"Logic 7 does not support hardware MIDI"- Agree?

[ski]

Looks like it's in the hands of the industry politicians. We'll have to wait and see. Anyone interested in a new career as a lobbyist? C'mon. How about you ski? . . . . Yea, that's it. I nominate you ski!

 

OMG... thanks for the vote of confidence, but, um, I must respectfully decline. One more thing on my plate and I'd explode like Mr. Creosote.

 

[xs4is]

I'm slowly trying to build up my reputation as a complete tool when it comes to MIDI, but this is an even broader question. How does a computer keep accurate time (real, clock time)? Sometimes when I get really bored, I'll watch my computer's clock and try to synchronize another clock to it. I am almost always disappointed that the computer clock's numbers don't update when they should almost every single time. I'm talking a delay or even a slight jump of a second or two.

 

If a computer can't keep real time accurately, how can it do MIDI clocking?

 

X

[David Nahmani]

AFAIK computers use quartz to keep track of time, much like watches.

[ski]

AFAIK computers use quartz to keep track of time, much like watches.

 

That's what I thought too.

 

And just to 'splain a little further... What David's referring to are actual pieces of quartz packged into an electronic component, like this:

 

http://www.score2picture.com/logicpix/quartz.jpg

 

Quartz molecules vibrate at a very high rate of speed and output a very small but measurable fluctuating electrical charge. The speed of vibrations ("frequency") remain stable under a wide range of temperatures. So what you have with quartz is a stable source of electrical current that alternately outputs a very small "high" current followed by no current ("low"), over and over again. This happens at very high rates of speed, typically at hundreds of millions of cycles per second.

 

It's a clock.

 

The signal from this quartz crystal is amplified so that it's usable. How's it used? Well, lets say for arguments sake that the crystal vibrates at 60 million vibrations per second. If you connected this crystal to a circuit designed to count every vibration, you could program that circuit to flash an LED every time 60 million cycles have been counted. In this case, the LED would flash every 1 second.

 

This is called a "divide down" mechanism, where a circuit (and/or software) counts the quartz clock's vibrations and indicates when a certain number of vibrations have been counted. Such circuits/software are usually designed to "tap off" indications of milliseconds (or even smaller increments of time), jiffies (1/10ths of a second), seconds, hours, days, etc.

 

There are even circuits/software that can take that 60 million vibrations per second and interpolate between cycles to derive even higher frequencies (smaller slices of time). So a 60 MHz crystal can actually be used to generate 120 MHz, for example.

 

So crystals are used as a master clock source for computers and lots of other electronic gear.

 

Anyway, gotta run.

[editbrain]

asher, sorry.

 

x, i use ntp for clock on computer.

 

ski, usb is a serial connection as well. I know you knew that.

 

fred b, 384 bytes/sec does not take a high priority IMO.. but is this per note, or data sent and received? is every packet from midi 384 bytes?

 

it is strange. we use USB midi devices. hmm....

so we are using a device with 10mb/sec transfer rate (USB 1.0) to transmit 384 bytes/sec, and Logic still doesn't send a constant signal? i know that the limits are on midi and not the connections.

 

there is no reason that midi cannot transfer more information. it isn't like the it is a limit of the cabling..

I am not a midi wizard. hell! i can't even program a VCR. so clocks are not my bag, but why is is that Logic the "Greatest Midi Sequencer" has issues with keeping a running time?

 

i wonder if it is possible to write a midi clock application that Logic would sync to as well as other midi information?

[ski]

First, I stand corrected. The MIDI transmission rate is 31.25K bits per second, not 384Kbytes/sec.

 

ski, usb is a serial connection as well. I know you knew that.

 

Yeah, I did 8) . My point was that a high speed (serial) communication like USB affords faster thruput of info, but it's not a dedicated system like MIDI is.

 

fred b, 384 bytes/sec does not take a high priority IMO.. but is this per note, or data sent and received? is every packet from midi 384 bytes?

 

If I may answer this one... The bandwidth of MIDI is fixed, but the length of messages are different and are allowed to occur at any instant in time.

 

Each MIDI byte is approximately 1/3 millisecond wide.

 

Examples (not complete lists here...)

 

1-byte messages: MIDI Clock, Start, Stop, Continue, Active Sensing

 

2-byte messages: Program Change, Aftertouch, Pitch Bend (7-bit)

 

3-byte messages: Note On, Note Off, Controllers (all)

 

Variable # of byte messages: sysex

 

So a 3-byte message like Note On (containing "note on", "note number", and "velocity") is approximately 1 millisecond wide. So a four note chord, with all notes played "simultaneously" takes 4 milliseconds to be transmitted from start to finish. Each note is sent serially, one after the other. So with MIDI there is no such thing as truly simultaneous events occuring when data is sent on one cable to a single device. Any chord where the notes are played simultaneously -- in real time -- are "translated" into MIDI as a very fast arpeggio.

