McSeem

Thanks! There's an impressive demo by Burr Johnson with MIDI. http://truefiretv.net/graphtech/index.html?id=2 However, in my opinion, any MIDI produces sound refined to death. Not alive. I'm sure, you can do a lot with software, but I'm also sure that MIDI just kills the live sound. Also, with a live string feedback it's a huge difference from any sample-synthesized sounds. So, I want to keep the live guitar signal, combined with fancy effects and controls, plus sustain feedback. And of course, hexaphonic. Not sure what I can achieve eventually, but so far it's being very exciting. For example, I'm now experimenting with double filtering - direct and "counter-motion" one. Basically, computationally cheap infinite impulse response filters, but applied in both directions - forward and backward. No big success so far, but we will see.

Yeah, I understand. I my case it's more like innovative research experiments, so, at this step I can afford some extra expenses.

Sustainiac has a piezo-acoustic transducer mounted to the headstock. It works pretty well, but of course, it's also mono. What is bad about the under-string mounted driver is that you can't sustain certain harmonics, namely those ones that have wave nodes right at the driver. My preliminary experiments confirm it, so, as expected. So, an ideal driver would be... mechanical, like you say - a speaker, or a piezo crystal. Just a crazy idea -- if you have a floating bridge, you theoretically can try to attach a speaker coil to it, with a speaker magnet system mounted to the body. So that, it would vibrate the bridge itself. The system is heavy, so, you probably can't sustain high frequencies. I'm also thinking of hex electro-mechanical saddle driver, so, it would vibrate the strings in the parallel-to-the body direction, but I'm not sure if it's doable in practice at all. As for hexaphonic in general, yes, your aversion makes sense, mostly due to the wiring nightmare. I had really hard time DIY'ing it. :-) Besides, my system requires an audio-interface, a laptop, and a 6-channel power amp. Plus there must be some control pedals, and/or on-body pots and switches, that I could read in my software. I'm thinking of using some USB DAQ like this one: http://www.hytekautomation.com/Products/IUSBDAQ.html But as the major benefit - you can control any parameter you want, or even assign a single pot to control ANY number of parameters simultaneously.

In my case it's not a problem at all. I use Graphtech piezo saddles, which are already well balanced, besides, 6 pre-amps with adjustable gain. I'm not a big fan of DIY everything, if I can find a ready to use solution, I'd rather buy it.

Yes, that's interesting - to place the driver exactly between the pickups, and connect them in such a way that the driver feedback would be compensated. Needs experimentations. My goal is different - to have a truly hexaphonic system, so that you could sustain big chords. And of course, I want to make my system totally stable and predictable. I already have a wiring nightmare with it, so, having two hexaphonic pickups would double it. :-)

The humbucker won't reject the driver mag field completely. It perfectly rejects only fields that have no gradient, that is, namely 60Hz hum, induced by power transformers, etc. In case of the driver, there's a significant field gradient, as the driver and the pickup are placed relatively close to each other (even in case of the bridge and neck positions). If you use a separate pickup as the sustainer source, the driver will induce the signal in the input pickup directly anyway. The only solution I can see is heavy magnetic shielding of the coils, so that, the driver mag field must be as local as possible. Probably double coil humbucker-like configuration may also help, but I'm not sure how efficient it will drive the strings.

