Re: Comparison of triple play and axon by the inventor

"JAMES HARRISON [email protected] [midiguitar]" <[email protected]> Mon, 16 Mar 2015 09:51:18 -0400
Newsgroups gmane.music.equipment.midi-guitar
Message-ID <[email protected]>
Thanks so much for that, Greg.

James
.
.


>         Sun Mar 15, 2015 7:52 am (PDT) . Posted by:
>
>
>       gregor_malecha
>       <mailto:[email protected]?subject=Re%3A%20Comparison%20of%20triple%20play%20and%20axon%20by%20the%20inventor>
>
>
> I now went through the mails related to the Fishman Triple Play that I 
> have kept on my computer...
>
> The earliest descriptions there are actually already from 13.11.2011 
> and 15.11.2011, when it was still called "Wireless Guitar Synth" WGS6000.
>
> The most relevant posts (plus some related ones) I'm combining here...
>
> (I hope yahoo doesn't mess up the line spacing as it's now often the 
> case.)
>
> :^) Greg
>
>
>
> ----- Original Message 
> ----------------------------------------------------------
> From: [email protected] On Behalf Of Wayne
> Sent: Thursday, October 13, 2011 5:05 PM
> To: [email protected]
> Subject: Odp: [midiguitar] Re: Guitar Synth Challenge? 20 ms not 30 ms 
> (GR-55)
>
>
> In my tests, the GR-55 tracks around 20 ms, both in internal and 
> external mode:
> http://www.joness.com/gr300/MIDI_SPEED.htm
> The slowest sound was, of course, the low E (30 ms) but on average 20 
> ms. This is about the same as the GI-20.
>
>
> Wayne
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Thursday, October 13, 2011 7:23 PM
> Subject: RE: [midiguitar] Re: Guitar Synth Challenge? 20 ms not 30 ms 
> (GR-55)
>
>
> I cannot display here the graphical hystogram of my GR-55 
> measurements, I try to describe it.
>
> I played the lowest four notes (E,F,F#,G) quickly, repeatedly about 
> 100 times, in a fast, continuous riff. I measured the MIDI response. 
> 75% of the latency values on the hystogram were under a wide Gauss 
> curve, centered at 28 msec, the tails at 24 and 32 msec. There was 
> only one single value below 20 msec, it was at 19 msec. The 25% rest 
> were randomly distributed between 35 and 75 msec.
>
> The same measurement for the new WGS6000 wireless guitar synthesizer 
> (will be introduced at the winter NAMM) has a thin Gauss curve 
> centered at 15 msec, with the tails at 12 and 17 msec. Only two notes 
> were 18 and 19 msec, 98% was below 17 msec.
>
> Andras Szalay
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> (http://launch.groups.yahoo.com/group/midiguitar/message/33907)
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Saturday, October 15, 2011 10:08 AM
> Subject: RE: [midiguitar] Re: Guitar Synth Challenge? 20 ms not 30 ms 
> (GR-55)
>
>
> > Do we have a list price on the WGS6000?
>
> I cannot disclose it until it is not yet final. Let this be a surprise :)
>
>
> >> "WGS6000 wireless guitar synthesizer (will be introduced at the 
> winter NAMM)
> >> has a thin Gauss curve centered at 15 msec, with the tails at 12 
> and 17 msec.
> >> Only two notes were 18 and 19 msec, 98% was below 17 msec."
>
> > Is that faster than the latest Axon???
>
>
> First of all, there is no difference at all between the oldest and 
> newest Axon models; the pitch detection engine is exactly the same in 
> all models, they only differ in the user interface.
>
> As the Axon already reached the theoretical mimimum of latency allowed 
> by physics, the answer could be no. But this is not true, the essence 
> is in the details. Even on some Roland models -- especially the GI-20 
> -- you can notice pitch detection in slightly more than one vibration 
> period, the same as the Axon or the new system. The question is 
> whether this happens in 1% of the cases, by cleanly picked individual 
> notes, or in 99% of the notes, even in dirty riffs.
>
> The really revolutionary thing about the Axon was not the device 
> itself, but the development system that I built to it. It allowed me 
> to detect and record every error that happened, and theoretically it 
> was possible to squeeze the software to repair the error. However, 
> this process came to an end as the necessary processing power reached 
> the top of the processor's capabilities, and the software became very 
> complicated as I had to concantrate on two things at the same time, 
> the functional performance of the software, and runtime optimization 
> to fit it into the available time.
>
> Now, on the WGS6000 I have a development system that is even far more 
> advanced than the Axon's; it took me four years to build it, while the 
> device itself was developed in three years.
