay times, you sum both delay times. and the difference between them will give you a "standing wave" or a plucking position really since this is a 1d model.

the nicest way to go is this:
delay time 1 = position
delay time 2 = Global delay time - position

so you always set your pitch with the delay time, and the position gives the standing wave. i hope thats clear enough.

fiy, this technique is patented. i think the patent is now void, but still its amazing to see such a fundamental physical principle being patented. This simple waveguide, with some filters and the right exciter, will make a great evp88/evd

too much herbs in my tea

nite all!

[kensuguro]

hmm... I had always assumed the stuff had to have feed backward.. so perhaps I was thinking of the waveguide karplus-strong variant... I mean, physically the sound does feed backward. I'm not quite sure what you meant by the interpolation tho... at what point of the process are you referring to? Just the interpolation that happens during the decay within the feed back/ feed forward loop? Either way, the idea of interconnecting, and subdividing these delay lines sounds interesting... and computationally heavy. I guess these would amount to physically connected Y strings and whatnot in the real world, but... well, it's sort of not worth it to try to imagine the pysical counterpart at this point.

lol, just saw your edit... so it's patented eh. That's like saying Tetris is patented... Or someone claiming the rights to the game of poker. lol...

[braincell]

Better guitars are made every day because there is a huge market for them, also you are allowed to experiment in guitar making. Any cello which is radically different in design would not be accepted. Obviously a composer expects his composition to sound a certain way. I believe the most recent instrument to be widely incorporated into the symphony orchestra is the saxophone which was developed in the 1840s by Adolphe Sax. The modern bassoon emerged 40 years before that. They use electric guitar sometimes but I don't think it's standard. Maybe I'm wrong.

[garyb]

Gary, I disagree with your laws. Use what you want to use and let listeners decide if they like it. I've recorded guitarists before and it worked out well but it is logistically more difficult. Setting up the amp and mic and preamp is tedious plus you have to worry about the room sound. I don't think there is a good guitar simulator but I expect something reasonable in the future. Personally I don't need a guitar for what I'm currently working on. I have nothing against simulators in theory. This seems like a religious issue to some. This is why you see so many guitar bands with no keyboard players. It's a rigid way of thinking and it's holding back music in my opinion.

if you don't need a guitarist, don't use him. otherwise, if he can do what you want, hire him. that's what i said. also, the hard part about it is what makes it great, if it was easy everyone would do it. also, also, the dynatube is pretty good...

[ReD_MuZe]

hmm... I had always assumed the stuff had to have feed backward.. so perhaps I was thinking of the waveguide karplus-strong variant... I mean, physically the sound does feed backward. I'm not quite sure what you meant by the interpolation tho... at what point of the process are you referring to? Just the interpolation that happens during the decay within the feed back/ feed forward loop? Either way, the idea of interconnecting, and subdividing these delay lines sounds interesting... and computationally heavy. I guess these would amount to physically connected Y strings and whatnot in the real world, but... well, it's sort of not worth it to try to imagine the pysical counterpart at this point.

lol, just saw your edit... so it's patented eh. That's like saying Tetris is patented... Or someone claiming the rights to the game of poker. lol...

yes, to get standing waves, you need the sound to travel both ways. forward, and backwards. so the basic waveguide has both feed foward and feedback.

Interpolation - meaning delay interpolation aka - fractional delay. there are many ways to interpolate delays. unfort scope only has linear interpolation for the delays - something thats fine for flangers, but hardly for waveguides. there are many ways to interpolate delays and each way has its pros and cons.

actualy waveguides are not computationaly heavy at all, and worth giving a go. maxmsp is a better choice imo for this kind of rnd, but its very easy to get a delay tuned in mod: take a constant freq and connect it to the X of the division module the y of the module should get the mvc frequency. that should output a good signal for the modulation delays in mod. to tune the freq, use the constant freq knob. quite simple.

Brain, no one is forcing you to use a guitarist. it has serve your music first of all. and i think the logistic mess is:
1) well worth it
2) not such a mess when you rig yourself to do it

in my studio it takes 1 phonecall and 15 mins of setup to record a guitarist (about the time it takes him to get to my place). and thats considered not very efficient. but in my book well worth the "mess".

actually the fact that ravel used a sax in an orchestra doesnt mean its a standard part of it. an orchestra is an instrument in its own right, im not sure that judging instruments by their orchestral role is of any musicological importance . the reason there is such a big market for the guitar is exactly that successful evolution path i'm talking about.