Hi, I get up this topic to report about my progress. Talking myself, basically. I finally made the sustainer coils, I described earlier http://projectguitar.ibforums.com/index.php?showtopic=43502 and by the way, I ended up with bobbins; I just figured out how to lathe nylon with a simple tool. The major advantage is that I could make them smaller. I inserted slightly shorter pole pieces and small neodymium disk magnets (3/16" in dia, 3/32 thick). The coils have about 12 ohms DC resistance. The driver looks like this: http://antigrain.com/hex_project/hex_driver_installed.jpg The technology, in case someone interested: http://antigrain.com/hex_project/tool_for_lathing.jpg http://antigrain.com/hex_project/equipment_for_lathing.jpg http://antigrain.com/hex_project/bobbins.jpg http://antigrain.com/hex_project/bobbins_w...pole_pieces.jpg http://antigrain.com/hex_project/coils_ready_to_install.jpg Well, I admit, it's not the best place for the driver, but it's just my initial experiment. Besides, it definitely wants some cover. I wanted to place the driver as far as possible from the hex magnetic pickup, trying to keep the neck HB. Also, the coils are not shielded anyhow. In the next version I'll try a heavy magnetic shielding, but it will take more space, as I expect, a humbucker cavity. I also tried light magnetic shielding, a cup made of a AAA battery, but the strong magnet completely saturates it and it doesn't have any effect. The shielding must be way much heavier. http://antigrain.com/hex_project/guitar_connector.jpg The guitar is connected with a 24-pin DVI dual link cable, which has very reliable pins (compared to Roland 13-pin DIN crap), and most of all, 7 pairs of separately shielded wires, so that, I can use a single cable for the input and output without any noticeable feedback. I use 3 cheap stereo mini-amps to drive the coils, which is not the best idea, but I have no more room in the cavities to install 6 power amps into the body. I used some cheating with the amps. The amps are bridged, which would require 12 wires for the coils. It would be possible, but I wanted to save some wires for the on-body controls (in future). So, I used a common ground and capacitors. The thing is the amps have single power supply, +12V, so, at the outputs there's a 6V DC, which allows me to use electrolytic capacitors and a common ground. In fact, I use only half of the amp, which still provides me plenty of power. The system looks like a monster, but I want to experiment with software a lot, which was my initial goal. There's a lot of work, but there're some early results. First, I wanted to simulate a simplest overdrive/distortion. In fact, I use a cubic Bezier curve to simulate any transfer characteristic, symmetric or asymmetric. Just linear: http://antigrain.com/hex_project/sample_linear.png http://antigrain.com/hex_project/sample_linear.mp3 Exaggerated "warm tube sound" with the mag hexaphonic and very asymmetric transfer: http://antigrain.com/hex_project/sample_tube.png http://antigrain.com/hex_project/sample_tube.mp3 Then, a soft symmetric overdrive, without any post-filtering: http://antigrain.com/hex_project/sample_overdrive.png http://antigrain.com/hex_project/sample_overdrive.mp3 The yellow lines are not actual signals, they show what happens to the sine wave at -12, -6 and 0 dB. Note that even with a significant non-linearity, there's no intermodulation, so, even big chords remain very detailed. However, it's not good for solo, because any noise from other strings is very noticeable due to the huge gain at low signal levels. When the signal is pre-mixed in the pickup, non-linear distortion has a side effect that suppresses low level signals, which is perfect for solo. I have a solution that would allow you to play both, solo and chords without losing any details and without intermod, but I haven't implemented it yet. Now about the sustainer. It doesn't yet work as I want, I only tried if it works at all. In fact, for now I only simulated a simple hexaphonic sustainer solution with a monster system, but I'm sure, with the software and DSP I can do a lot of things that you can never achieve with circuits. Well, at least I can now sustain big chords. The sustainer completely confirmed my expectations. I can use it with a mag pickup, but the sustain is kinda weak, or, otherwise I get a huge direct E/M feedback. Exactly as expected. In future, with heavily shielded coils it will be possible to vastly improve it, but at this step it's good enough to use it with Graphtech saddle pickups. I can feed the coils with a lot of power, which gives me just crazy sustain. In fact you have to damp the strings heavily all the time. I also was amazed how many harmonics you can excite. Just a sample with distortion: http://antigrain.com/hex_project/sample_sus_overdrive.mp3 The same, with linear transfer: http://antigrain.com/hex_project/sample_sus_linear.mp3 (don't worry about the "dzzz" buzz, it's not because of any interference, it's just because I'm not a good guitar player and don't hold the strings well) So, here's my experiment: I hold my finger across the strings, not touching the frets at all (finger, not any metal), starting from the 18th fret and move it slowly to the nut. There're 3 stops: at 12th, 7th and 5th frets, which correspond to the major natural harmonics. The amazing thing is how many OTHER harmonics there are. http://antigrain.com/hex_project/sample_sus_move_finger.mp3 So, it drives the strings crazily, but for now it's totally out of control. First, I plan to dynamically adjust the level and the phase, to make it perfectly predictable and stable. Then, I want to synthesize the signal, that will provide me full control on the harmonics. Well, something like Moog Guitar, but way much better. This is my very ambitious plan, we will see, maybe it will be an epic fail, that's OK too, but it's so exciting to work on it. I also have a lot of other ideas with hexaphonic. For example, you can control certain strings with other strings. Say, you use the low E string to control the distortion by just its level, or the pitch. And so on. McSeem

