wow gibsons new toy!

[disto]

i know many people here are into analouge and more "vintage" efftects but...

wow!!! http://www.record-producer.com/learn.cfm?a=2281

a digital guitar the possibilities are amazing!


almost makes you wonder whats happening next i love the bit

Imagine using a different effect on each string! You can't say that's been done before

the gibson sites interesting to //www.gibsonmagic.com/


what will come next...

[d95err]

Looks like it's mainly a regular guitar with a built in A/D converter for each string. Line 6 has gone a lot further with the Variax, especially when used with the Workbench and/or the software which enables you to model your own guitar.

[d95err]

...on further analysis, perhaps Gibson is the cutting edge technology. The Line 6 Variax is a proprietary platform with a single set of features.

The Gibson digital guitar appears to be the founding of an open system where multiple developers can produce their own software to interact with the digital guitar. Now it gets really advanced...

[SeanCostello]

Hmmm...when did Roland first come out with their hexaphonic pickups with hex fuzz for their guitar synths? 1978 or so?

I think the most interesting part about this guitar is the Ethernet connection. Roland has been making proprietary interfaces for their hex pickups for years, but presumably there is a published protocol for getting the data from the Ethernet packets, and splitting it into the individual signals.

It will be interesting to see if this guitar goes anywhere besides Wired magazine and the like. Gibson and CNMAT (the computer music center at UC Berkeley) have had a connection for years, and this seems to be one of the first tangible things to emerge from that research. I wonder if such a guitar will play well with the standard Gibson audience. Plus, I would presume that the cost is rather high (it looks like an American made Les Paul), while the competition is a relatively inexpensive hex pickup and effects system/guitar synth from Roland that can fit on various guitars.

Sean Costello

[RandomRedLetters]

Honestly I think this guitar will be completely cost prohibited to most of the market because its already a les paul (probably a studio, but thats still a $1,000 guitar) plus this technology (which Im sure they'll charge a good amount for)

I think that if they really wanted to bring the price down to compete with Line 6's Variax (models start at $500) then they might have to at least have an entry level model that resembles a les paul jr or an sg.

[disto]

:oops:  ok seemed a kind of new idea to me i wasnt aware of any such systems i knew people had tried making midi type guitars but they had not caught on...

it seems like an amazing way forward though to me anyway

[lovekraft0]

You know, it might not be a bad idea, in the hands of just about anybody but Gibson - their corporate record of utter failure in designing and marketing any new(ish) technology will probably insure this project remains a footnote in industry history. C'mon, these guys have had both Bob Moog and Bill Lawrence on their R&D staff, and the best we got out of it was the RD series and the L6S.  :roll:

Does anybody really like the idea of onboard A/D converters, battery powered, and designed by a Gibson engineer?  

[Paul Marossy]

C'mon, these guys have had both Bob Moog and Bill Lawrence on their R&D staff, and the best we got out of it was the RD series and the L6S.

The Lab Series L5 is a great amp, IMO (for a SS amp).  :wink: I hear what you're saying, though. At least Fender is always trying to be in the middle of current events - I could mention other manufacturers as well, but I'll stop here...  :wink:

[Peter Snowberg]

A guitar with A/D on-board and digital transport offers some rather amazing possibilities.

The way Gibson did this (MAGIC) requires an FPGA and a really fast CPU. It's not DIYable.

Is there interest in a DIY digital guitar interface?

I have one all designed up using off-the-shelf parts that offers "us" most of the benefits of this technology, without the high cost or the FPGA / ARM requirement.

This dovetails with my DSP project so I though I would ask now.

Mine is based on the Alesis light-pipe interface format, so the guitar would have power on-board with a lithium polymer battery and the "cable" would be a piece of fiber-optic. No ground loops, but no ground either. :?

[tommy.genes]

I think the most interesting part about this guitar is the Ethernet connection. Roland has been making proprietary interfaces for their hex pickups for years, but presumably there is a published protocol for getting the data from the Ethernet packets, and splitting it into the individual signals.

Well I've read a little bit of the MAGIC specs, and they do say their technology is based upon IEEE 802.3, which is the standards document that defines Ethernet. They also mention 3Com and Xylinx in the above-linked article, so that would also indicate that they are committed to Ethernet compatiblility (3Com founder Robert Metcalfe invented Ethernet.)

Still, the MAGIC spec hedges a little bit, clearly saying only that it is compatible with Ethernet at the physical layer (Layer 1 for you OSI geeks). To be fully compatible with off-the-shelf Ethernet hardware, it would have to be compatible at the data link layer (Layer 2) as well. The MAGIC spec notes this, saying that to be compatible with standard Ethernet hardware, MAGIC MAC addresses have to be unique worldwide. That actually shouldn't be too hard to do since the IEEE manages MAC address ranges, and I don't think they're all gone yet (there are 2^48 of them, which is over 200 trillion). So I find it interesting that they state that caveat.

