Slashdot Mirror
More Channels for The Digital Musician
syrupdude writes "For those of us out there who love making bad music, Harmony Central has a story about a new digital wiring scheme from Gibson called GMICS, which uses standard cat5 cable to deliver 16 channels of 32 bits at 96kHz, or 8 channels at 192kHz. Definitely cool, but apparently not open. "
Isn't there a (short) practical limit to 1394 cable lengths that would render them inappropriate in this application?
Essentially, what's been done here is to lay a time-division multiplexed audio-data stream on a protocol designed for a shared medium, that, in this application, well, isn't.
That's O.K.: Ethernet hardware is rather cheap, and Cat5 cabling is ubiquitous. I've often thought of designing something similar for distributing uncompresed digital audio throughout my home, though IP-based so routers could handle it.
There's a couple problems with this approach, though:
1) Combining async and sync traffic is problematic. Here, a dedicated data channel has been piggybacked on the syncronous packet stream. ATM works pretty much the same way, except allows for dunamic allocation of bandwidth end to end.
2) Notice that there is one master clock source. This is fine, if you're sending data to one amp (which has a stable master clock, and so doesn't have to endure arrival sample jitter. But what if you have amps all over the place? If one is a master clock source, the others will suffer sample jitter. Furthermore, this makes combination of audio packets on different segments problematic.
You'd need a switched star arrangement if you have multiple sources and multiple destinations. Furthermore, a single source could only talk to a single destination at a time over a single segment. Scratch the idea of an autio server.
This is why most devices in the technical spec are daisy-chained, by the way.
3) Finally, regardless of how much audio traffic is on the link, each receiver has to work almost at wire speed to process each packet in time.
IOW, you have to be prepared to send or receive 16 audio channels regardless of how many you need for your end to end application.
I do wonder, though, if ATM (which can also run over Cat5 cabling) wouldn't have been a better choice: there's even a low-speed ATM-25 (25 Mb/s) variant.
In Liberty, Rene
It's not compatible with 100Mbit Ethernet - they use the physical and electrical layers from the specification (sensible - there are a lot of cheap, readily-available components that deal with these), but not the data link layer. If you plug GMICS equipment into an Ethernet hub or network card, nothing will happen - the hub won't see a link heartbeat.
It's easy to imagine a device that has analogue and MIDI connectors on one side and a GMICS type A connector on the other side. It could be powered by the 'phantom power', and would make it trivial to interface an analogue instrument and existing effects boxes to a GMICS system.
(Incidentally, I think they are abusing the term 'phantom power' a little in the specification. It usually refers to power sent over an audio lead using the same conductor as the audio signal; at both ends the DC power is filtered out leaving the AC audio signal. The GMICS 'phantom power' is just sent using the four otherwise-unused conductors in a Category 5 cable.)
The raw signal from the various instruments will pass through several layers of processing: mixing, filtering, effects processing, etc. At each of these stages distortion will be introduced (through rounding errors, etc.) - it's better to start out at high resolution and sample rate, have the distortion be small at each stage, and only reduce the resolution and sample rate for the final medium than it is to use the bare minimum rate all the way through the system and suffer larger (probably audible) distortions.
I can see where this sort of thing would be very useful for keyboards and synths, effects modules, mixers, and so on, but I'm not quite sure what it buys someone who is playing a guitar. Good guitar amps are made using tubes, and the response of a good tube amp is totally dependent on the electrical qualities of the guitar, so cleaning it up digitally is sort of a waste of effort since you're killing the whole point of having a tube amp. Any use of this interface in a guitar is going to need a an analog channel to carry the signal out the good old fashioned way as well. Although the prospect of some sort of traditional like interface ( a dial) on a guitar that is wired to digital effects on the amp/mixer/preamp is very interesting.
-Rich
This is indeed a problem. Linux could, at least, do with something like OMS..
:)
Since Gibson bought opcode a while back, you know who to whine at
Seriously though, audio is the main reason I keep my windows install working.. Until I can get a copy of Rebirth for linux, a decent sequencer, patch librarian, audio and MIDI control etc, I can't ditch it alltogether...
What's worse that that drivers for soundcards of a higher quality than a basic SB16 are virtually impossible to find. Sure, there's an amazingly glitchy and unreliable sblive driver.. but what to do, if you want to drive a creamware pulsar, SW1000XG or an Event Gina?
One size truly does not fit all- in this case, linux is a lousy fit. I'll use it for what it's good at, for now...
While on the topic...now all I need is some professional quality sound editing tool's for LINUX. Anybody have any sugesstions? I've used Cakewalk for years and I am sick of it. As soon as I find this solution, my last Windows box will be gone!