Of course, it can be even 0.5 mm, and it will sound well, but you can't play hard. A typical recommendation is 2.5-3mm at 12th. I also set it lower, about 2mm. I was talking about the difference between a parallel fret board and sloped. Well, actually, the system is kind of self-regulating. When the string hits the frets, it restricts its amplitude, but still it's undesirable. If it happens at the very attack moment, it's OK (slapping), but it may happen later. The thing is the string vibration is polarized. Initially, the polarization is about parallel to the frets, which theoretically allows you to have a very low string setup. But the polarization is rotating. And when it gets perpendicular, the string may easily interfere with the frets. I'm sure you heard that -- you pluck the string, and after a while (1 second or so) it starts buzzing.

I'm sorry for being annoying, but I can easily prove a theoretical necessity of a compensated nut even with the zero fret. Straight physics and geometry. Firstly, no way! A parallel fret board wouldn't work at all. It would simply prevent the strings from vibrating (the sizes on picture are exaggerated for clarity). The amplitude of the fundamental frequency in the middle may easily be more than 1.2 mm. Then, about the geometry. If you remember the Pifagor theorem, you сan easily calculate the change of the length when fretting. Suppose, your string deflection is 1 mm along the whole fret board. Taking more or less realistic sizes, you add about 0.0005% to the string length, when you push the string in the middle. But with the same deflection distance at the 1st fret, you add about 0.005% - ten times more! That's why it's so hard to play with a high nut (not to mention bad tune), while in the middle you easily push the strings. So, the slope fretboard is essential! Then, if you calculate the extra length with a slope fretboard (assuming your favorite 0 fret), you will figure out, that from 1st to 12th frets you add almost exactly the same extra length and so, extra tense! Then, since you inevitably add tension along the entire fretboard, you want to compensate it. So, you adjust the bridge intonation. But! The bridge intonation almost DOES NOT affect the low frets! Just because the relative difference of the length is small (with and without saddle compensation). And, since different strings have different characteristics "delta-tense" to tune, the slot positions at the nut must be different. So, it proves at least a theoretical necessity of having a compensated nut. Although, the zero fret improves the tune, it's still theoretically necessary to compensate. I say, "at least theoretically", because I'm not really sure if there's any significant effect in practice. Probably not, otherwise a compensated nut would be already popular, as the saddles with intonation are. McSeem

Well, of course, the "zero fret" gives you pretty good tune. However, the nut does make sense in terms of better behavior with the tremolo bar -- the nut must be permanently lubricated, for example, Graphtech sells their special teflon nuts. Another reason is that, open strings produce stronger high harmonics, with bigger vibration amplitude near the next fret. It happens because strings are also vibrating lengthwise, and your finger dumps high harmonics, while the nut doesn't. So that, open strings with the "zero fret", or a nut adjusted exactly to the fret height, will tend to interfere with the 1st fret when you play hard, so that, you have to set the strings higher at the 12th fret to compensate it -- I did that experiment long time ago with an acoustic guitar. That's incorrect. Although, all frets have exactly the same height, a string fretted on 1st is higher than open (assuming the zero fret) a string, fretted on 12th is significantly higher at 13th -- just straight geometry. One more possible reason (maybe even the major one) - is that you can easily add extra 3-10 cents if you push your string stronger. And this effect is significant namely on the low frets, while the high frets can be compensated with the bridge. So that, you can potentially adjust the nut slots positions to compensate the average pressing strength. This is my rationale. Although, I agree, it doesn't make much sense for an average guitar player.

I don't have any serious goals on it. My first goal is to check whether or not it makes sense at all. That's a low priority project, so, before drilling and milling, I want to collect your opinions. Thanks a lot!!! Maybe one day I'll try it, but not sure. As for potential marketing, well, it's big question anyways. The musicians are conservative, and I totally realize that.

It's trivial -- with a precise strobe tuner. You set the bridge intonation first, then adjust the nut slots so that the first frets would be exactly in tune. If the 1st to 3rd frets are too sharp, you move the slot closer to the bridge. Long time ago I also filed and sanded an acoustic guitar neck to make a very low-profile fret board. I didn't notice ANYTHING bad about it, compared with the regular-height frets. So, why they make them so high?

