so, uhhh.....

im aware that an apogee big ben is a clock for digital audio. i maybe wrong that it has much to do with midi directly. i was under the impression that a good masterclock and wordclock does drastically improve a computer's ability to understand digital time. a hardware sequencer vs. an overglorified wordprocessor im sure that the hardware sequencer's timing is more stable. however, i assume that a computer with a good clock, proper drivers, and a dedicated and properly maintained operating system one would be able to achieve similar results. how bout the rme fireface as an example then instead of bigben.

http://www.rme-audio.com/english/firewire/ff800.htm#FEATURES

yeah arsen: wrong type of clock

if you take a very simple computer with direct CPU access to MIDI, then if it wants to send out a clock signal every 1000 ticks there's no reason why there should be any timing problems at all

a PC/Mac is a different entity - for it to send a signal to an external MIDI device involves a high-level instruction sets talking to high-level scripts, at the same time as the computer's carrying out 50 or 60 other high-level tasks on fifo via various resource managers and multitasking scripts - the hardware then has to communicate with external hardware via a device manager script, which is also coded high-level... for an extreme example, just look at the windows clock - it gets minutes, hours and days right, but the timing of the individual seconds is all over the place... this doesn't directly relate to MIDI, but it shows you how a PC or Mac's processing bottleneck changes from moment to moment

it's not really theory, i work with £25k Pro Tools rigs every day - i come home and sequence on a laptop or h/w sequencer - if i couldn't hear an obvious difference it wouldn't be worth commenting on... i also help test and develop a lot of software and hardware for companies like korg and yamaha - MIDI timing jitter (random timing errors - not latency you understand) is one of many things you have to comment on and often measure on large samples of wave data

a PC is a great tool for graphic design and 3D modelling - but it's an awful design for music app's... a 1980's eventide fx unit with just a few old DSP chips can pitch shift and harmonize digitally, in real time, without breaking a sweat... the best computers we use with dual processors and tdm hardware can only manage a single instance of that standard of harmonizing on such a sloppy, inefficient operating system... you need literally hundreds of times more power to fuel an rta like that

Anyone could really easily solve this for themselves by simply sampling a click or a drum hit from a piece of MIDI kit.

Then make a .wav file of 60 seconds of 2 channels of audio, on hard left, the other hard right. In this wav file, have the L panned channel play the sample of the drum hit, and then R channel play the sequenced MIDI gear.

Then do the same deal, but just use the sample on both the R and L channels.

Then, do a google search for ABX (double blind testing software). There's tons available for free. Usually they are used to test audio compression, but you can just as well give it one of these as an A sample and the other a B sample. Then the computer will give you a double blind test to see if you can actually tell the difference between the two.

Here, if anyone would like to satisfy their curiosity: PC ABX Software (for Mac, Windows, and Linux)

the simplest way is to understand how and to what degree timing affects musical and rhythmic phrases, then measure MIDI timing jitter by sequencing and recording a MIDI click, measuring and aligning to modal deviation then manually measuring maximum deviation/error

remember, a/b testing in general can be interesting under certain circumstances, but by definition, an engineer/producer should be far more discerning about details than his/her audience.... the problem with a/b testing non-professionals is that there's a psychological tendancy to gravitate towards familiarity above objective quality - but you put someone like that in charge of a recording studio and you might as well funnel diahorrea into your ears

you have to remember, half the general public are practically tone deaf, and even fewer have an accurate sense of rhythm... diahorrea

take a look at the new version of MidiBoxSEQ - i know i'm gonna build me one!

(also watch other clips)

killed the thread?