Priest of the Order of the Butterfly
Posts: 510 from 2003/2/25
From: France
Quote:
djcollins wrote:
Hi Tcheko,
thank you for your answer. I will try the cam.library in 68k version. the thing I forgot to say is that I tested 3 midi interfaces, one very cheap, one at 300 euros and another even more expensive, the result is exactly the same. I work all day but tonight I'll give you an example, it's very interesting. It will be necessary to listen well in the rhythmic, it is there that that shocks when one listens. We feel that the kick is offset from the rest and that the song does not have the same impact. In fact, in USB, it just feels like listening to a non-quantized track. I'm coming back here tonight. It is clear that if we can improve the midi in USB it is the simplest solution but apparently it is a general problem, on all systems, even in 2023 at the age of USB 3 or 4! explanation here:
https://midi.org/midi/forum/4025-jitter-and-latency-usb-firewire-thunderbolt
Well. USB is quite a complex thing.
I would advice to use the MIDI USB interface on an unpopulated USB port (ie: no competing device on the bus). At least, that way, all the packet frames will be allocated at a more precise time location. I guess all your MIDI interfaces are working in 'bulk mode', which translate to 'send my data when you can' on the USB bus (ie: the slot is allocated when some room is available on the bus).
Also, the link you passed isn't complete on the subject. USB frame rate can go up to 8KHz (aka 125us microframe) for hi speed devices which will likely help with quantization by reducing the jitter by the same factor. Finding an interface working in hi speed should really help here with those quantization artefacts produced by the USB bus timing.
Still, I'll be really happy to get two recordings, one made with an Amiga and one made with the Pegasos to get an idea of how different it sounds.