Yes, there's a lot of sound processing. As I said, it's Guitar RIG 4, with a pre-set called "Arpeggio Delays". This is a screenshot: http://antigrain.com//hex_project/GuitarRig01.jpg But the combination with the sustainer coil sounds REALLY fancy.

www.elutherie.org - yeah, that's a monster design. It also requires a body-mounted tuner. My design, that I keep in mind, is way much simpler. Although, it requires some minimal woodworks. We will see. Yes, that's another way of adjusting the height. However, this method is less precise; your distances depend on the string height at the 12th fret. I just saw the recommendation I mentioned several times. Graphtech also recommends to do the same. I agree, standard gauges are too wide. I used stacked pieces of clear tape, measured with a fine caliper. Then you cut your gauges as you want. The "3rd fret" method makes sense, because it's more accurate. If your strings at 3rd fret slightly touch the 2nd, it means the nut height is exactly equal to the fret height. However, you do want your nut to be slightly higher than fretted strings to prevent from fret buzz on open strings when you play hard. High harmonics are stronger on open strings.

I assume Ive got you with my posts, I accidentally found a nylon spacer of 10 mm outer diameter and decided to make a bobbin. It's good, but it's really hard to lathe nylon with just a drill and files. Much harder than iron; nylon is very slippery. But I could wind a much nicer coil and with the 36 AWG wire it has 8.8 ohms DC resistance. So, how it looks: http://antigrain.com//hex_project/coils022.jpg http://antigrain.com//hex_project/coils023.jpg http://antigrain.com//hex_project/coils024.jpg http://antigrain.com//hex_project/coils025.jpg Then I tested it, also with the Guitar RIG and VERY fancy virtual devices, like that: http://antigrain.com//hex_project/GuitarRig01.jpg It sounds crazy and also drives the strings like mad. It also tends to produce a lot of squealing harmonics even on low strings. I could never imagine that you can so easily excite 10th or even 15th harmonic. And, no, there's no direct feedback, I used the saddle pickups. http://antigrain.com//hex_project/sus05.mp3 I fed the coil with the very same signal you hear. Of course it's now all random; no phase coherence, no rocket science, but it proves that some fancy processing may make sense. McSeem

Yes, there's a point. Sharper notes appear mostly because of too high nut. However, the proper nut adjustment is described here: http://www.projectguitar.com/tut/nuthigh.htm And even in this case, and with the proper neck adjustment you still have some sharpness, up to 3-5 cents on low frets. I even made the nut slightly lower than described, like 0.003". Another thing to consider is that, when you "deflect" your string at the nut (1st fret) you add more extra tension than when you do that in the middle, assuming the same height, say, 0.5 mm. you can take a calculator and recall the Pifagor theorem. That's pretty straightforward. And you can't compensate the intonation on low frets with the bridge. As I said you can achieve perfect tuning if you add one more fret and a spacer behind it. But in this case open strings will rattle. Besides, different strings have different tension parameters, this is why you have to set up the saddles differently. Say, unwound G string needs more compensation. Besides, the tune depends on the strength of pressing. If you press the string stronger, you add extra 3-10 cents. It will be also nice to compensate it for average pressing. Just straight physics. Well, I'm just trying to convince myself that it all makes sense. :-) Maybe it doesn't...

Yes, AFAIU, there's no way to tune buzzfeiten without a precise strobe, because you have to take care of special offsets. While an adjustable nut would require a precise strobe only to set the intonation itself, and then you can tune your guitar as usual. Well, any nut with individual compensation requires different positioning, that's why it looks ugly, but maybe it's just a matter of habits. Afterall, you can say the bridge with a strange saddle pattern also looks ugly. Instead, maybe this uneven "nut pattern" with individual screws may look super cool... to players who understand it :-) as an option, it's also possible to use some cover to hide it.

The buzzfeiten nut is not adjustable, basically they just shorten the first fret space and then compensate everything with the saddles. And, yes, the compensation is different for different guitars, but it mostly depends on the strings, so, when you decide to use different strings, you have to re-adjust the intonation. Both Earvana, and Buzzfeiten use just average compensation. Maybe that's good enough, but definitely not perfect.

