Great time now to buy a Mac Pro

[kgdrum]

There are also Firewire MIDI devices that should theoretically outperform USB devices if bus speed is the question. I use the Track 16 for MIDI.and it doesn't seem any faster or slower for MIDI data than the ME128

I've always assumed its mostly about how well the MIDI drivers are written,I suspect MIDI takes a very small part of the actual bus bandwidth.

[mikehalloran]

And I also have to wonder if MIDI latency can be improved with thunderbolt. I have always felt that usb was sluggish compared to the old serial ports from back in the day, let run MIDI through some lightening fast thunderbolt along with audio and see what we get yes?

Not sure where you came up with that notion. Sample accurate timing is possible with USB MIDI. This was never possible through serial ports nor did anyone ever claim it could be.

[toodamnhip]

And I also have to wonder if MIDI latency can be improved with thunderbolt. I have always felt that usb was sluggish compared to the old serial ports from back in the day, let run MIDI through some lightening fast thunderbolt along with audio and see what we get yes?

Not sure where you came up with that notion. Sample accurate timing is possible with USB MIDI. This was never possible through serial ports nor did anyone ever claim it could be.

I came up with that "notion" by means of speed of throughput and latency. Perhaps computers can adjust to make things sample accurate now. But the input latency of usb has always been suspect to me. I have read articles on this in the past and wish I could recall where they were to provide a link. I am sure anyone interested could do a google search. But right or wrong, my point is I hope thunderbolt provides improvement in throughput in computers.

[James Steele]

And I also have to wonder if MIDI latency can be improved with thunderbolt. I have always felt that usb was sluggish compared to the old serial ports from back in the day, let run MIDI through some lightening fast thunderbolt along with audio and see what we get yes?

Not sure where you came up with that notion. Sample accurate timing is possible with USB MIDI. This was never possible through serial ports nor did anyone ever claim it could be.

I came up with that "notion" by means of speed of throughput and latency. Perhaps computers can adjust to make things sample accurate now. But the input latency of usb has always been suspect to me. I have read articles on this in the past and wish I could recall where they were to provide a link. I am sure anyone interested could do a google search. But right or wrong, my point is I hope thunderbolt provides improvement in throughput in computers.

Yeah, but you're overlooking the system that MOTU designed into their USB MIDI interfaces like the MTP/AV USB that had a propriety timing system built in when used with DP. How fast the data went from the Mac to the MTP/AV was irrelevant as it was buffered at the MTP/AV and then sent out and precisely the correct time. I think it was called MTS now... MIDI Time Stamping. I'll look for a link.

[James Steele]

Here you go: http://lmgtfy.com/?q=MOTU+MIDI+Time+Sta ... +Performer

[toodamnhip]

Here you go: http://lmgtfy.com/?q=MOTU+MIDI+Time+Sta ... +Performer

HEY James, thanks for the link. But to me, speed of throughput is even more imortant than post performance adjustment.
I think the digital recording world has been around long enough for through-put latency to be a thing of the past and I am hoping thunderbolt has the capability to make that a reality or at least CLOSER to reality. For example, Having a drummer come in and play Roland V Drums or even virtual pianos, basses, whatever, I don't want ANY lag.
When I moved over to VE Pro I was able to run buffers at 64 more of the time and I said to myself, "dayum, this is nice...I don't have to quite quantize as much".
I am a MIDI wiz. I always make MIDI sound great, but I work too hard and would have to work way less if I could have 100 tracks of full audio and still record MIDI at 64 or even ZERO buffers. Yeah, I know, I am always the guy complaining...but that's what I want and I gotta think that SOMEDAY, MIDI latency will be a distant memory right? I mean, come on man! Are we going to be relying on post performance computer "timing make-up" for another 10-20 years? I hope not. I want blazing speed and zero latency in audio and MIDI, and I have little patience for less than that. Thunderbolt sounded exciting when it's specs were 1st announced yet I;ve seen no new MIDI products or even many audio products that seem to really take advantage of it. Am I missing something? That is all I have been talking about, the potential in the new macs and in thunderbolt to finally take a step forward in realtime latency.

[James Steele]

Maybe we're talking about input Latency then or something, because I'm talking about playback latency and my point is that on MIDI playback latency there's doesn't appear to be ANY improvement to be gained by a MIDI interface being connected via something other than USB. Granted input latency is another story perhaps. That's all I'm saying and external modules aren't much of a factory any more anyway.