>
> I use now a processor that has practically unlimited available 
> processing power. This means, I can write much simpler software, I can 
> concentrate on the functionality and not on optimization. In spite of 
> that in the current state the software requires just slightly more 
> than 10% of the available processing power, so there are practically 
> unlimited reserves for further improvements, if it would ever be 
> necessary.
>
> Therefore, the new system reaches the minimum latency allowed by 
> physiscs in practically 100% of the cases, which Roland devices 
> perform very rarely, and the Axon maybe in 50-70%.
>
> The enhanced reliability is reflected especially in very stable input 
> to notesheet translation, as John McLaughlin reported me who started 
> to use it middle of September.
>
> Andras Szalay
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: [email protected] On Behalf Of mnfraga
> Sent: Sunday, January 22, 2012 1:14 AM
> To: [email protected]
> Subject: [midiguitar] Re: Fishman Triple Play
>
>
> Hi all,
>
> > I wonder what this thing will give that the Axon doesn't.
>
> 1. Freedom from 13-pin cables;
> 2. Extreme portability;
> 3. Same accuracy with less errors;
> 4. Lower starting price (my Axon NGC77 was US$1200, 4 times more);
> 5. No pickup needed -- magnetic or piezo.
>
> There are drawbacks too. You'll need one of these for each guitar you 
> have.
> Still, you can buy 3 for the price of one Axon 100+pickup.
>
> Marcelo
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Sunday, January 22, 2012 9:01 AM
> Subject: RE: [midiguitar] Re: Fishman Triple Play
>
>
> >I am still at the NAMM, but let me answer to some of the questions 
> raised here:
>
> >> You'll need one of these for each guitar you have.
>
> 1.) The whole system clips on the guitar with a magnetic fixture, 
> something like the power connector of a MAC laptop, so changing from 
> one guitar to the other will be a matter of a few seconds.
>
> 2.) For a 13 pin system you need a GK-2 compatible pickup for each 
> guitar. It costs almost as much as the whole system here, so even if 
> you have separate sytems for each guitar it is much less investment 
> than having 2 * GK-2 + 13 pin cable + an AXON or ROLAND.
>
> I deleted some messages, but I remember remarks like that:
> >>"it looks like a toy"
> The palm-size box which looks like a toy contains more than 1000 times 
> higher processing power than the original AXON.
>
> >> Does it have fret, string and pick splits like the AXON?
>
> It does not have pick splits, but fret splits and string splits are 
> much more flexible. While in the AXON the corners of four segments 
> meet each other in one common point, here two such points can be 
> defined independently, one for the lower splits and one for the upper. 
> This means that the lower and upper splits can overlap, even fully 
> cover each other as two layers, and the string split of the lower and 
> higher range can be different. This is possible because the lower and 
> upper splits are transmitted on separate MIDI channels, the lower on 
> 1...6, the upper on 11...16. Burr used these features extensively in 
> his demo.
>
> Andras Szalay
>
>
> ----------------------------------------------------------
> Another interesting post by Marcelo Fraga describes how the occurrence 
> of glitches can vary depending on the software the AX-100 (or the FTP) 
> is used with
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "mnfraga" <[email protected]>
> To: <[email protected]>
> Sent: Sunday, January 22, 2012 2:30 PM
> Subject: [midiguitar] Re: Fishman Triple Play
>
>
> Hi Ken, Andras, Mike, and all,
>
> > The Fishman Triple Play does include a magnetic hex pickup, similar 
> to the Roland GK-3.
>
> Pardon my phrasing, what I meant was: no need to BUY a pickup! ;)
>
> I mentioned my first Axon was US$1200. Well, add $179 for the pickup, 
> $159 for the DB50XG soundboard, and the I also spent about $400 in a 
> good midi interface for the computer. Do the math...
>
> Andras: I thought it was screwed in place, so what you are telling me 
> is excellent news. :)
>
> Mike:
>
> > "it looks like a toy"
>
> My old plastic Macbook looked pretty much like a toy, and yet was a 
> million times better than my first PC, heavy, bulky, noisy, and 
> useless for making music.
>
> > "And it's not glitch free. There are still glitches, not many, but 
> they happen."
>
> Have in mind a lot of the "glitches" you may have seen/heard may have 
> to do with Logic more than the system itself. I have worked 
> extensively with Logic. The time I've spent trying to make it respond 
> to midi guitar in a nice way is almost the same I've spent setting up 
> the Axon itself. Since version 8 it became even worse, further 
> stripped from certain functions. I had to shift the basic channel of 
> my Axon to 2, because it won't accept ProgramChange messages on 
> channel 1, for instance.
>
> Then, you hit the very input of the system, which is actually a (bad) 
> script! It further removes midi information from the midi stream, so 
> you have to manually cable the midi input to the instruments in use.