[braincell]

My mistake Red, I looked it up. The saxophone is not an important orchestra instrument. It was actually invented for marching bands originally.

My point was really that the design of an instrument is essential for a genre. I disagree that instruments and genres should be static but this is what hard core fans and musicians want. The electric guitar is newer and isn't subjected to such ridged standards but in fact when it was introduced, a lot of acoustic guitarists abhorred it because it doesn't sound the same at all. I could imagine a trumpet made of wood would sound very nice but it aint gonna happen. People are too conservative to buy such a thing.

Any development which expands the sound of the synthesizer is good. There are too many analog sounding ones. Enough is enough.

[Fluxpod]

Look up the yamaha vp1.It can do a LOT of really neat tricks with acoustic modelling like strings out of stone and severel meters long.It was around 60000€ new but the rack version sometimes show up cheap.I had the pleasure to play the demo in colonge around 1998 i think and man that thing is SICK!It maybe get you a wooden trumpet.

[ReD_MuZe]

woop, a musicology buff here
trumpets started out wooden

http://en.wikipedia.org/wiki/Cornett
http://en.wikipedia.org/wiki/Trumpet
http://farm2.static.flickr.com/1086/140 ... 62.jpg?v=0

and actually your sax story is not that baseless. Ravel - who is considered as one of the greatest orchestrators in the 20th century (allot has to do with his work on musorgski's pictures at the exhibition - a true masterpiece - recommended!!) has introduced the sax as a part of his orchestrations as a standard. it is actually a heavily debated matter if the sax should be considered an essential part of the 20th century orchestra or not. and this subject is indeed interesting since it has consequences to seemingly unrelated things, such as how the orchestra seating should be arranged, who has the right to set standards and so on...

I see your point. I think its a beautiful feedback process between design and usage. different designs lead to different usage, which leads to different designs and so on.

lets look at 3 electroacoustic instruments:
1) Electric guitar - a massive hit, a great sounding instrument, with ALOT of expression
2) fender rhodes - a classic but not a hit by any means. sound is nice and expression is still possible
3) clavinet - a total dud. good for funk riffs and thats it. about as expressive as a cembalo (not very).

all three sound different from their originals, and if you would go by original instruments (classic guitar, piano, cembalo [harpsichord] ) you would guess that a piano or keyboarded version would have been more popular. indeed a classical piano has more expression ranges than a classical guitar and the ratio of classical piano/guitar pieces written prove im not the only one how had this idea.

The trick is not to judge the instrument by its ancestor (a cornett sounds awful, and humans are not monkeys), but to try and get a fresh look at it and see what its really good for. which is what jazz and rock did for the guitar (and drums and much more), and what psy-trance, ambient and idm (and other few electronica genres) do to synths today.

Fluxpod - in all honesty are those sounds you would use in your own music, or are just just fun to tweak? i heard the VP1 and wasn't that impressed with its musical potential. it is a very impressive instrument, but not very musically usable imo. prophecy aswell, the emulations are impressive because they are the best. but in this case being the best, doesn't necessarily mean being any good at all, just not as bad as the rest.
kind of like the simulanalog amp modeler vsts. they sound the best, but i still wouldnt use one in my music. and i tried. but once a real amp was recorded it was clear as daylight.

having mile long strings and nonsymetric string models (enharmonic) is just a matter of delay length, and allpass in the path. with some more imagination it is possible to make indefinable sounds and not just strech the numbers of a very distant aproximation of 1% of the physics that actualy take place in vibrating string. in my opinion the whole physical modeling approach is based on some kind of marketing hype, and a promise for something unforfillable. those ambiguous settings of string length are kind of like giving you a crash cymbal and telling you, you can make it bigger and smaller by changing its pitch. you know better than that because you know how it works. i think its time we unset the vail from waveguide synthesis, show you guys the guts of the engine and see what you can come up with. without trying to model non existent stuff. but rather use waveguides as a way to create musically useful sounds.

