Page 1 of 1
So what exactly is the ultra-low latency time?
Posted: Mon Jun 20, 2005 10:21 am
by loff56
So, I'm debating purchasing Mach Five if for no other reason than the advertised "ultra-low software latency". I currently have EWQLSO Silver edition, and I do like the piano from that collection and use it a lot, (though I may step up to something better as well), but as a pianist, I actually find the 10ms latency in Kompakt to be too slow for live playing. Does Mach Five beat that 10ms? 5ms, 3ms, 1ms? Can someone who has it actually tell me exactly what it is.
Also, am I missing a piece of the puzzle here? How much latency is added by going from the keyboard through the MIDI FastLane device, then through the USB cable, then through the computer hardware, DP (4.52), the MIDI Track which routs to the Instrument Track, then out through the Firewire audio card (M-Audio's Firewire 410) before the sound even gets to the speakers? Is a near 0 latency pipe dreams for the way I have this set up?
Thanks.
(Dual 2.0 G5, 2Gig Ram, 250Gig HD, M-Audio Firewire 410, Delta66 Omni I/O, OS 10.4, DP 4.52)
Re: So what exactly is the ultra-low latency time?
Posted: Fri Jun 24, 2005 4:11 pm
by tcollins
I'm using hardware samplers, but with DP and Moto interfaces. I have a usb MIDI interface (moto microlite, that I bought for extra channels), and the firewire 828mk2. I noticed latency in the USB interface when playing, but when I used the MIDI input in the firewire 828mk2, there wasn't a problem. The moral of the story is: the firewire MIDI interface seems much faster.
Re: So what exactly is the ultra-low latency time?
Posted: Sun Jul 03, 2005 2:42 am
by terra
Set the "bufer" for 256 or 128 in setup of the Hardware... with this, you will not have more "delay""(latency).
Re: So what exactly is the ultra-low latency time?
Posted: Sat Jul 09, 2005 7:17 pm
by mhschmieder
I can verify the statement about the MOTU 828 mk II not introducing noticeable latency vs. any of the MOTU USB MIDI interfaces, in regards to the demos I downloaded of Kontakt 2 and MOTU MX-4 (since there's no demo of Mach Five and I'm still debating buying it). It is in fact the improved MIDI performance and reliability that the MOTU 828 mk II gives us, that finally moved me in the direction of soft samplers and away from hardware samplers and workstations last month (when I bought the 828 mk II).
Back to the specific question, is the "ultra-low latency time" marketing hype, or is Mach Five measurably less prone to latency than other soft samplers (such as Kontakt 2)?
Re: So what exactly is the ultra-low latency time?
Posted: Sat Jul 09, 2005 7:19 pm
by mhschmieder
Again referring to the original post: is the 10ms latency in Kontakt for version 1, version 2, or both? This could be a make-or-break decision for me in deciding on Kontakt vs. Mach Five, as I really cannot accept any latency greater than 5ms for critical work. Too many years playing acoustic instruments to not notice!
Re: So what exactly is the ultra-low latency time?
Posted: Mon Jul 11, 2005 10:58 am
by jeremyroberts
FYI, most hardware instruments (korg, roland, yamaha, etc...) have latency in the 10-13ms range. As electronic musicians, we've learned to play on top of the beat to hear the sound we want.
the problem is when you quantize to a grid - you won't like how it feels (it's late) and some poorly designed instruments, where latency is cumulative depending on polyphony.
Keyboard did a test on this maybe 10 year ago - and the one thing we learned to look for, was how an instrument deals with 10 notes of polyphony, as opposed to 1. the better instruments had no change. Some got progressively worse as polyphony increased.
I haven't tested M5 - has anyone?
Re: So what exactly is the ultra-low latency time?
Posted: Mon Jul 11, 2005 12:45 pm
by mhschmieder
Ah right, now I'm remembering the issue regarding cummulative latency with added polyphony.
Yes, every instrument, acoustic or otherwise, has latency, and we as musicians adjust to it.
It is harder with software instruments because it is less predictable and as other variables other than "instrument design" can affect the performance, thereby making it more difficult for the musician to adjust to the "natural" latency of the instrument.
With that said, less latency is still better

. Preferably under 20 ms and closer to 5-10 ms.