>
> Just so you have an idea. With my Axon properly set but not tweaking 
> the software, I'd say:
>
> 1. Playing the internal soundboard: all works fine
> 2. Playing Kontakt standalone with Gtak: all works fine
> 3. Playing Kontakt as a plugin in Ableton Live: few glitches
> 3. Playing Kontakt as a plugin in Logic: glitches and more glitches
> 4. Playing any Logic instrument: hell seems very close by
>
> Marcelo
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: [email protected] On Behalf Of hippietim
> Sent: Monday, January 23, 2012 7:14 PM
> To: [email protected]
> Subject: [midiguitar] Re: Fishman Triple Play
>
> Is there any plan to have a receiver that has just a MIDI out? The USB 
> connection to a computer is great but I really don't want to use a 
> computer live - I've got rack synths already.
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Monday, January 23, 2012 10:46 PM
> Subject: RE: [midiguitar] Re: Fishman Triple Play
>
>
> Believe us, we thought about all these ideas of piezo pickup, nylon 
> string guitar, floor unit with MIDI out, bass guitar compatible unit, 
> etc., and these will come. Also, the Triple Play supports a 
> Hold/Arpeggio/Looper + Chain pedal set, Burr used these in the demo. 
> This will be optional, not part of the basic package. Only if these 
> options would come all at the same time, it would take a hell lot of 
> time until release, and we did not want to let you wait much longer. 
> So, the product that was shown on the leaflet at the NAMM is a first 
> release, different packages and auxiliary extensions will follow soon.
> Do not forget that also the first release of the AXON did not work 
> with bass, with piezo, did not have Arpeggiator, etc...
>
> >> I've got rack synths already.
>
> Specifically, a floor unit with MIDI out was my priority, but I had to 
> admit that if we would start to sell a unit with a MIDI OUT only, I 
> cannot name a single rack synth today that would be a good match for 
> the Triple Play, except as second hand :(
>
> Anyway, if you have still more ideas that we probably did not think 
> about, we have open ears for that ;)
>
> Andras Szalay
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: [email protected] On Behalf Of James
> Sent: Tuesday, January 24, 2012 6:29 PM
> To: [email protected]
> Subject: [midiguitar] Re: Fishman Triple Play
>
>
> Hi, Andras.
> Congratulations on a fine achievement and thanks for monitoring here.
>
> One thing I'd like to be able to do [very important to me] is to use 
> the unit somehow to get an output of a 6 string bass from a six string 
> guitar. IOW, tune the strings down 17-17-17-17-16-16 [Or is the last 
> one 17; I forgot] semitones. What I want is a BEADGC output, one 
> octave down from a regular guitar. Can you do that with the first unit 
> [or the software it connects to]? In case you haven't noticed I'm a 
> midi dummy...
>
> Thanks,
> James
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Wednesday, January 25, 2012 12:46 PM
> Subject: RE: [midiguitar] Re: Fishman Triple Play
>
> James,
>
> just like on the AXON, you cannot transpose strings individually, but 
> you can define segments and transpose each separately. For your 
> example, you can define a string split with four strings transposed by 
> 17, and the other two by 16. This works.
>
> Andras
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "mnfraga" <[email protected]>
> To: <[email protected]>
> Sent: Wednesday, January 25, 2012 8:30 PM
> Subject: [midiguitar] Re: Fishman Triple Play
>
>
> Hi James and all,
>
> > One thing I'd like to be able to do [very important to me] is to use
> > the unit somehow to get an output of a 6 string bass from a six string
> > guitar.
>
> This can be achieved at the software level. The Triple Play will send 
> one string per channel (like the Axon), so you can transpose them 
> individually per channel. There are other options involving detuning 
> the guitar and transposing in the synth too. For instance, you can 
> lower strings 1-4 (or 5) by one semitone and transpose them all 16 
> semitones, then you get the 17 and 16 combination you want. (I'm not 
> familiar with bass tunings, sorry!)
>
> Since this first run of the Triple Play will be USB only (midi out 
> will follow, as Andras said), you'll probably use a DAW software and 
> softsynths/plugins. I'd put a vote down for Kontakt with Gtak, IMHO!
>
> (You can still output midi to external synths, transposed the way 
> you'd like, but that's something for another topic.)
>
> Marcelo
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "Andras Szalay" <[email protected]>
> To: <[email protected]>
> Sent: Thursday, January 26, 2012 11:45 PM
> Subject: RE: [midiguitar] Fishman TriplePlay details
>
>
> >> There is a Fishman TriplePlay OSX Control GUI app, and I assume a 
> windows version will exist too?