:)
AFAIK, you'll find links to the gammut of linux sound s/w here
The short answer is that it all pretty much sucks. Some suck less than others, of course. But if you hate Cakewalk, you are really going to hate this mess of s/w. I've used Rosegarden a bit, and it seems to be the most complete editor, but even still, it's not all that great.
Happy Hunting!
Arrgh, I hate Cakewalk, too. I've never heard of anyone actually using it professionally, just hobbyist stuff.
Here's a nice list of sound apps for Linux:
http://www.4front-tech.com/ossapps.html.
There's some really nice stuff there.
I hope Gibson doesn't expect huge licensing fees for this. Domination by one company often hurts acceptance. Look at Yamaha's XG and Roland's GS. Have you ever seen an XG synth NOT made by Yamaha?
/me shivers, thinking of the hard disk space that audio will take up.
192 kHz * 8 channels * 32 bits per sample / 60 seconds per minute / 8 bits per byte = 100 MB per minute.
I wouldn't bother adding compression. I don't think pro audio guys would go for it.
--
"I was a fool to think I could dream as a normal man."
B. B. Buick
Why do we need another standard? There is already a way to send uncompressed audio and MIDI information over 1394, as well as video and everything else.
Yamaha is the biggest proponent of the audio over 1394 standards. And they're a much bigger player in the digital music market than Gibson.
The timing might be slightly better in the Gibson spec, though. 1394 is great for stereo signals and would work well for source signals, but the timing is not quite good enough to seperate a mixed signal. You need picosecond resolution for that. In other words, 1394 is good enough to run signals from the mics to the mixing board, and from the mixing board to the amplifiers, but you wouldn't want to run 1394 from the amplifier to the speakers. (well you could, but some audiophiles might complain)
1394 is an open standard that has been in public discussion for a long time, and promises to interconnect a large variety of devices.
I don't think that the Gibson spec will be any cheaper than 1394. A 100Mb Ethernet PHY is cheaper than a 400Mb 1394 PHY, but the data/link IC is as yet unknown....
I have had to do some research for work on digital audio and the best 16 channel (32 with a sister board) part I could find was at http://www.frontierdesign.com. The lack of Linux support with this kind of (very sweet) high-end hardware is frustrating.
Here are a few reasons why:
Already supported by a number of manufacturers including QSC, Rane, Peavey, Crest, Level Control Systems, Eastern Acoustic Works and Crown.
Uses standard Ethernet - uses off the shelf switches and hubs
Each 100m link supports 64 channels of audio
A single switch ( with enough backplane bandwith ) can support upwards of 600 audio channels
Can coexist with regular ethernet data so you can play Q3A on the same network
I am obviously biased ( check the sig below ) but what the hell...
-John
--
John Dunn
Peak Audio, Inc
According to the press release:
Our vision is to embrace and extend the capabilities of the guitar - and all musical instruments ...
Hmm, where have I heard the phrase embrace and extend before?
First, to address this issue:
:)
Definitely cool, but apparently not open.
The specs are there on the GMICS site in pdf format. How open do you want? This is definitely better than the MIDI standard. Sure you can get the MIDI specs, but the the MIDI Manufacturers Association (MMA), who maintain them, expect you to pay 40 or 50 bux for them.
Having put that aside, I've only skimmed the specs, I haven't got a complete view of how this all works, but it appears that it will solve a number of problems that many of us musicians currently deal with.
[0] MIDI works and is useful, but it's getting old and crispy. Connecting MIDI devices can be a royal pain, and MIDI patch bays are equally annoying.
[1] Mixing analogue and digital signals is another pain in the butt, particularly in live situations.
[2] Somewhat related to [1], with any decent sized MIDI setup, the number of midi and patch cords becomes almost unmanagable. GMIC will at least cut this mess in half.
These are just a few of them. Most importantly, GMICS should, if accepted by manufactures, go a long way in unifying studio and live setups, and making them easier to deal with (even with fairly small "studios" like mine).
On the other hand, I'm not going to get too excited untill a few things happen.
[0] Manufacturers other than Gibson need to accept this as a standard and start building gear around the standard.
[1] Decent software needs to exist before I'll consider it usefull.
[2] It better deal well with my pre-existing gear in some sane and reasonable manner, or it is pointless. I like the gear I have and I don't intend to sell it all off just so I can retrofit. (fortunately most of my gear is MIDI enabled, including FX processors, so this might not be too big a deal).
Anyhow, it sounds pretty cool, and I'm looking forward to seeing how it turns out in practice. It's about time somebody is trying to update digital interfaces for instruments. I don't think the MMA is going to like this much