Right, I just forgot about it. For alnico you have to use real plastic bobbins. BTW, it's also possible to lathe bobbins from a suitable plastic rod, using just my homemade lathe drill. It's more tricky, but quite possible. I tried to request for a quote at ablecoil.com (for hex pickups), it costs more than gold: The request: I need a number of self-supporting coils, with the following parameters (approx): Internal diameter: 3.2 mm External diameter: 8 mm Height: 12 mm DC Resistance: 400-500 ohms The response: The cost will be a $500.00 lot charge for the 10-20 coils, plus $1,000.00 for the tooling. Lead time will be 10 weeks. So, it only makes sense to order a 1000 or so. BTW, are there any places to make custom coils for a reasonable price? It's OK to DIY coils for experimentation, but becomes unreasonably tedious, when you come up with the parameters and want to make many copies for production. McSeem

Not necessarily, I may try to put it directly on the body. On the other hand, I have a feeling it's nice to have individually adjustable height. So, we will see. Maybe I'll use the neck HB space, maybe I'll mill another cavity. But first I'll try simplest ways. . That's OK, you can sound as negative as you want :-) I'm aware of this problem, and as I said, I'm going to use a lot of processing in software - pitch detection, automatic phase adjustment, depending on the string and the note, and so on. Not sure I can achieve anything, but I'll give it a try. And, of course, I can dynamically change the gain to start the sustain quicker, and keep it under control: if the amplitude is getting too high, I just reduce the gain and balance the signal - it's easy to do in software. In general, I noticed that a dynamic range compressor helps a lot and makes the sustain feedback almost perfectly stable. So, I'm going to use some compression too. It's even more complex. I'm going to try a very fancy response characteristic, an expander at low signal levels (or, even a threshold), and a compressor at high levels. It should prevent the strings from autoexcitation. With a simple, but powerful amp, the system, as a very high-Q one, can easily get out of control. And yes, I nearly burnt the coil. :-) McSeem

This is why I'm asking the question, because I'm not sure either. But assuming I can come up with a reliable design, I have a feeling it makes sense. Thanks for the article. They also use a kind of a pre-fixed positioning, and BTW, it looks very different from this ones: http://www.earvana.com/products.htm In particular because of a wound G string, which proves that the adjustment heavily depends on the strings. McSeem

Hi, While tuning the saddle intonation I've noticed the first frets sound too sharp, up to +6 cents, while the tune is perfect at the 12th fret. I use a very accurate software strobe tuner: http://www.tbstrobetuner.com/ It's quite understandable: the nut is supposed to be slightly higher. That is, the distance between the string and the 1st fret strip must be a bit bigger than the distance on the 2nd fret, when you hold the 1st one (yes, I've read an article about the nut tuning). So that, the difference in string tension is higher on the first frets, while on the last frets it's compensated with the saddles. Long time ago I tried to replace the nut with a metal strip and a spacer behind it, and then polished it together with the other strips, as if your fret board was extended. I did everything very carefully, and the intonation was perfect. But there was another problem: open stings interfered with the first strips! Why? The string height was the same and it was very accurate. The answer is: open strings sound brighter, with stronger high harmonics. Your finger damps high harmonics, even though, the string lies on the metal strip! The string is also vibrating lengthwise, so, your finger removes high harmonic energy. The difference is very slight, but it's enough to make open strings interfere with the strips. This is why the nut must be slightly higher, or, otherwise there must be a damper above the nut that imitates your fingers. But I assume the guitar players do not appreciate it, because it's nice to have a bright sound of open strings. As the result, first frets sound sharper and you can't compensate it with the saddle intonation. The solution is to have an adjustable nut. Not just adjustable height, but most of all, individually adjustable positions! I could only find these ones: http://www.earvana.com/products.htm But they all are pre-fixed and adjusted for average strings, while there can be a huge difference, especially on light and heavy strings. Wound and unwound G string also makes a huge difference. So that I came up with an idea of fully adjustable nut. It's not about the height, it's only about the position. I have a couple of versions of the design in my mind, and my main question is: in your opinion, how cool would it be to have this kind of a nut? Is it worth giving a try, or it doesn't make much sense? McSeem