>
> We had three demo stations at the NAMM, two on MACs, the third one at 
> the back of the booth was a Windows PC.
>
>
> >> Can a user buy a base kit, and then purchase additional USB dongles 
> to connect to multiple MIDI hardware devices? If so, how much for an 
> extra USB dongle?
>
> If this is what you mean, one transmitter can send data to one 
> receiver only, this is strictly limited, since communication is 
> bidirectional and every message is acknowledged back from the receiver 
> to avoid the possibility of data loss. But of course, you can chain an 
> arbitrary set of MIDI and USB devices to respond to the transmitter, 
> but these must be connected with each other through MIDI ort USB 
> cables, depending on the setup. So, only one receiver for one 
> transmitter.
>
>
> >> I understand the intial version will not have wireless transmission 
> of the normal Guitar signal. Since the audio from the guitar is not 
> transmitted, what are the functions of the Mix and Guitar positions of 
> this switch?
>
> If you go to the PC with the guitar signal and insert our plugin in 
> the signal chain you can turn the guitar signal on or off. This is for 
> the starter kit. On stage, if you go to the guitar amp directly, for 
> the floor unit -- what exists as a prototype, that John McLaughlin is 
> using - you can couple the guitar signal through the stomp box, and it 
> is switched on or off by a relais.
>
>
> >> If your band has two guitarists, each with a Triple Play equipped 
> guitar, can you set each to transmit on a different frequency? Or must 
> you transmit on different MIDI channels?
>
> The system uses one of 80 radio channels, so the number of device 
> pairs is almost unlimited. You just press a pushbutton on both the 
> transmitter and the receiver that you want to work together, and they 
> tune to each other automatically.
>
>
> >> I never saw Burr playing more than two instruments or a different 
> instrument on each string. Assuming you had a fast enough computer to 
> handle the six softsynth plug-in instances, can the Triple Play 
> transmit a different MIDI channel on each string?
>
> This is what it basically does, and that was the case in all demos. It 
> sends the lower fret split on channels 1...6 and the upper split on 
> channels 11...16; the hold, arpeggio and looper is on channel 7, so 
> normally it uses altogether 13 channels. But just like the AXON it can 
> be configured to POLY MODE, sending all strings on channel 1, if a 
> compromise cannot be avoided. But in this case there are no splits.
>
>
> >> Is Pitch bend range adjustable?
>
> Yes, 0...24 semitones.
>
>
> >> What are the dimensions of the control box on the guitar?
>
> It is about the same surface on the guitar as the GK-2, but much more 
> flat, so it fits easily into a hard case without carving it out...
>
>
> >> what are the minimum and "ideal" system requirements for the 
> computer in question? Is it best to use a Mac? or will a fast Windows 
> PC work too?
>
> System requirements depend on the soft synth that you use, it 
> determines 99% of the needed computing power. PC and MAC are equally 
> good to use.
>
> Andras Szalay
>
>
>
> ------ Original Message 
> ----------------------------------------------------------
> From: "'Andras Szalay' [email protected]
> To: <[email protected]>
> Sent: Wednesday, February 11, 2015 3:29 PM
> Subject: RE: [midiguitar] Re: Fishman TriplePlay Wireless Guitar MIDI 
> Controller Versus??
>
>
> >>> Then after the Axon demise, FTP goes to a computer-based platform 
> (which
> >>> Szalay should arguably have done long before now ) to get the 
> processing
> >>> speed up.. but which has major disadvantages ( like having to rely 
> on a
> >>> laptop to gig with) and then it throws in a radio system to help with
> >>> the speed and the 'spaghetti cable syndrome'
> , all of which adds extra
> >>> costs...but for what?
>
> There are several misunderstandings here.
>
> The FTP is a completely self-contained device (the whole processing 
> happens on the guitar unit), the computer is just receiving the MIDI 
> signals from the wireless dongle. In January at the NAMM a new product 
> was introduced by Fishman, a stomp-box that hosts the wireless dongle 
> and can drive hardware synths, so you do not have to take a laptop to 
> your gig.
>
> The processor in the FTP can perform the typical signal processing 
> tasks significantly faster than computers, not to mention that it does 
> not have to fight with the speed-limiting factors of computer OS, like 
> input and output buffers that are necessary on a computer but cause a 
> noticeable latency.
>
> The radio chips used by the FTP are significantly cheaper than a 
> Roland 13 pin cable. So the radio solution is not only convenient, it 
> is cost efficient as well, and since the latency of the radio is less 
> than 0.5 msec there is no mentionable drawback.
>
> Andras Szalay



------------------------------------
Posted by: JAMES HARRISON <[email protected]>
------------------------------------

--to unsubscribe send a blank message to [email protected]