No sound or continuous beep w or w/o original sound on top on EMU APS

XDread

Hi and thanks for this project!

I just noticed that the following messages can be found in the event viewer:
"File replacement was attempted on the protected system file c:\winnt\system32\ksclockf.ax. This file was restored to the original version to maintain system stability. The file version of the system file is 5.0.2165.1."

The same message can be found for the following four files:
c:\winnt\system32\ksclockf.ax
c:\winnt\system32\ksvpintf.ax
c:\winnt\system32\ksdata.ax
c:\winnt\system32\ksinterf.ax

These files seems to be Microsoft Specific, although they are used by the kx drivers.

(Could this be the problem that these files is not updated during install?)

Anyways, the problem is still as follows:

This is a really annoying bug, and for those who have read my update (which is now removed) it seems like an coincident that I could fix the bug by stopping and starting the sound a lot of times. Hovever I have found the different beeping sounds and silences to cycle between each stop/start of a song in a period of 16. If you have silence a lot of times, then a low beeping sound, a high beeping sound then a low beeping sound it is exactly 16 clicks on stop/start to get the same sound.

My soundcard is an EMU APS as in " http://www.emu.com/products/aps/aps.html ". My system is of no importance since the very same error also was present on my former system, but my operating system MAY be of interest, MS Windows 2000, sp3 5.00.2195.

When my computer is fresh started everything works fine. Everything allways works fine during play.

The error does not allways appear during the following conditions, but when it does it is allways between two different sounds. (Between a program releasing a handler and reaquire the handler it seems.) That is between films in media player, songs in WinAmp or even between two beeps in Trillian which is a kind of "all in one" messenger. These are just examples. The clue is between, not during.

The error is pronounced as:
- No sound
- Continous beeping sound with or without the original sound on top.

It actually sound as if a (small) buffer with part of a former played song (that is a guess, it may as well be random data) is played continous with no update within the buffer. BEEEP in other words. The beep may sound different stopping and starting a same song, but the sound is alike during the whole play. When I stop play, beep goes away. When I start song, beep comes back. Perhaps in an other volume, but the same frequency. Different sounds may produce different frequencies.

The times I can hear the original sound on top of the beep is when the frequency is lower than the high pitched beeps, just as my buffer analogy.

I have had the following problem for a while now and I have upgraded with every new version released hoping the problem is solved now, but no.

I do not use the Drive Bay Input/Output Module since it is crap. It is not even connected to the base card. Therefore it is the analogue outputs on the card I am using. If I don't remember wrongly, the same problem existed also using the headphone out on that card.

I have tried to reset the settings and also to reinitialize DSP, none works. In Cubase SX using ASIO drivers everything works fine.

This problem is rather annoying because I have to reboot my computer every time this happens, and I have a feeling that it happens more often with the latest driver versions than the previous ones.

When we're at it I have also had some problems in WinAmp when searching within a song. (That search-bar.) Sometimes the sound level gets double and in only one speaker, but this problem is solved pressing the search bar once more and is therfore not important for me at all.

I'm pretty sure this is a buffer-handling problem, and now I have to reboot. *again* It seems like the SDK does not include the source for the driver itself... I have to investigate this further...

Hanz Petrov

Hello XDread.

I have an Emu APS working here without problems under Win2k SP2 with kX 3530, and file protection is active, so that is not likely to be the source of your problem.

Beyond that (and you may have tried all of the following, but for what it’s worth):

-Is hardware acceleration set to full (control panel->sound& multimedia->audio->advanced)?
-Do you have DirectX 9 installed?
-Have you tried a clean install (i.e. re-install win2k, or install on a separate partition)?

Regards,

/Hanz

p.s. You can watch kX debug output with Numega’s monitor.exe – might hold some clues..

XDread

Hi! Thanks for an answer!

You say you have no problem. That's interresting since i have used this card on two different computers and also with two clean installations.

All HW accellerations are set tu full.
DX 9 is installed and working well on my computer.
Yes, this is a new and clean installation.

I think i'll have a go at monitor.exe as you are mentioning, that's probably my only solition right now. Especially since other users with the same configuration does not experience these problems. (I see that you have sp2, I am using sp3, but that should not be a problem...)

Again, I am really happy that someone did come with an answer! Thanks!

XDread

... monitor.exe was rather not possible to find.... Seems to be some tool incorporated in a commercial driver development package.

Seems like I have to throw away my sound card. Better Buy something else like Pro-tools or something...

