USB and CAMD (Audio) problem (Solved) Reason: Retarded user

I have an EDIROL UM-1X USB to MIDI converter which should work
(I have set the switch to OFF as told in the Hints section, otherwise it won't be detected at all)

In a way it does work already but only if i plug in the USB AFTER Workbench has loaded

If i leave it in while booting up or rebooting i get a "System Request"er a second after the boot screen shows, telling me "Unable to open USB Function driver "camdmidi.usbfd". When i click OK or wait two seconds it goes away and boots into Workbench.
The UM-1 is shown in USBInspector (as "AUDIO" in "Functionality" and "UM-1" in Topology" but NO cluster is created whatsoever, so no MIDI software, player or anything can send data to the hardware)

Further down is the USB log (which does list the UM-1 being added!)

It seems the cuplrit is the camdmidi.usbfd is started too early? (But do i know?)

Is anyone in here having a working installation of OS4.1u1 and an USB 2 MIDI switch that could throw in some hints?

@Dwyloc

Thanks for your answers in PM, but every piece you mentioned is already installed.

It IS working, my hardware gets to play the MIDI data that is sent, but ONLY if i plug in the switch AFTER Workbanch has loaded and that is not the way i want to use it.

@lylehaze

I hope you'll have the time to come around and maybe answer some of my questions? After inspection of the camdmidi.usbfd i found that it was built for OS4.0. Maybe that is the problem ... grasping at straws here

The other thing is that i first used MIDI through the onboard VIA chip and had to install a driver in an extra directory in devs: called "MIDI".
Could it be that i need something in there?

Thanks in advance for any help

USB log (USBInspector tells me that the UM-1 would be Product No: 0x0052 and Vendor No: 0x0582)

I: [08:26:39] USB stack | | ==========
I: [08:26:39] USB stack | | USB System Software Task started
I: [08:26:39] USB stack | | Starting HCDs...
I: [08:26:39] USB stack | | ...HCDs started.
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FF86DC0 is {Vendor: 0x0000, Product: 0x0000, Class: 09.00}
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FF86F00 is {Vendor: 0x0000, Product: 0x0000, Class: 09.00}
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FF86DC0 initialized
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FF86F00 initialized
I: [08:26:39] hub.usbfd | HUB | Device found on port 1 of hub "Rear Roothub (uhci.usbhcd/0)".
I: [08:26:39] hub.usbfd | HUB | Device found on port 1 of hub "Front Roothub (uhci.usbhcd/1)".
I: [08:26:39] hub.usbfd | Hub port | Device at port 1 of hub "Rear Roothub (uhci.usbhcd/0)" is low-speed.
I: [08:26:39] hub.usbfd | HUB | Device on hub port 1 successfully added.
I: [08:26:39] hub.usbfd | HUB | Device found on port 2 of hub "Rear Roothub (uhci.usbhcd/0)".
I: [08:26:39] hub.usbfd | Hub port | Device at port 1 of hub "Front Roothub (uhci.usbhcd/1)" is full-speed.
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE041F0 is {Vendor: 0x6D04, Product: 0x17C5, Class: 00.00}
I: [08:26:39] hub.usbfd | HUB | Device on hub port 1 successfully added.
I: [08:26:39] hub.usbfd | HUB | Port 2 of hub "Front Roothub (uhci.usbhcd/1)" disabled - port is empty.
I: [08:26:39] hub.usbfd | HUB | Driver Running
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE04290 is {Vendor: 0x2404, Product: 0x0425, Class: 09.00}
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE04290 initialized
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE041F0 initialized
I: [08:26:39] hub.usbfd | Hub port | Device at port 2 of hub "Rear Roothub (uhci.usbhcd/0)" is full-speed.
I: [08:26:39] hub.usbfd | HUB | Device on hub port 2 successfully added.
I: [08:26:39] hub.usbfd | HUB | Driver Running
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE043D0 is {Vendor: 0x8205, Product: 0x5200, Class: 00.00}
I: [08:26:39] USB Fkt Init | Init Fkt | Fkt 0x6FE043D0 initialized
I: [08:26:39] hub.usbfd | HUB | All ports are empty
I: [08:26:39] hub.usbfd | Sys_EndInitialAttachmentPhase | Initial USB Attachment Phase terminated
I: [08:26:39] USB stack Process | | USB System Software Process started.
I: [08:26:39] USB stack Process | | Reading FD descriptors...
I: [08:26:39] USB stack Process | | FD infos read
I: [08:26:39] USB stack Process | Rebind FD | Weak FD for Ifc 0x6FE5BE90 is being preempted.
I: [08:26:39] USB stack Process | Rebind FD | Weak FD for Ifc 0x6FE5BF10 is being preempted.
I: [08:26:39] USB stack Process | | Starting HCDs...
I: [08:26:39] USB stack Process | | ...HCDs started.
I: [08:26:39] USB stack Process | | USB stack fullbooted
I: [08:26:44] USB FD fkt start | Sys_EndInitialAttachmentPhase | Initial USB Attachment Phase terminated