And then at Layers 3 and 4, they appear to forego the more common TCP/IP protocols for their own MCP (Magic Control Protocol). They may have been forced to do that since TCP/IP isn't known for its ability to handle real-time streaming media, but I wonder if they'll regret that decision the first time somebody wants to record their guitar tracks over the (capital-I) Internet. I guess you could handle that well enough with routers, but it's still a thing that makes me go hmmmm...

But the MAIN thing that makes me suspicious is the diagram at the bottom of THIS PAGE, where they show the guitar plugged into the breakout box, and then six analog channels patched into the sound card of a computer. Hello!? If you have an Ethernet port on your guitar, why do you have to go through extra D/A and A/D conversions to get it into a computer? You should be able to just plug into the Ethernet port right?

Ahh, whew. I'm OK now...

-- T. G. --

[SonicVI]

And knowing Gibson it probably only costs $8,000!

[Paul Marossy]

And knowing Gibson it probably only costs $8,000!

Yeah, that's probably disgustingly close to the truth!  :roll:

[StalfoS]

They really missed the mark on this one...  Ethernet cables are sooo 90's!  Where's the 802.11 wireless interface!

[DavidS]

Peter, I'll bet there would be some interest in the lightpipe output!

That's something I always wanted to build. I looked at those chips several years ago (when they still had the Alesis name on them) and spent some time fantasizing about it, reading and re-reading the datasheet, and realized I had NO idea what I was doing with designing something like that.

You could get pretty elaborate with it too, with 8 channels to work with...

[lovekraft0]

Is there interest in a DIY digital guitar interface?

Oh, absolutely! Lightpipe sounds like a really good idea! I'd definitely like to see somebody pursue that.

Actually, the MaGIC standard looks pretty good, but I simply can't see Gibson ever successfully developing and marketing an open-source project, and anything proprietary will quickly get orphaned because they can't/won't keep up with advances in the technology. I guess that's what happens when all your best known products are all at least 30 years old. BTW, haven't they been bragging about this thing since 2003? I seem to remember some early adverts about demo kiosks being set up around the US sometime in the first or second quarter of last year!

[tommy.genes]

Ethernet cables are sooo 90's!  Where's the 802.11 wireless interface!

Or even 802.16 WiMax, so that you could be recording your tracks while on the train on the way to work!

BTW, haven't they been bragging about this thing since 2003?

Oh, even earlier than that. The original spec was released at the 107th AES show in 1999. The technology was originally called something like GMCS, which Gibson actually encouraged people to pronounce "gimicks." Apparently someone in marketing took notice and said, "you know, maybe we shouldn't be telling everyone that our wonderful new technology is nothing more than a gimick."

-- T. G. --

[Peter Snowberg]

Again... this is NOT Ethernet. Ethernet couldn't pull this off.

MaGIC uses the same physical layer, but that's where it ends. You could plumb a MaGIC signal using an Ethernet hub, but not with a switch. The switch will just isolate the source port thinking its looking at noise.

MaGIC is a little more like a super high speed synchronous serial connection with complex framing. The 802.11 physical layer is used because it's the right speed for the physical transport task and piggybacking on another standard means that parts like PHY interfaces, isolation transformers, and filter magnetics are already in huge supply and CHEAP. MaGIC bears a lot of similarity to the standard for controlling digital dimmers and lighting gear (DMX-512). Both standards are just digitally multiplexed data inside frames.

If you look at the eval board, the legend still says GMICS (Gibson Musical Instrument Communications Standard).

One of the original authors of the spec is the guy I hold personally responsible for addicting me to DSP back in 1987.

It's a VERY cool idea and I had plans to commercialize it at one point.... but somebody has to buy some gear first and it needs to get some momentum in the market place..... otherwise why not just use a cheaper technology with the same benefits?

I built a Dan Armstrong 12-way strat switch out of some guitar magazine in the mid 80s for a friend of my dad. There was a bug in the illustration that I was very proud to figure out and correct. I got an electric shortly after, adding the same switching to mine and I added a phase switch for the the neck pickup, bringing the number of switching positions up to 19. In those 19 were the original 5 plus 14 NEW SOUNDS!!!! :mrgreen:

That got me hooked on wiring guitars and exotic pickup wiring and mixing. It was always a pain to find the "ideal" rotary switch or ON-ON-ON switches in the same finish as the rest of the guitar. Also, going between two different sounds would require up to 4 switch togglings which isn't practical for live use, but those tones are locked in there just asking to be let out. Forward to today.....

The Alesis Light-Pipe interface gives you 8 channels @ 48K to play around with plus MIDI on the side. If you had a hexaphonic pickup in one position and two single coils (or any setup reduced via switching to 2 wires) in the others.... there's your 8 channels.

Two humbuckers? OK.... 4 channels required. Less power consumption.