Hi, a couple of weeks ago, while waiting for the parts, I decided to make a sustainer coil. Since my goal is to make a truly hexaphonic sustsainer, I'll wind coils for each string. 6 of them, and, maybe, even 12. So, this is also my early experiment, I came up with just a single coil, that works perfectly and has proper parameters. I took pole pieces from a cheap humbucker, but you can order them for a much lower price. You can also use alnico magnet poles for that. The reason I used just metal is to tweak the permanent magnetic bias, using different magnets. But if alnico works well, as I assume, it's even better. So, you take a pole piece, you take a drill like that: http://antigrain.com/hex_project/coils001.jpg Well, there's a trick. You don't want your drill to rotate at its full speed. You need to adjust it. And you also don't want to hold your finger all the time to keep proper speed. So, what you gonna do? It's called TRIZ, http://en.wikipedia.org/wiki/TRIZ. I'm a big fan of TRIZ. I have a 300W lamp with a dimmer, and I just rewired two plugs, so that, they connect the drill and the lamp in series: http://antigrain.com/hex_project/coils002.jpg http://antigrain.com/hex_project/coils003.jpg So, with the dimmer you can perfectly control the RPM and use your drill to wind the coils. And you also use it as a lathe machine! Next, you saw the pole piece in the following way: http://antigrain.com/hex_project/coils004.jpg http://antigrain.com/hex_project/coils005.jpg http://antigrain.com/hex_project/coils006.jpg The tricky part is to accurately measure the height. But you can make a coil of any height you want, even 3mm or so. Then you take a suitable plastic pack to make the washers. You cut them, stack, and drill a slightly smaller hole like this: http://antigrain.com/hex_project/coils007.jpg http://antigrain.com/hex_project/coils008.jpg You also cut it of course, so that the washers would be approximately round. An octagonal shape is OK. Then you slightly cut the hole and put them on: http://antigrain.com/hex_project/coils009.jpg http://antigrain.com/hex_project/coils010.jpg And, of course glue them. I use Elmers Ultimate Hight Performance glue. Make sure to use vinyl gloves, because this glue is really bad for your fingers. The tricky part is to set up the washers as perpendicular as possible. I use a toothpick, while rotating the drill: http://antigrain.com/hex_project/coils011.jpg And then you gently remove it trying not to touch the washers. You can use a magnet for that, with a combination of plastic tweezers. http://antigrain.com/hex_project/coils012.jpg The next day, when the glue dries, it's time to lathe the plastic and wind the coil. http://antigrain.com/hex_project/coils013.jpg It can be dangerous for your fingers and your eyes! So, do it at your own risk! Eyeglasses are MANDATORY! After that, you insulate the core with a piece of plumbing seal tape. It's very thin and provides enough insulation: http://antigrain.com/hex_project/coils014.jpg http://antigrain.com/hex_project/coils015.jpg You gently tighten it with a piece of plastic, while the bobbin is rotating. Then I use the very same drill to wind the coil: http://antigrain.com/hex_project/coils016.jpg http://antigrain.com/hex_project/coils017.jpg http://antigrain.com/hex_project/coils018.jpg Well, I admit it could be more accurate, but as the practice shows, it works perfectly. The modern magnet wires have very reliable insulation, so, don't worry too much about potential shortcuts and the accuracy. The only problem is how to count the winds. With the drill I have no good solution for that. If you wind it manually, you can use a simple step counter, put on your hand. It works perfectly. But with the drill it's not that simple. I didn't count the winds at all. Just I assume that 10-15% of difference is OK. I'm anyways going to adjust the response with the amp gain, since I will have 6 individual channels. So, more or less the same physical size is good enough to me. I used 36 AWG wire, and with this size, 6mm polepiece, about 4 mm space between the washers, and 10mm outer diameter, I've got almost exactly 10 ohms of DC resistance. It drives the string perfectly at about 4-5 mm distance. Well, after attaching the magnet you will have about 8 mm height, which is still OK in my case. I still have about 1.5 mm extra for the underneath binding, without having to mill a cavity. But once again, with this homemade technology you can wind coils of any size, 3 mm if you want. You can also use thiner plastic. The only thing is left is to cut it off. I use the very same drill and the very same metal saw: http://antigrain.com/hex_project/coils019.jpg http://antigrain.com/hex_project/coils020.jpg http://antigrain.com/hex_project/coils021.jpg It may look very hard to make, but if you make a number of coils, at least 6 of them, it takes about an hour per coil. Well, plus the time to let the glue dry, of course. So, enjoy with homemade engineering, as I do! My girlfriend does not appreciate it, though... :-)