XDread

Update:

I formerly stated that this error did not happen with ASIO. That is wrong. I get this bug using ASIO drivers with Cubase SX as well as with other programs.

XDread

Further update:

I degraded the driver version to build 3520 and the bug is pronounced a little different now. Funny actually....

In Winamp while playing a song, clicking the search bar various times:
Music, beep, music, music, music in right speaker only (reversed?), beep, music, music, beep, music, music, beep, music, beep, music, music, beep, music, beep, music, music, music, music, music, music, music in right speaker only, music, music, beep, and so on....

The reason I displayed so many "tries" is to see if there is a repeating pattern as in the reported bug. It is not. Actually the error is much better in build 3530, but when it strikes, it is harder to get rid of it...

In other words, this is a highly reproducible bug, but you may have to try a *lot* of times to manage to produce it...

Now I am more confident that this is not a fault in my sound card as I eventually begun to be afraid of....

More updates will follow....

...and then for the possible ASIO fault as posted in the "General" forum...
...it appeared to be a fault in an effect in Cubase SX...

Hanz Petrov

http://kxdev.narod.ru/tools.html

;-)

/Hanz

p.s. Which version of winamp?

XDread

Thanks!

Winamp? The newest there is... But that is not the problem for sure. It just happens to be THE tool to help me in and out of the bug... The bug appears in whatever application you want if it's appearing.

What happens is probably that Winamp during each click releases some handles/resources or what it could be and reclaims it before playing the new section. That is probably why the error only appears between switch of song, application or two sounds within an application. (Releasing the "claim to use sound resources" or aquire it.) It also seems to be a fault in the buffer handling as I have suggested above. I no longer think that the fault is my own card since the problem is different on different builds of the kX drivers.

Every application handles this (resource claim/release) differently. Mediaplayer 9 for instance, clicking the searchbar during a film won't produce the same bug. Stopping and starting a film on the contrary. I must admit I don't know the exact details of how the system communicate with the drivers and the drivers communicate with the sound card(s), but I know a *lot* of these things enough to make an educated guess of what happens. (I have been involved with a lot of different aspects of computer science since about 20 years ago ranging from how a computer is created and is internally communicatin to have programmed a chipset in assembly language.) Just to put some weight to my words... Then you can judge for yourselves!

XDread

I had to edit the "ImagePath" registry key to get monitor.exe to work. IE set the path to winnt/system32/drivers directory.

The output produced by monitor.exe is pretty simple. Whether I get a beep or I get music the output is like this:

"waveout open[f8569d68:f7dae000]: open(freq=48000; n_channels=2[2]; bps=16; type: ;"
"waveout open[f8569d68:f7dd2000]: open(freq=44100; n_channels=2[2]; bps=16; type: ;"

The addresses may vary, but there is no difference between beep and music.

...but this is rather little amount of data to tell anything of what goes wrong...

Eugene Gavrilov

please post your hardware configuration here

/Eugene

XDread

HW Config as follows:

MOBO: MSI K7N2G
CPU: AMD Athlon XP 1800+
RAM: 500MB
HD: 200GB

(But had the same problem on a HP VL8i 400MHz desktop computer.)

----Hardware / Software configuration---------------
Driver Name:kX Audio Driver (Debug)
Driver Date: Mar 2 2003 03:57:33
Driver Version: 5.10.00.3530 - debug
DB Name:none
EMUAPS [d400]
PCI Information:
Device: 21102 Subsys: 40011102 ChipRev: 4
Card has MPU device
Card is EMU APS
Card HAS NO AC97 codec
Port: d400 Irq: 8
Playback buffer: 20a0
Record buffer: 4000
Number of AC3 buffers: 8
Tank memory: 256 kb
----------------------------------------------------

I have had this problem for the last couple of versions of the kX Driver, but there is a difference:
On earlier driver version (i.e. 3520) the problem is more easily reproduced and is also easier to get rid of.

I have a trick to get rid of the problem for a while (or to reproduce it), and that is to start a song in WinAmp and then repeatedly click the search bar. When the problem occurs, it seems like there is a cycling between 16 buffers. If you say that n stand for no sound and b stand for beep and you hav got the following 16 steps in sequence:
n,n,b,n,b,b,b,n,b,n,n,n,n,b,b,n (16 clicks on search-bar) then this is the same sequence rotating over and over.

On version 3520 I have not found a pattern. It is completely random.

Also I occational experience louder sound often combined with no sound in one of the speaker, but that is usually easy to fix, i.e. stop and start or press the search-bar. This problem is not only within WinAmp, but in all other applications like Mediaplayer, Messeger, Cubase SX (using ASIO) as well.