 

Again, MIDI isn't a clocked system like digital audio, so a MIDI event is allowed to occur at any instant in time. But the maximum amount of data that can travel down a MIDI cable is limited to that 31.25kbit bandwidth. It's not a lot of bandwidth.

 

With respect to MIDI Clock, each "tick" at 60 BPM would have to be calculated every 41.6 millisconds. I don't think that's too much to ask for on computers running in the GHz range. As Fred said, yes, MIDI clock would have to be made a priority. But I must point out again that there are many more superfluous things in Logic that seem to take priority over MIDI events.

[xs4is]

ski, is there anything you don't know at least a little of?

 

Stamp collecting maybe?

 

X

[David Nahmani]

So a 3-byte message like Note On (containing "note on", "note number", and "velocity") is approximately 1 millisecond wide. So a four note chord, with all notes played "simultaneously" takes 4 milliseconds to be transmitted from start to finish. Each note is sent serially, one after the other. So with MIDI there is no such thing as truly simultaneous events occuring when data is sent on one cable to a single device. Any chord where the notes are played simultaneously -- in real time -- are "translated" into MIDI as a very fast arpeggio.

Now imagine trying to play back a 16 part multitimbral MIDI sequence with piano chords, drums, guitar chords, with volume pedal, pitch bend info, and maybe some MIDI CCs controlling filter cutoff or such... let's just say that the "1" might not fall where you expect it.

[ski]

David,

 

Well.. when you put it like that, I'm very tempted to say...

 

MIDI SUX!

 

 

 

sKi

[David Nahmani]

Actually, I remember back in december 2005 I had the chance to meet 2 of the Logic developers and Logic's software architect when I invited them to be the guest speakers at one of our first L.A. user group, and one of the questions I asked was whether they were aware of/participating in the development of HD MIDI, aka MIDI 2.0 (http://www.midi.org/newsviews/hdmidipr.shtml).

 

They were aware of it but I don't think were very much involved in its development.

[editbrain]

Now imagine trying to play back a 16 part multitimbral MIDI sequence with piano chords, drums, guitar chords, with volume pedal, pitch bend info, and maybe some MIDI CCs controlling filter cutoff or such... let's just say that the "1" might not fall where you expect it.

 

wouldn't most of that be handled internally by the keyboard itself, and only the Start, Stop, and clock be sent to the keyboard by midi from Logic?

[David Nahmani]

wouldn't most of that be handled internally by the keyboard itself, and only the Start, Stop, and clock be sent to the keyboard by midi from Logic?

Not if you're sequencing in Logic, only if you were sequencing in your keyboard and syncing your keyboard with Logic... and who wants to sequence on a 4 line led display when you have Logic and a 20" LCD display!

[editbrain]

i knew you were going to say that, but i just assumed that the topic was "sync with midi clock" is the reason i did not make that apparent.

 

i meant only syncing. so if what you said above is true why can't logic send a constant clock on the bandwidth that is allowed if all is being sequenced in an outboard sequencer?

[editbrain]

Double Post ---oops!

[David Nahmani]

I see.. I had forgotten about the original topic for a sec, to focus on the limitations of MIDI in general.

 

Well... Logic can and it does send a constant clock. I don't have enough time to start going on another board and troubleshoot the issues of a non-LogicProHelp member, but I have used MIDI Clock both over IAC and over physical MIDI Interfaces to sync external sequencers with no problems.

Edited June 26, 2007 by David Nahmani

[editbrain]

so where is the drift from the OP coming from, and where are the limits of midi being reached on this thread?

[David Nahmani]

so where is the drift from the OP coming from, and where are the limits of midi being reached on this thread?

I don't know. I haven't experienced those problems.

[editbrain]

okay. well it seems that there are varying opinions on what logic can and does do.

 

Asher says midi clock is broke.

You say it works correctly.

 

There are limitations to midi, but if your doing all your sequencing on the keyboard and not in logic then your not over running messages because logic is just sending a clock that is broke or not broke depending on who you talk to.

 

this is all pretty much guess work, and seems that nobody knows what the hell is going on with Logic. pretty much the same as the speculation that Logic 8 is coming out, or not.

 

midi is going away, and/or needs improvements, but the last generation of samplers and workstations have implemented it.

Midi has worked for years. despite any issues people have had it is the only thing we have.

Now you have midi over gigabit ethernet. talk about a waste of bandwidth.

 

you have inter application control that uses midi.