This is logged when i plug it in again:

I: [08:49:12] hub.usbfd | HUB | Detaching Function removed from hub port 2
I: [08:49:12] hub.usbfd | HUB | Port 2 of hub "Rear Roothub (uhci.usbhcd/0)" disabled - port is empty.
I: [08:49:12] hub.usbfd | HUB | Device found on port 2 of hub "Rear Roothub (uhci.usbhcd/0)".
I: [08:49:12] hub.usbfd | Hub port | Device at port 2 of hub "Rear Roothub (uhci.usbhcd/0)" is full-speed.
I: [08:49:12] hub.usbfd | HUB | Device on hub port 2 successfully added.
I: [08:49:12] USB Fkt Init | Init Fkt | Fkt 0x6FE043D0 is {Vendor: 0x8205, Product: 0x5200, Class: 00.00}
I: [08:49:12] USB Fkt Init | Init Fkt | Fkt 0x6FE043D0 initialized

So it seems it IS added, but for some reason it doesn't create clusters...Why?

Bump

No one any ideas?

@Raziel

Lylehaze would probably be the best person to ask, at amigaworld.net.

@Raziel

I'm using the Edirol UM-1SX and have no problems since
the new Update 1 of the OS.

I do have it plugged into the front USB on a Sam440ep.

USB Inspector shows it as UM-1 ,Class:Audio, Subclass:3,
Driver camdmidi.usbfd. And topo of Product 0x0052, Vend 0x0582
version 2.0, Class:Multiclass, Subclass:0, Mode:Interface.

At this point I don't think it's camdmidi that is the problem since
it does work. I have no idea where to point you. Sorry

@trixie

I dont visit aw.net, could you maybe point him to this thread please?

He is a member here too, but doesnt seem to visit too often

Thank you

@mechanic

Thanks

I also dont think its camd.library.
I need to do some tests but i think i could workaround the problem by starting Sirius a little later than it is in startup-seuqence...then again i probably have to only figure out which part of the system needs to be up to make it work

Back at home tomorrow, so ill see

Throw in any ideas you may still have, please

@Raziel

You could try;

1. Add a wait to startup before or after USBstart. C:Wait 2
2. Keep a small USB mem stick always plugged in. This helps
steady the ports on my rev C sam for some reason (?)

???

@mechanic

1. No change
2. No change

3. I tried something other
I put USBCtrl START at

;C:USBCtrl START
C:Mount QUIET DEVS:DOSDrivers/~(#?.info)
C:AddNetInterface QUIET DEVS:NetInterfaces/~(#?.info)
C:USBCtrl START

and the little window popping up is gone.
Unfortunately i still don't have any clusters on WB

Note:
When i put the line even further down (this time at last line in user-startup) the device is load in correctly (given the information from USBInspector).
I also have the clusters available, but...no sound coming from the device (notes and messages are being sent to the device, but nothing is sent back to the soundcard) getting stranger by the minute
btw: if i let it sit there in user-startup and do a USBCTRL RESTART it stays the same, no sound output. If i put it back to where it belongs and do a RESTART i have sound back

Tis a hard one :-/

edit: Updated no 2
edit 2: Note: added

@mechanic

It seems it does something different for me:

After the window popped up i get in USB Inspector

Quote:

and

Quote:

So, two problems here, the driver is obviously not loaded --> Why?
and the Mode display (Topology) shows "Available" instead (as for you) "Interface" --> Why, again?

btw: after a "USBCTRL RESTART" the device is added right, the driver being loaded in and Mode set to "Interface".
I wonder why i always have to have such strange kind of behaviour :-/

@Raziel

Quote:

On my Samep rev C board a USBctrl restart will just kill the USB if
I have a mem stick plugged in.

Before Update 1 the USB ports would only do input, keyboard, mouse,
MIDI in. No output, no memory, etc..

My UM-SX1 has a LED that flashes when MIDI data is being
transfered. Before Up1 Showkeys, from LyleHazes' camdtools,
showed that MIDI data was going to the UM-1 but in reality
nothing was leaving the USB ports. However, using showkeys, I
could see data was going into the computer through the UM-1
from a MIDI keyboard. It now works both ways.

Which motherboard, soundcard, etc..?

Did you do a clean install of Update 1 ?