The front panel controls can just be wired to digital inputs and A/D inputs on a microcontroller which then sends out the status periodically via MIDI over the light-pipe.

The interface could be a regular TOSLINK connector so you could get patch cables at RadioShack or CircuitCity. You could also use bulk fiber connections which use a twist to fasten nut system for a really secure connection. Finally there is always twisted pair too.

For me, it's about the technology and the benefits... not the media or signaling format interoperability.

On the receiver side you would have the Alesis receiver chip with its 4 stereo outputs feeding directly into an Alesis AL3102 DSP which would then mix the signals down to one or two channels. The mixed signal might include one for tone and one for envelope control in the effects down the line, or you might want one pickup combo to drive clean effects and then another to get some screaming harmonics with distortion which are then mixed together later in the chain. Imagine a stomp-switch switching your guitar and effects at the same time . Each of the 8 input channels can have unique EQ and the results can be additively or subtractively mixed for output, which would then possibly feed directly into the digital input of a DIY DSP multi-effects processor.

[JimRayden]

Again... this is NOT Ethernet. Ethernet couldn't pull this off.

MaGIC uses the same physical layer, but that's where it ends. You could plumb a MaGIC signal using an Ethernet hub, but not with a switch. The switch will just isolate the source port thinking its looking at noise.

MaGIC is a little more like a super high speed synchronous serial connection with complex framing. The 802.11 physical layer is used because it's the right speed for the physical transport task and piggybacking on another standard means that parts like PHY interfaces, isolation transformers, and filter magnetics are already in huge supply and CHEAP. MaGIC bears a lot of similarity to the standard for controlling digital dimmers and lighting gear (DMX-512). Both standards are just digitally multiplexed data inside frames.

If you look at the eval board, the legend still says GMICS (Gibson Musical Instrument Communications Standard).

One of the original authors of the spec is the guy I hold personally responsible for addicting me to DSP back in 1987.

It's a VERY cool idea and I had plans to commercialize it at one point.... but somebody has to buy some gear first and it needs to get some momentum in the market place..... otherwise why not just use a cheaper technology with the same benefits?

I built a Dan Armstrong 12-way strat switch out of some guitar magazine in the mid 80s for a friend of my dad. There was a bug in the illustration that I was very proud to figure out and correct. I got an electric shortly after, adding the same switching to mine and I added a phase switch for the the neck pickup, bringing the number of switching positions up to 19. In those 19 were the original 5 plus 14 NEW SOUNDS!!!! :mrgreen:

That got me hooked on wiring guitars and exotic pickup wiring and mixing. It was always a pain to find the "ideal" rotary switch or ON-ON-ON switches in the same finish as the rest of the guitar. Also, going between two different sounds would require up to 4 switch togglings which isn't practical for live use, but those tones are locked in there just asking to be let out. Forward to today.....

The Alesis Light-Pipe interface gives you 8 channels @ 48K to play around with plus MIDI on the side. If you had a hexaphonic pickup in one position and two single coils (or any setup reduced via switching to 2 wires) in the others.... there's your 8 channels.

Two humbuckers? OK.... 4 channels required. Less power consumption.

The front panel controls can just be wired to digital inputs and A/D inputs on a microcontroller which then sends out the status periodically via MIDI over the light-pipe.

The interface could be a regular TOSLINK connector so you could get patch cables at RadioShack or CircuitCity. You could also use bulk fiber connections which use a twist to fasten nut system for a really secure connection. Finally there is always twisted pair too.

For me, it's about the technology and the benefits... not the media or signaling format interoperability.

On the receiver side you would have the Alesis receiver chip with its 4 stereo outputs feeding directly into an Alesis AL3102 DSP which would then mix the signals down to one or two channels. The mixed signal might include one for tone and one for envelope control in the effects down the line, or you might want one pickup combo to drive clean effects and then another to get some screaming harmonics with distortion which are then mixed together later in the chain. Imagine a stomp-switch switching your guitar and effects at the same time . Each of the 8 input channels can have unique EQ and the results can be additively or subtractively mixed for output, which would then possibly feed directly into the digital input of a DIY DSP multi-effects processor.

At least 50% of the words in this post are NOT rock n' roll!

Digital is evil.  :?

----------
Jimbo

[Peter Snowberg]

Digital is not inherently evil......

Products designed by MBAs are aother story.

MBAs have given digital a bad name.

Digital is capable of LOTS of things that analog is not.

When done at high quality, the only place that digital does not leave analog in the dust is in creating distortion.

When done at low quality, yes.... digital is its own "flavor" of effect, and its not the best flavor for most things.

[lovekraft0]

When done at high quality, the only place that digital does not leave analog in the dust is in creating distortion.

I don't think I've ever seen a more telling assessment of the current state of audio technology - mind if I quote you? Oh, wait, I guess I just did!  :oops: