keyspan sx serial cards

[supernova777]

http://macintoshgarden.org/apps/sysex480
did u ever try this thijng syntho?

[Syntho]

That's really of no use since this is probably a hardware-related thing. I'll try again when my Studio 5 gets here.

[supernova777]

That's really of no use since this is probably a hardware-related thing. I'll try again when my Studio 5 gets here.

i take it thats a no;)

[Syntho]

OK, good news. The KeySpan is completely fine, it's the Express XT that's giving issues. You can't double click on it and change the speed in OMS, and FreeMidi/Clockworks has an option to change it but it never saves the configuration for some reason.

I just hooked up my Unitor via serial and it's transferring completely fine. I imagine the Studio 5 will do the same, it's being delivered tomorrow.

The Express XT happens to work completely fine with the parallel port, but its serial ports seem to have trouble... I guess those things were programmed more with parallel ports in mind than serial. I'll still hang onto it for a backup.

[supernova777]

OK, good news. The KeySpan is completely fine, it's the Express XT that's giving issues. You can't double click on it and change the speed in OMS, and FreeMidi/Clockworks has an option to change it but it never saves the configuration for some reason.

I just hooked up my Unitor via serial and it's transferring completely fine. I imagine the Studio 5 will do the same, it's being delivered tomorrow.

The Express XT happens to work completely fine with the parallel port, but its serial ports seem to have trouble... I guess those things were programmed more with parallel ports in mind than serial. I'll still hang onto it for a backup.

maybe its damaged..?
or doesnt support this speed that u require for the sysex bank dump

[Syntho]

VERDICT:

The studio 5 rules 

[supernova777]

VERDICT:

The studio 5 rules 

damn u got it already? thats fast.
yeah i like my opcode 128x so far..
i kind of think for midi opcode knows best considering they wrote oms.
and everything requires oms.

[Syntho]

I'm running a dual port setup. The manual says that if you need more 'midi throughput' to use a dual port configuration but it doesn't explain how it does this. It does say that it distributes midi information 'equitably' on both ports, whatever they mean by that. OMS shows that it's connected to both the modem and printer ports (it has a / indicating it) but I don't see an area where you can actually assign different midi inputs/outputs to the A and B serial ports. Maybe the Studio 5 utilizes both of these automatically or something?

The manual also states that any timecode the Studio 5 generates is sent on the B port, while midi information is sent on the A port. It says generates though, so I'm not sure if that includes incoming MTC.

If this were 1997 all I'd have to do is make a phone call to figure this out 

[supernova777]

yea sure or if gibson guitars didnt kill opcode 

[coachla]

Syntho... Glad to hear the news! No you can't like with the MTP console and Motu gear. But it's so robust, there isn't any need to. All incoming midi is through the A port. Are you receiving SMPTE IN or MTC over midi cable? Run it full throttle on one port and see if it handles that.

Get another if you can....You can never have too many BLB's (Blinking Light Boxes)!

[Syntho]

I'm going to have incoming MTC for sure. The generated/converted audio SMPTE to MTC I'm sure would be sent on the B port, but plain MTC itself from a midi cable I'm not sure about. It says it distributes the information 'equitably' on both ports, but also that the midi information is sent on the A port. Almost contradicting info... maybe it means it SENDS information to the Studio 5 on both ports, but that the incoming stuff is always on the A port (besides SMPTE)?

I need to test this with a Beige G3 or lower with actual modem/printer ports and Logic 4 with a cabled physical input object. I'll have a 9600 someday so I guess I'll test when that time comes.

[coachla]

If it's not generated by the Studio 5 data comes in via the A port, however the manual does not specify the distribution of the outgoing messages which if divided between the serial ports would significantly reduce traffic, given that your incoming MTC as well as sysex receives are added to the A port's overall data flow. Let us know if you compare the configurations with differing results please.

Logic 4 even with a keyspan and OMS has discreet inputs. No problems with Logic 4 and that Sum-input-only phenomena on my MDD. But you're right, because OMS preferences can only allocate to the first 8 M ports on the physical input object, you can't test it on the MDD. Without OMS means using MTP mode which may not be as efficient to compare accurately, if in fact you can actually get Logic 4 communication preferences to address an MTP on multiple ports. A big question mark on that one....