Eugene Gavrilov

all this looks very weird
well, I suggest you try the following:

1. if possible, remove 256Mb of your memory -- (that is, try to work with 256Mb)
2. re-install your Windows
3. re-install directX
4. try using any other sound card
5. open kxmixer's 'voice management' ('analyzer') window and check if there's anything of interest
6. switch your winamp to directx mode
7. re-install the drivers. make sure you use the correct distribution CD / hdd folder [ksclock.ax-related message is -very- strange]

anyway, at the moment we cannot reproduce / solve the problem

/Eugene

XDread

1. It is kindda hard to divide the one 512MB chip that I am using. I am not sure what you want me to test by doing so, but if I am to use only 256MB RAM, then I will have to go to the computer store to buy smaller chips.

2. As I said earlier, I have run an entirely different system with the same audio card. That means all components is replaced since then, and I did have the same problem on the other system as well as on my present one. I really don't think reinstalling will help at all.

3. AS for number 2. DX has been upgraded several times (even uninstall and reinstall) with no help earlier.

4. I do have a on-board audio card, but it does not support ASIO.

5. Uhm.. The analyzer has never worked, except for the VU meter that obviously has been removed in the later versions of kX. The view I can see is just a lot of cyan bars and a "frozen" view of the moment just before i opened the analyzer. With that I mean that when there is music, there are some patterns, but they don't move. When I close and re-open the pattern has changed, but I still haven't seen any movement. I have never seen any changes with my own eyes yet...

6. I could do that, but the problem is present and (re)producable in other applications as well... To mention a few, Cubase SX, Media Player, Trillian... But I will of course try!

7. The drivers has been deinstalled and reinstalled a lot of times in a lot of different versions already...

"[ksclock.ax-related message is -very- strange]"
Do you have any suggestions on how I can force an installation of that file?

"anyway, at the moment we cannot reproduce / solve the problem"
Well, then there is nothing that can be done! I fully understand that you are unable to fix something that you are unable to reproduce yourselves, it's just bad luck for me. I will have to buy another card then...

Eugene Gavrilov

1. there might be problems with certain chipsets and large amounts of memory
so, I wanted to know if you could test this issue

3. "even uninstall and reinstall" -- well, as far as I remember, it is impossible to -uninstall- DirectX if it is installed. are you sure?..

4. on-board audio: try to disable it

5. the 'analyzer' is actually a 'voice management monitor'. each bar represents a hardware-allocated voice.


.ax file: it seems that your Windows distribution is not correctly located
if you installed any service packs, make sure -not- to insert the original CD [and rename the Windows installation folder, if you installed from a HDD] -- if you are asked for a file, browse to \winnt\system32

I suggest you re-install service pack 3

/Eugene

XDread

"1. there might be problems with certain chipsets and large amounts of memory
so, I wanted to know if you could test this issue"

Ah. That is the only thing common with my present and former system except the audio card... I will try to borrow some lesser ram from a friend!

"3. "even uninstall and reinstall" -- well, as far as I remember, it is impossible to -uninstall- DirectX if it is installed. are you sure?.."

Ehm. You made me very unsure there! Well on this system I have only upgraded and reinstalled DX, but on my former computer I had DX debug drivers and SDK installed. I think there was a possibility to either deinstall the debug version of the drivers or at least replace them with production ones. Well, my point is that this bug has been on two different systems (but with the same audio card).

"4. on-board audio: try to disable it"

I have already disabled that. I disabled that option even before installing my os.

"5. the 'analyzer' is actually a 'voice management monitor'. each bar represents a hardware-allocated voice."

Ah. I thought it supposed to be a spectrum analyzer or something...

Here are some pictures:
I. Normal audio playing:
http://www.galaasen.com/kx_normal.bmp

II. While beep playing (the problem):
http://www.galaasen.com/kx_beep.bmp

As you can see, there is no difference.

III. Other problems:
Music plays loud in both speakers:
http://www.galaasen.com/kx_loud_in_both_speakers.bmp
Loud in left, normal in right:
http://www.galaasen.com/kx_loud_in_l...l_in_right.bmp
No sound:
http://www.galaasen.com/kx_no_sound.bmp
Also I sometimes get no sound in just one speaker, no sound in one and loud in another...

I used to solve this by restarting the media, but changing the volume meter in the app fixes the loud/no sound problem. This makes me suspect two different bugs that can not be connected.