Thanks! My major problem is what to do next. :-) Well, I feel a lot of potentials here, but I also need help from real guitar players. What kind of sounds musicians would like to try? What kind of controls? And so on. I can't be a good player and an engineer at the same time. I'm an engineer. So, any cooperation with musicians is highly appreciated. I aim for new kind of sounds and fancy processing. I know, for example, the great Guitar Rig software, but it mostly simulates different devices, and it's not hexaphonic. Well, you can run 6 instances of Guitar Rig, and it works, but routing and configuring becomes a nightmare. The coils I used for the mag hex pickup are, Paul's Type 1 with step-up transformers: http://ubertar.com/hexaphonic/ It's some kind of industrial coils from old telephony. They have about 314 ohms DC resistance and about 10 mv output. Paul also makes other type of coils. His original design is very simple, and although he claims the coils are individually adjustable, in practice it's useless, because after you install the pickup you can't adjust the height, so, you have to do that in advance, which is tricky and inaccurate. And the adjustment is critical to balance the signal levels from each coil. So, I designed my own binding. It looks nice, but it's very impractical in assembly. The design could be way much simpler. And yes, it takes the entire humbucker space because I wanted to have adjustable spacing too, so that, you can place the poles very accurately, exactly under the strings. As for the sustainers, I plan to design a hex driver, but first I want to handle at least the simplest processing via the ASIO SDK (pass-through, and multichannel recording, at least). I only did some very early experiments, and it proves that the latency is not a problem, although, I expect tricky time when I tweak the phase of the signal. So, the first experiment I did is with the HB signal, passed through the Guitar RIG, with a compressor, slight overdrive, reverb, tremolo, and something else. And... I recorded it via a USB Samson Go Mic. :-) So, there's a lot of hum and noise and the quality is poor. There's nothing musical in those samples, mostly random dissonances. It's because my right hand was busy with the driver coil, and the guitar was on the carpet. Very tricky to produce anything musical. http://antigrain.com/hex_project/sus01.mp3 The second experiment is also with the Guitar RIG and very heavy distortion on the driver coil. But I figured out how to use the Guitar RIG loop machine to record the signals -- rather poor tool, but still can be used. http://antigrain.com/hex_project/sus02.mp3 - signal recorded from the Guitar RIG input http://antigrain.com/hex_project/sus04.mp3 - signal recorded from the Guitar RIG output (lower the volume!) Another experiment is also interesting. I run Audacity and generated very harsh E2 tone -- sawtooth. Then I powered the coil with this signal through a separate amp, and recorded the signal from the saddle pickups. As I expected, the only low frequency, E2, 82.4 Hz perfectly drives all E notes, from E2 to E5. http://antigrain.com/hex_project/sus03.mp3 Which proves that it's not obligatory to power the drivers with the original sound, instead, you can achieve interesting results, if you drive the coils with an octave lower signal, heavily distorted, namely a sawtooth. And, yes, the phase was arbitrary and totally out of sync, because the signal was artificially generated. It will be very interesting to tweak the phase. McSeem