Could you please give your complete midi setup that your
working with. I'm confused....

@mechanic

The same for me, showkeys shows that there is data flowing and even my MT-32 shows that it receives data on the display, but nothing is going back to the sound card

A1XE fixed (more on the extra info page)
Aureon Space7.1 (EnvyHT24)
The UM-1X and a genuine Roland MT-32

I tend to believe that i won't be able to fix it unless i use it on new hardware
As someone on IRC told me, USB is maybe broken beyond repair on the XE's

@Raziel

MT-32, always wanted one of those.

Does the MT-32 have an output you can plug powered speakers into?

If it's getting the MIDI data then it should be producing sound. I think(?).

@mechanic

Heh, yes, awesome piece of hardware (you should still get those babies on eBay or alike, got mine from there)

It has two 6,3mm cinch on the backside which i bought a cable for to plug it into my soundcard.

As i have written above it WORKS as expected, but ONLY if i
a) plug it in AFTER WB has been loaded completely (which makes me still think it could be a CAMD initialisation problem, or a problem with the USB driver)
b) restart the USB stack (which is probably the same as a) anyway)

@Raziel

Nope. I'm out of ideas. Wish I had an A1Xe to play with.

Right now, because I'm too lazy to do a complete setup, I'm
using a Roland PMA-5. Not the greatest, but good enough.
Everything else Amiga related is pretty much waiting for X1000.

@Raziel

have you tried moving the usbctl line to alter point in your startup sequence?

It may give some clues as to what isn't yet initialised...

Try moving it forward a line at a time till it works, backing up your original ofcourse.

@broadblues

See my post no 7 in this thread

The problem is that everything LOOKS normal when i move it forward, but no sound reaches the sound card, unless i attach the device AFTER WB has loaded completely.

To me it seems ONE part of the system is not yet ready. AHI, CAMD, USB to name a few, too much possibilities

Unless i can get at least one confirmation (preferrably from CAMD maintainer lylehaze) i'm stuck with too many unknowns

@Raziel

Quote:

Which is why I suggested connecting speakers to the MT-32. Seems
like perhaps the soundcard is not ready/initialized as you said the
MT-32 was getting MIDI data, but no sound was being produced. So
if the MT-32 is indeed playing then the problem moves to the
soundcard or trying to get camdusb to initialize later in startup.

Or something??

@mechanic

Quote:

Heh, i missed that part

Yes, that might be it, it might have to do with AHI not having initialized the soundcard yet and, after it did, not respecting the fact that there is an USB audio device already there (USB wise)

Hmm

@Raziel

Forehead*BIGPALM* - I am so thick

Please, somebody, come around my house and take the Amiga from me, i'm too dumb to use it.

There was nothing wrong with the installation and the hardware

Sorry for blaming anything else than my own retardness

I had camd.library NOT installed in SYS:libs, rather had i installed in in another libs drawer which is added later in the boot process to the system in user-startup.
One can tell that this draws every kind of strange behaviour afterwards.



The only thing i still have to test right now is that as soon as i start Mixer the soundcard goes deaf, but only for MIDI input, seems i have the input muted by mistake somewhere in the prefs.

Thanks for all the help guys

oh Raziel.... raziel....

@Raziel

So sorry.. I got here late.

It sounds like you've figured it out.. if I understand you correctly.

So, the way usbfd works is not quite like a normal CAMD interface.. It is called by the USB stack, NOT by CAMD.. when it finds a workable MIDI interface, it opens CAMD library just like any other application... NOT the same as the other drivers would. This also has the advantage of allowing the USB interface(s) to exchange complete messages, without CAMD having to parse them from bytes.

So, in reality, it's NOT even part of CAMD, it's just another program that opens CAMD for communication.
SO if your CAMD was not in the library path, it would SEE your interface, but would not be able to register it into the CAMD port list.

Like everything else, it makes sense once you know what it's trying to do.

I regret my slow reply, I have moved, and I am now working a job with ridiculously long hours, sometimes working seven days a week. I just don't have enough time to "play" anymore.

I can tell you that usbfd was developed with a Roland UM-1, and later with an EDIROL UM-880, so it should be completely compatible with them both.

I can also (happily) tell you that I have banked enough to get an A1X1000 as soon as they are available, and my very FIRST install will be CAMD and USBFD.. I have really missed my little playground.

I'm very happy to hear that CAMD and usbfd are still getting good use. I'm hoping to do some more application development once the new Amiga arrives.

Oh, and if someone needs to reach me, my nick at gmail dot com is checked daily.

LyleHaze

Oh, and thanks for the great laugh.. I needed that!
Nice to know I'm not the only one who makes occasional mistakes.