".ax file: it seems that your Windows distribution is not correctly located
if you installed any service packs, make sure -not- to insert the original CD [and rename the Windows installation folder, if you installed from a HDD] -- if you are asked for a file, browse to \winnt\system32"

I did install from a CD, and have never used it later for installation of other devices drivers. When applying service packs I can not remember ever having any questions regarding replacing old drivers, but as a former professional computer support crew member, that was allways the rule when it came up such situatons installing service packs. (On nt4.0, w9x.) I have never seen it on W2k and XP is something that I will *never* touch!

"I suggest you re-install service pack 3. "
Ok, I will, but as said, I had the same problem on my former computer system.

...and my system otherwise works as a charm!

Lusive

I just recently installed kX on a friends machine, and i got this exact problem (i'm not to sure about the repetitions of 16 or whatever though)

It is a brand new SBLive 5.1 with these specs:

----Hardware / Software configuration---------------
Driver Name:kX Audio Driver (Debug)
Driver Date: Mar 2 2003 03:57:33
Driver Version: 5.10.00.3530 - debug
DB Name:none
SB0102 5.1 [ffc0]
PCI Information:
Device: 21102 Subsys: 80641102 ChipRev: a
Card has MPU device
Card model is '5.1'
Card HAS AC97 codec
Codec name: SigmaTel STAC9708
3D Extension: SigmaTel 3D Enhancement
Codec is 2.0 compliant
Capabilities[6940] :
DAC resolutions : -16-bit- -18-bit-
ADC resolutions : -16-bit- -18-bit-
Ext Capabilities [80]: -PCM surround DAC-
Port: ffc0 Irq: 12
Playback buffer: 20a0
Record buffer: 4000
Number of AC3 buffers: 8
Tank memory: 256 kb
----------------------------------------------------

It's a windows 2000 sp3 system, freshly installed, intel motherboard chipset, onboard video (no onboard sound). When i say fresh install i mean i installed windows without the card in the system, then i renamed wdma10k1.inf so it wouldnt detect as a sblive, then i put the card in and installed kX.

Is there any other info that you guys would like to have, to investigate this problem? (i know its hard when you cant recreate it)

This is the second time i have come across this problem (both on friends systems thank god), the first time was on an old, non-5.1, live value in a gateway2k system. I can get specs on that aswell if needed.

I am using an SB0060 on a via kt7e motherboard, absolutely fine. no worries, you guys rock, i love these drivers. I will try swapping my card into my friends machine, but i am sure it will create the same problem as my hunch is its some other hardware (like mb chipset) thats conflicting.

Thanks,

Lusive

Max M.

>Port: ffc0 Irq: 12 - this is the worst configuration for audio device that could ever exist... well, typically happens with W2K/XP and ACPI enabled... try to move your card to another PCI slot... and disable some unused devices (like USB for example - do you use all four/eight?)

__________________

XDread

Are you suggesting this may be an IRQ conflict? (I'll check!)



*** After checking ***

I can see that my on-board graphical adapter does indeed share IRQ 16 with my sound card. (My system reports using irq 16, while kX mixer reports IRQ 8...)

XDread

Additional information (or a question at least):

In the device manager, on location, it says that the card is located in PCI slot 8 (PCI bus 1, device 10, function 0). The resources tab says that I/O range is D400-D41f, IRQ 16.

The HW/S CFG says:
----Hardware / Software configuration---------------
Port: d400 Irq: 8
----------------------------------------------------

IRQ's don't match! What could this be? A bug in kX? I can see that PCI slot does have value 8? Could the value that kX about box displays as IRQ really be the PCI slot?

Max M.

Interrupts shown by Win XP/2K are virtual... the ones shown by hw/sw cfg are provided by OS too and they are more correct maybe although still virtual... this is what ACPI is... There's no IRQ16 in IA32 PC Architecture.. there're only 16 interrupts in hardware (from 0 to15)... So with that OSes you can't be sure is there conflict or not - for example windows shows you that video uses irq 18 and audio irq 22.. haha.. actually in real they probably use the same irq... well, from expirience i can say that d000, d400, 1000 or similiar (two zeros at the end) are somewhat "normal" addresses but ffe0 is not... Well, it's a black magic actually - i wonder if someone knows how that works (except MS maybe)...

__________________

XDread

That is actually rather strange... If the last byte should be 00 then it means that the memory multiplexer is not connected to the first 8 bits. (bits 0-7) That is every card you insert into the computer should have 256 addresses each which they can communicate through... Not all of these is nessesary in use.