Hi, This is my hexaphonic guitar project. It's totally for experiments with new sounds and sound processing, not for any kind of a performance. I bought quite a budget guitar in case I break it I wouldn't cry. But eventually I installed a lot of stuff in it, I can't even say how much money it cost. Maybe $1K or so, but it doesn't matter, I do it just for fun. I'm not a musician, I'm an engineer and a computer programmer. But I fond of music and I truly enjoy building unusual guitars. So that, I started my hexaphonic project. Long time ago I even made my own guitar strings, metal and nylon, some of them had double winding! It's not that hard by the way, to make strings. Anyways, what is this project about? - I'm not sure. The minimal goal is to play with it and just to throw it out when I'm bored. The maximal goal is to create some hardware/software system where you can configure all 6 channels independently and can control the sound of each string with a kind of a MIDI controller connected via USB. Or, even on-body pots, or even with something wireless, in a way termenvox works. But it's not MIDI! Everything must be truly analog. So that, for the beginning I ordered a hexaphonic magnetic pickup from Paul Ribinstein, http://www.ubertar.com/ but I wasn't satisfied with its design. So that, I used the coils and designed my own fully adjustable pickup. It's not practical, and I made a lot of mistakes that I had to fix on demand, so that, the design is very fancy and it took me a lot of time. Nevermind, I now know how to do that simpler. The major problem was the separation of the channels, the crosstalk. I estimate it as about 24-28db, which is good for stereo panning, but might be problematic if you want to use DSP with pitch detection. Then I learned about Graphtech piezo pickups, they produce very nice acoustic sound and have excellent channel separation. So, I eventually ordered 6 Graphtech preamps with their saddle pickups, of course. This is how it looks (I don't use "IMG" tags in order not to pollute the message to much, so, just click it) http://antigrain.com/hex_project/hex_project01.jpg http://antigrain.com/hex_project/hex_project02.jpg Wiring nightmare: http://antigrain.com/hex_project/hex_project03.jpg http://antigrain.com/hex_project/hex_project04.jpg http://antigrain.com/hex_project/hex_project05.jpg Yes, I found 6-pole potentiometers (actually, 8-pole ones) so that, I can control the volume of the magnetic pickup and the saddles. Each Graphtech preamp has two channels - one with a fancy filter for the piezo saddles, and the other - just straight for magnetic. The preamps are very fine and based on a precision low power 2022C chip, so that, they don't color the mag sound at all. This is the preamp with a commutation cable: http://antigrain.com/hex_project/hex_project06.jpg Six preamps and the commutation board: http://antigrain.com/hex_project/hex_project07.jpg http://antigrain.com/hex_project/hex_project08.jpg Assembled and connected: http://antigrain.com/hex_project/hex_project09.jpg On the right there's a box with transformers for the magnetic hex pickup. The direct voltage is about 10 mv, which it too low, while after the transformers the signal is about the typical instrument level, that is, 100-1000 mv. Yes, I could tweak it with the preamp gain, but it would be too tricky to experiment with the assembled preamps. It would require some reverse engineering... too much hassle. Pickups: http://antigrain.com/hex_project/hex_project10.jpg http://antigrain.com/hex_project/hex_project11.jpg Added one more pot for the neck humbucker. http://antigrain.com/hex_project/hex_project12.jpg http://antigrain.com/hex_project/hex_project13.jpg The workspace: http://antigrain.com/hex_project/hex_project14.jpg And, of course, I'm going to experiment with sustainer drivers. I made coils, while waiting for the parts: http://antigrain.com/hex_project/hex_project15.jpg http://antigrain.com/hex_project/coils14.jpg Of course, I made a lot of mistakes. The major one is the 13-pin female connector. I used a round one, without the locker, which was easier to install. So that, it doesn't hold the cable well. But it's still OK for experiments. I should have ordered the original ones that Roland use. But just laziness. Of course, there's also a breakout box with a battery. I use a budget MOTU 8pre interface, and I also bought a cheap 8-channel Behringer ADAT converter to extend the ins and outs, when I play with a sustainer. So, how it sounds. Once again, I'm not a good guitar player, I can play some very simple basic chords, long time ago I used to play an acoustic guitar, so, the narrow string spacing is very tricky to me. The sound is not processed anyhow, just to estimate. It's a very-very beginning, so, don't expect anything spectacular. I just panned hex pickups left-right, but it can potentially be used for very spatial sound, 5.1 or even 7.1. Regular neck humbucker (very fat): http://antigrain.com/hex_project/sample01_neckHB.mp3 http://antigrain.com/hex_project/sample02_neckHB.mp3 Hex magnetic: http://antigrain.com/hex_project/sample01_hexmag.mp3 http://antigrain.com/hex_project/sample02_hexmag.mp3 Graphtech Ghost hex: http://antigrain.com/hex_project/sample01_hexghost.mp3 http://antigrain.com/hex_project/sample02_hexghost.mp3 Mixed all together: http://antigrain.com/hex_project/sample01_mix.mp3 http://antigrain.com/hex_project/sample02_mix.mp3 So that, if someone is interested in this project, I'll be very happy to cooperate, share the ideas, and so on. And, most of all, if someone in U.S., Washington DC area wants to test it, play, record some demo samples, you are very welcome to visit my place. I'm in Maryland, College Park. Just leave a personal message. McSeem