[toodamnhip]

Maybe we're talking about input Latency then or something, because I'm talking about playback latency and my point is that on MIDI playback latency there's doesn't appear to be ANY improvement to be gained by a MIDI interface being connected via something other than USB. Granted input latency is another story perhaps. That's all I'm saying and external modules aren't much of a factory any more anyway.

Yes, playback latency is a solved issue. And yes, I am talking about speed of signal on input being so fast as to approximate analogue..

[mikehalloran]

For example, Having a drummer come in and play Roland V Drums or even virtual pianos, basses, whatever, I don't want ANY lag.

Well isn't that nice? At some level, we probably all want that. BTW, you are asking for an event to trigger a simultaneous event. Ain't never gonna happen as long as the laws of physics still apply.

Einstein's theory that the speed of light is the maximum physical limit has yet to find a workaround in modern electronics. TB vs FW vs USB is not the real bottleneck. Resistors, coils, transformers, transistors, ICs, processors, memory chips etc are. True, ICs and transistors have less of a path than tubes and transformers but as long as there is any physical path for those electrons, there is going to be latency. Every object put in the path of those electrons slows them down - some way, way more than others.

That there are ways to buffer paths to achieve any kind on synchronicity is a fantastic achievement. It may be possible to make it better but some form of smoke and mirrors will always be involved (MIDI Time Stamp, Cue Mix etc.). An engineer has to find a way to accomplish the task by working with the latency issue... it will never go away.

[FMiguelez]

Einstein's theory that the speed of light is the maximum physical limit has yet to find a workaround in modern electronics....

Yup.

Scientists are making their tests with primitive quantum-based computers. And if they ever figure out the quantum entanglement conundrum, and how to make technology out of it, that's when the next big break-through in computing will take place.

I personally think it's a matter of time.

And that's when TDH's wishes regarding latency will finally come true...

[Dan Worley]

Even a real piano has latency.

[BKK-OZ]

Sometimes I wish there was more latency between my thoughts and my mouth.

[FMiguelez]

Sometimes I wish there was more latency between my thoughts and my mouth.

Exactly. Same here!

Same applies for email... or even answering posts in MOTUNation.

[toodamnhip]

For example, Having a drummer come in and play Roland V Drums or even virtual pianos, basses, whatever, I don't want ANY lag.

Well isn't that nice? At some level, we probably all want that. BTW, you are asking for an event to trigger a simultaneous event. Ain't never gonna happen as long as the laws of physics still apply.

Einstein's theory that the speed of light is the maximum physical limit has yet to find a workaround in modern electronics. TB vs FW vs USB is not the real bottleneck. Resistors, coils, transformers, transistors, ICs, processors, memory chips etc are. True, ICs and transistors have less of a path than tubes and transformers but as long as there is any physical path for those electrons, there is going to be latency. Every object put in the path of those electrons slows them down - some way, way more than others.

That there are ways to buffer paths to achieve any kind on synchronicity is a fantastic achievement. It may be possible to make it better but some form of smoke and mirrors will always be involved (MIDI Time Stamp, Cue Mix etc.). An engineer has to find a way to accomplish the task by working with the latency issue... it will never go away.

I think “zero" latency might have been a misstatement, but as close to zero within Einstein’s laws would be nice. That means the fastest protocol for data transfer, (thunderbolt), and a computer that can handle smaller chunks of data resulting in almost immediate throughput. There is much room for improvement, and , again, I have yet to see new equipment come out to match the new protocol for data transfer, (thunderbolt), nor the new mac’s supposed speed. For such I am hopeful.

You see guys, when I see something like thunderbolt hit the market, I really expect to see hardware rise up to utilize the new capabilities. I am surprised no one else has wondered about this. How about a new THUNDERBOLT MIDI interface with latency “throughput” 4x current capabilities...something like that

[Gravity Jim]

USB 2.0 already outruns many, many channels of MIDI. USB 3.0 is something like 100 times faster than that, and so represents massive overkill. What would be the point of Thunderbolt MIDI?

More buss speed will not improve MIDI latency, most of which occurs in host software. There are some clear advantages to Thunderbolt audio interfaces if one is running many channels of 96K or higher audio, but I push 32 channels of 96K/24 bit audio thru a FireWire cable without a hiccup.

So all this drooling and imagineering over what TBolt could do for MIDI doesn't mean a thing. In short, it won't do anything.