Of course things have gotten rather complicated the last years. (There is probably no address line 0 anymore for instance, and both memory management and communication with connected pheripherals has developed during the last couple of decades and with plug and play.)

Max M.

-> it means that the memory multiplexer is not connected to the first 8 bits. - well, it is just means that bios typically assigns such addresses to most of audio devices i have ever seen.... nothing more.. and address ffe0 shows that XP/2K somewhat have overriden hardware assigned configuration... somthing like that...

__________________

Lusive

I took out all unneccesary cards in his system, (modem, net card) and moved the sblive to the top pci slot. It changed the port to ffeh, as opposed to ffch.
As i cant disable ACPI on his system i don't know what to.

I do remember i tried to instal windows 98 on his system and it wouldnt install it because the system had "run out of motherboard resources" or something. i had never got this error before EVER and so i put win2k on. I'll have a look around for a bios flash for his board to disable acpi (but as its a mobo thats rebadged as Dell, that may be hard).

Rox271

FYI I've got exactly the same problem.
I did play with IRQs, moved the Sound card from a slot to another etc... I have Win XP and just upgraded from DX8.1 to DX9. Still no success.
That could be connected to the Direct X secondary buffers. When the beep occurs, I cannot reset the configuration. I get a message that tells me it couldn't allocate DX secondary buffers (the message isn't in english, so this is my personnal translation).
May be there's a tweak with buffer settings?

SkinJob

Just thought I'd report that I'm experiencing the same problem. I can confirm that the problem occurs from both DirectX (WinAmp) and ASIO (Cubase SX) applications. I can also confirm that starting & stopping WinAmp 16 times will correct the problem. I haven't upgraded to DirectX 9 yet (DirectX 8 was such a disaster for Cubase users), but this didn't seem to correct the issue for other users in any case.

Reset settings and reinitialize DSP do not correct the problem. The only way to recover is to do the WinAmp 16x trick or reboot.

If it helps, I don't think I ever experienced the problem with 3528 or earlier.

I'm sure this one is going to be a bitch to track down. I don't have a lot of time to start tearing my machine up right now, but if there's anything relatively quick that you'd like me to try, or any other info you need, please let me know.

BTW, the developers are doing a really great job. I'm truly amazed at what's been accomplished. I sure hope you guys can figure this one out.


Config:
Dell 8200
P4 1.4
768MB RDRAM
XP Pro SP1 + all critical updates


----Hardware / Software configuration---------------
Driver Name:kX Audio Driver (Debug)
Driver Date: Mar 27 2003 00:01:41
Driver Version: 5.10.00.3531 - debug
DB Name:LiveBay0
CT4760 10k1 [ece0]
PCI Information:
Device: 21102 Subsys: 80401102 ChipRev: 7
Card has MPU device
Card HAS AC97 codec
Codec name: SigmaTel STAC9721/23
3D Extension: SigmaTel 3D Enhancement
Codec is 2.0 compliant
Capabilities[6940] :
DAC resolutions : -16-bit- -18-bit-
ADC resolutions : -16-bit- -18-bit-
Ext Capabilities [200]: -slot/DAC mappings-
Port: ece0 Irq: 10
Playback buffer: 20a0
Record buffer: 4000
Number of AC3 buffers: 8
Tank memory: 256 kb
----------------------------------------------------

TheDingo

Reporting in to say I too have the exact same problem. There is no one cause I can find to make it happen, but it seems to be common when my PC is on for long lengths of time. Open/close Winamp 16x works for me as well. I have had this problem for a while with quite a few different versions of KX, even after clean installs. Here's my specs:

MSI KT7Pro
Athlon XP 1800+
512mb RAM
240 GB
Sound Blaster Audigy
XP Pro SP1 & DX9

One thing to note is that I install KX over the Creative drivers so that my midi/gameport will work.

Eugene Gavrilov

the bug is probably fixed in 3532b (development release)

pls check this

/Eugene

XDread

Where do you get that build?

The UpperTaker

http://kx.maincore.nl/kxdrv3532b-update.exe

Link for 3532b

XDread

No it didn't work... (Can you hear me cry?) :.( sob

For you other people out there, I would like to hear if you solved your problems with this build?

For you kX devel people's information I can say that there is no differences between the last build versus this beta regarding this issue.

Also: I am now so tired of this bug that I am going to promise you twice the suggested donation if you manage to fix this bug. (I know you are doing an excellent job!) I can't afford a new card, so it is well worth it!