[Syntho]

Hey Coachla, I just did some testing. I don't remember all that was said but here's my findings on a 9600 with an old version of Logic using a Studio 5:

When using OMS, you're stuck with using nothing but the Sum part of the physical input object. In order for me to get the individual ports working, I have to disable OMS and click 'Modem' and/or 'Printer' in the Logic Communication preference screen. The Sum still works then, and as soon as I cable an individual port to something else, that port is taken off of the sum and is only coming through its actual port.

I think another step I had to do to get that working was to go into OMS and the Studio 5 setup screen and set up the 'generic midi interface' settings. Haven't tried without it but that's what I found.

I was hoping that somehow I could have OMS active and use the Sum, but also be able to break the connection for individual ports when I need to. That doesn't seem to work with OMS enabled unless there's a step I'm missing.

[supernova777]

keyspan card still working great?

[Syntho]

Of course.

I need coachla to come back and fill me in on the issue with the physical object...

[supernova777]

just a note to say that ive updated the original post in this thread to include the manual for the keyspan sx pro in pdf format
the manual makes reference to mac os x 10.1 so it is probably from the time that this was the current mac os x version

im also adding the "keyspan manager" which is to be added to the system folder (not sure if its compressed or what?)
theres a good chance this file is screwed from being uploaded improperly.. syntho did u have the sx pro manager system folder extension for the keyspan card?

[Syntho]

I posted the driver on the last page, it's there. KeyspanSXProV20.sit.bin

[supernova777]

http://www.macupdate.com/app/mac/4633/keyspan-sx-pro
What's New
Version 2.0 adds the following:
    Fixes issue with sleep on certain Macs

[supernova777]

http://www.cnet.com/news/keyspan-sx-pro-software-update-is-available/

http://web.archive.org/web/20060102181013/http://www.keyspan.com/products/sxpro/downloads/

this seems to lead back to the same file

ok this definately has the manager!!!
it has options for editing + fine tuning the speed of the fifo chip
and supports multiple cards aswell!!

options for:
card
port #
port name
recive fifo [8,16,56,60]
default buffer size
TX acknowledge
double baud rate
emulate printer port
epson support
and some kind of stats  + test button
with a small feedback window

[DaCat]

Does anyone know of any serial cards either PCI or USB that work in both OS9 & OSX for MIDI? I saw that Keyspan announced a driver for OSX but my understanding is it doesn't handle MIDI, I'm wondering if any of the USB ones do.

The other workaround I'm wondering, say if using an Opcode Studio 5LX, could one use two Macs, one in OS9 & the other in OSX and either using pass thru or synchronizing both computers, run the MIDI from the OSX computer while the OS9 one handles routing & maybe static librarian preset changes?

My basic setup is a G4 Quicksilver DP running Opcode SVP & Galaxy using a MOTU 2408MK3, while I want to experiment with MOTU DP 5 in Leopard on my 2008 MBP, and possibly at other times the MBP will be in Mavericks using current software.

The Studio 5LX seems ideal for running MIDI but then there's the challenge when working from the laptop in some relatively current OS.

Would a practical solution be a modern USB MIDI interface that would also work in OS9 and with Opcode SVP? My head is spinning a bit following this subject in several forums, maybe there's issues for SVP & Galaxy using current USB interfaces? I also notice Opcode had some MIDIport USB interfaces likely for USB 1 so would this work in modern Macs? Considering extension/driver issues it might be too much to expect a MIDI interface that could function across OS9 to Mavericks.

The snail workaround is simply re-patch all the MIDI cables with two dedicated interfaces while I suppose one could rig up a complex matrix to get around this but too many boxes with in/outs might start lagging. I must be getting lazy because I only have about a half dozen MIDI pieces at present.

Seeing that there's some success with the Keyspan in the MDD gives me reassurance this would be good for my Quicksilver, I'm thinking of upgrading to the MDD but have seen other Opcode users claim older computers are more stable, while I went with the Quicksilver thinking it was closer to the G3s, I also have my original G4 Yikes! that is essentially a G3 with G4 processor so that could be an alternative if there's any stability issues. They were referring to Vision/SVP re stability so maybe that's not an issue with MOTU or Pro Tools setups, but I did some basic cassette transfers into SVP on this Quicksilver & everything seemed to function OK.

My "fetish" area is seeing how maxed out a G4 could go running Opcode SVP, I currently use a Seritek PCI E-SATA card and it screams with an external SATA drive, also allows quick reboots to OSX partitions for using Leopard for Internet downloads, etc..

I also wonder if newer USB interfaces have better timing than serial versions? I don't need a lot of MIDI in/outs at present but could use for future expansion.