 

where in the world is this going?

 

you have wireless midi triggers, midi controlled dj mixers, etc, etc.

 

 

so which is it. BROKE or WORKS?

[David Nahmani]

Now you have midi over gigabit ethernet. talk about a waste of bandwidth.

You can use your ethernet connection for MIDI connection. Doesn't mean that the MIDI connection will take over the ethernet connection: you can still use it for many other things!

 

As for saying that a feature is broken or does work... I find most people are too quick to conclude something works or doesn't work. All I've said is that in my experience it does work, but I don't use it every single day, and I have heard of people having problems with it. The poster who posted on Gearslutz seems to have tried and tried again several time before he concluded that it doesn't work for him. Who knows, maybe the guy really couldn't make it work and he's got a legit point, or maybe he simply forgot a setting, got a MIDI feedback loop, or corrupted preference file... hard to tell. As for "my friend who's an apple guy told me that..." ... well... you and I both know that unless Apple says something officially, it is all to be taken with a grain of salt.

 

I would say who cares about what he says. If you need it, try it, and see if it works. Maybe it will work for you. If it doesn't find a workaround, or use another DAW. As a Logic consultant I don't always insist on people using Logic all the time. Certain things are better achieved with other daws. At the end of the day, what matters is that you can move on with your music production!

 

But hey, just the other day, I synched two machines running Logic using MIDI Clock over a wireless network: a laptop and a G5. Worked flawlessly. Go figure.

[editbrain]

awesome.

 

well there you have it folks.

[ski]

Getting back to the subject for a second...

 

I'd like to ask once again what I think is a key question here: the guy on Gearslutz that was complaining about MIDI clock being sloppy --- did he mention what kind of interface he was using? I think this is a key question...

[gsilbers]

most of the thread was or being defensive or poking at it agresively. so not a lot of the little details that could help. the thread was more of a final statement.

 

i tried to mention about not using the 808 and sampling it via autosampler and the guy said something like Im the consumer apple wants.. to ditch the hardware and go software (or something like that kinda mentioned here. )

 

i didnt want to coment back to fuel more useless argument but unless u are doing a lot of realtime pitch or frequency events on an 808 then by sampleing your own youd basically have the exact same thing and the programming is 10x easier. how can u program an 808 record it and then do the exact same programing but with samples from that exact 808 and not have the same outcome? is there something in my 808 (909 for me) that im not aware of. does each time i program and record it sounds diferen thus better? is there a magic quantize in the 808. last i checked my 909 sounded like crap and it was dificult to program so i sample it raw and add the nesesary efect to make it sound desent. (id say same as the 303 but the 303 has more filter and envelope movement and more expresion which makes sense to have the real thing.).

 

I dont agree that an 808 is like a lespaul thoug as someone mentioned. as the les paul was going for the high end market whle the 808 was a failure when it came out to be later sold in pawn shops and later explode in popularity and be use for something completly diferent for what was intended for.

and of course its a guitar and the 808 is more "static".

 

 

 

what a nerve that guy hit...

 

 

i dont use much of external hardware cause basically its a pain no matter what program. conect midi , route midi, conect audio, route audio, at the end something like drums i can do far better and easily with samples or a drum machine. as for the synth sounds, id say 80% of stuff we hear on the radio or from artist comes from softsynth and most people wont notce a difrence.

that said, i have a virus TI. got it for the convinieve of the plugin cause the midi to control the virus C just sucked, one thing or another logic and the virus C just didnt get along. difrent midi CCs , patch recal etc. bt also hapen with PT and LIVE.

 

hopefull like Dave said there will a better or more standarize version of MIDI or something that all companies will implement the same thatll make it easier to just conect and it works. the whole thing of USB MIDI rocks for me , I wish it was in every synth so audio and midi go though it.

[ski]

The whole samples vs. original source thing could be a whole thread in and of itself, but I'll just say this... That the static predictability of samples is a convenience, for sure, but the "live", never-exactly-the-same-twice aspect of analog gear is, pardon the pun, "precisely" what makes a lot of vintage gear have so much personality. But "personality" can mean "unpredictability", i.e., "problems".

[gsilbers]

The whole samples vs. original source thing could be a whole thread in and of itself, but I'll just say this... That the static predictability of samples is a convenience, for sure, but the "live", never-exactly-the-same-twice aspect of analog gear is, pardon the pun, "precisely" what makes a lot of vintage gear have so much personality. But "personality" can mean "unpredictability", i.e., "problems".

 

i understand that but dont agree 100% but dont want to go too far away form the original thread. its true ot could be a whole other thread.

[ski]

I agree. Let's pretend I never said anything... OK? OK!

 

8)