[braincell]

The piano was once very popular. I did a little research on this. Commercial pop music on a wide scale started at about 1885 in NYC. At that time there were no recordings so songs were distributed by means of piano sheet music. The first recordings were around 1904 but recording technology didn't become widely available until around 1940. The early song writers were the first pop stars. At the turn of the century many song writers were making what would today be millions of dollars, lyricists too. They led wild lives with lots of parties, drinking sex and some say pot which was referred to as "spinach", although this may not have been until around 1930.

Besides the invention of records, the reduction of the piano in popularity was due to the fact that they are not portable and can not be easily tuned by the owner and they are expensive. As a piano player, I also find that when you play it with other instruments, it becomes difficult to hear. It starts off percussive but then fades and the notes tend to blend together and they are normally mellow. An electric guitar has a very piercing, irritating sound which cuts through all the other sounds quite well. You need an instrument like that when you have a drummer banging the hell out of his kit behind you. The cymbals particularly take up a huge amount of spectrum.

[next to nothing]

"Besides the invention of records, the reduction of the piano in popularity was due to the fact that they are not portable and can not be easily tuned by the owner and they are expensive. As a piano player, I also find that when you play it with other instruments, it becomes difficult to hear."

It never got less portable, and i might be wrong but pianist never was known for carrying their instruments along... i dont know what era you refer to, but the piano has been a venue-hold for oh so long.. i admit that trumpets and guitars are more easily carried around, but as you have stated yourself, maybe the drop of popularity was more a symptom of new styles and sounds?

[ReD_MuZe]

Here is another project aimed at jingle makers:
http://labcast.media.mit.edu/?p=35

Its amazing how much time and energy people spend on fixing perfectly good instruments.

What is this about? saving space for guitars? hmm...

but then i stumbled upon this thing:
http://futurecraft.media.mit.edu/amitz/ ... uitar2.jpg

which imo is much more interesting, providing it sounds good

i think musical instruments have to be an artwork in their own right. if its just a scientific/mechanical concept, it just wouldn't work. people love gibsons and fenders because of their personality, not their functionality.

[braincell]

Things change. If you are an instrument maker, you want your creations to have an edge over other instruments. The point of the saxophone was to make a louder wind instrument. Sax played flute which is why saxophones have a similar fingering to the flute. You work to improve Flexor too.

The piano was the main source of music. Other instruments were available but with the piano, you can be a one man band. Two hands and your voice. It wasn't so much the sound of the piano that made it popular. That was the music you could buy. Go to any large antique shop and you will see hundreds or thousands of pages of piano sheet music. It is still cheap because there was so much of it printed. Pump organs were available too. My friend has one he found at the dump with a fan in it. The fan makes more noise than a modern computer. The instrument sounds like crap.

[astroman]

Here is another project aimed at jingle makers:
http://labcast.media.mit.edu/?p=35
Its amazing how much time and energy people spend on fixing perfectly good instruments.
What is this about? saving space for guitars? hmm...

first I found this 'modelling by wood' an interesting approach, but as soon as they added their 'virtual resonance chamber' it became clear what instrument making really is about...
these people pretend to know what they are talking about, but can they listen ?
it was the most lifeless tone of a guitar I've heard in weeks
should have made it clear that they are at a (conceptual) dead end, but somehow they manage to ignore it

cheers, Tom

[ReD_MuZe]

yep its more an intellectual escapade than a musical one.

this is a common symptom of a scientist that was never very good at violin, but very good at math. who didn't really like math, but loved the violin.... (or guitar in this case)
the poster in the background looks better than the guitar sounds...


actually, it also sounds better than that guitar

[braincell]

Besides expert craftsmanship, cooler weather made very dense wood which is why the Stradivarius violins sound so nice. That is the current theory anyway. A Stradivarius was recently put under a medical CT scanner.

[ReD_MuZe]

well cremona at the time was full of violin builders all using the same wood...

the strad is a total mystery any attempts to copy it were unsuccessful, every decade someone else says he know what it was. once its volcanic ash, once its weather. but none of that accounts for the otehr violin makers at cremona at the time that didnt get famous and their instruments dont sound as good.

its not all about technology
sometimes its more important that something actually works, than to figure out how.

[garyb]

can't argue that...