Graffias, I just sent you an email about this, but everyone else, I took out the sam 460ex_setup (hack) out of my wbstartup and cold reboot. Now, the static is gone from my delitracker when I play mods. The audio still sounds like it has no filters on it and a bit crunchy but this got rid of the static for me.
Juliet and Maya both uses the same Envy24HT-S chip but other chips are different so don't know if our Envy24HT driver works with Maya or not. It looks like Juliet XTe (PCIe) model is not easy to find. Besides the Dutch site you can try www.thomann.de also. Google is your friend too.
Rock lobster bit me - so I'm here forever X1000 + AmigaOS 4.1 FE "Anyone can build a fast CPU. The trick is to build a fast system." - Seymour Cray
The PCI-E to PCI bridge on this Envy24DT sound card is the AsMedia ASM1083. Doing a google search turns up instances of people complaining of stuck interrupts with this chip. This makes sense as it looks like workbench is stuck waiting for something when it loads. I wonder if this is something that can be adjusted or compensated for. Apparently this ASM1083 chip is pretty common among PCI-E to PCI bridge chips.
As an update, I tried again with the Envy24HT PCI-E Card and also a low profile CMI8738 card on a PCI-E to PCI adapter. This adapter happens to have a different PCI-E to PCI bridge chip. This eliminates the possibility that it is just an incompatibility with just one brand of chip. Regardless, it seems to be the same problem, but I think I have a better understanding of it now;
I know that AHI sees the sound chips through the PCI-E to PCI bridge chip because they appear in AHI's list of devices installed. When selected, AHI expects the chips to work like they are connected to a PCI bus. I have noticed that when the chip (for instance, the CMI8738, which we know to be a well supported sound chip) is selected, it behaves like normal, except when you go to play a sound. The card does not play the sound and the program that sent the command to play the sound becomes hung up as if it were waiting for something to finish or return.
As an example: I would select the sound card in AHI and the Mixer, and then open TuneNet. When I press play, nothing happens. The program does not go through the motions of playing, it just hangs there. I cannot close the program. What is it waiting for? Why can't the signal reach the program from the sound card? The driver can see the chip and Ranger can see the vendor ID and address, etc.
I just had a thought: My earlier problem of the Envy24 sound card keeping Workbench from loading is probably due to the startup sound. I should try to disable the startup sound while the card is in there so I can at least look at the list in AHI showing the Envy driver info.
I don't think that PCI-E to PCI bridges of any kind are supported. At least that's the evidence that I've gathered. I can get the AOne/500 to "see" the devices but not use them.
Currently the AmigaOS4,1 kernel for Sam460ex boards doesn't support I/O over PCI Express, it support only Memory Mapped access. That's the reason why gfx cards like the Radeon HD works, and sata/audio controller don't, since the laters require I/O accesses. Using a PCI-PCI-E bridge doesn't solve the problem, if the PCI card requires I/O accesses to work too.
Using Ranger it's possible to check if the PCI-E (or PCI) board requires I/O accesses, just select the PCI Bus tab, choose a board from the list on the left, and in the Resource Ranges check for the Flags files, if it reports only IO, than the board won't work with the current AmigaOS4.1 kernel.
In the Ranger screnshoot above, that audio board use only I/O, therefore it won't work atm.
I have the slow audio problem on my SAM460 (amongst other issues).
PCI info shows up as "Unidentifyied Multimedia Device" Vendor code 0x1002 product code 0xaa90. Don't know if this is some new device Acube is now using as apparently sounds works for other people.
Older driver: Sound is ok except for playback occuring at about 95% speed. Current driver: Sound plays at the right speed but there are crackles and pops
PCI info shows up as "Unidentifyied Multimedia Device" Vendor code 0x1002 product code 0xaa90. Don't know if this is some new device Acube is now using as apparently sounds works for other people.
That device is: Turks/Whistler HDMI Audio [Radeon HD 6000 Series]
It's the HDMI audio device on your Radeon HD 6xx0 graphics card. There is currently no driver support for this device.
t's the onboard sound which I have my speakers plugged into. That must mean it doesn't show up as a PCI device at all then? All other devices I in Ranger I account for.
Anyway, I'd be interested to hear how many SAM460 people have problems with sound.
t's the onboard sound which I have my speakers plugged into. That must mean it doesn't show up as a PCI device at all then? All other devices I in Ranger I account for.
IIRC, the audio is done by the SM502, which also happens to be the on-board graphics card (best described as a "multimedia chip"). Depending on how intertwined the graphics and audio functionality of the SM502 is, this could explain why it's taking longer than usual to get the audio working 100% correctly. That would mean sharing interrupts and possibly even clock sources with the graphics driver. An audio chip that appears as a separate PCI device is probably easier to program than one that is integrated into something else.
Quote:
Anyway, I'd be interested to hear how many SAM460 people have problems with sound.
I have audio issues with my Sam460ex too. I haven't tested it extensively, but I get audio gaps when trying to listen to CD audio (via AHI, as my DVD drive is one of the newer ones that don't have an audio out).
EDIT: The slight audio gaps occur only with PlayCD. Playing the CD via Tunenet works fine. The gaps still shouldn't happen, though. Also, I am using a beta version of AmigaOS, so you're better off getting feedback from others.