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

dripgoss

I too would get an intermittent loud beep (only followed by a BSOD!) at various points during operation when I first tried to install 3553b for my SBL! Platinum (CT4760) in WinXP SP1 on my P4/845PE/512DDR400 system.

Then I uninstalled kX driver, uninstalled all (even hidden) devices in "Sound Video and Game Controllers" in Device Manager, ran msconfig (which win2k doesn't have), trimmed all unnecessary services, did all the tweaks on musicxp.net, and then disabled all unnecessary ports and features in the BIOS.

Then I re-installed 3533b in safe mode, rebooted and no more beeep. *HOWEVER*, I'm still stuck with distorted/crackled playback in Sonar XL 2.0, but that's been with me since I installed this infernal sound card!!

Hope this helps!

__________________
Custom "All Purpose" Rig
-------------------------------
• WinXP Pro SP1 (Tweaked and Partitioned)
• Nuendo 1.6.3
• Sonar 2.2
• Various Plugins

• P4 2.4B(C1) @ 2.88Ghz Stable
• 512GB Corsair XMS PC3500LL (DDR400)
• Albatron PX845PEV Pro Mobo
• (4x) 7200RPM Maxtor FMB HDD's
• Radeon 8500LE 64MB (Cat 3.2, No Control Panel)
• Albatron 17" DVI LCD
• Echo Mia
• SB Live! Platinum w/ kX 5.00.3533b Driver (5.1/Stereo Outputs)
• Logitech z640 (For Now)

• Clavia Ddrum4SE kit
• DW Acoustic kit / Paiste Sig Cyms
• Various Snares
• Various Guitars
• Oxygen 8 Controller
• Various Mics

Rox271

Hi Eugene,

I haven't used kx drivers for a few weeks and was back on Creative and these ASIO dx drivers from that german supplier... Though, not as good as kx
Anyway, I just reinstalled the latest 3533b and so far so good but I know from previous experiences that I'm easily pleased so I will refrain myself from being too quickly excited . Is it the same version as the 3533 we tested before?

Aslo, I can see a new combobox on the main screen named AC3 and 3 choices of SPDIF. What is this?

Thanks again for your work. Actually I was wondering who you guys were; will you publish some kind of bio on the kx web site one day?

Eugene Gavrilov

3533b is not 100% the same, but there were no changes related to this issue at all
I'm still waiting for somebody to test it more thoroughly (the mentioned above technical information 'for volunteers')

ac-3 combo box: it is for ac-3 passthru, but it seems to be useless at the moment (for 10k1 boards)

--
who are we?
well, programmers/musicians/sound engineers
[mostly] located in Russia. partially in Canada...

/Eugene

Rox271

Hello Russian music/technology lovers ,

Well, I had the beep back this morning. I'm getting a little bit confused on how to help on this issue (This thread is getting very large). I did a clean install a couple of weeks ago and didn't notice any particular improvement. I do not have monitor installed anymore but we couldn't see anything special anyway.
The problem is this issue is completely random. I can spend 4h with no problem and all the sudden it becomes a nightmare. I don't mean to sound awkward toward you but wouldn't it be a good idea to publish sources in order to allow more people with different settings to check them?

XDread

Hmm.. No BSOD here! Everything works just fine, beep playing instead of music, that's all. I don't think this is the same error!

XDread

I am going to test it, but I am in a moving phase for the moment. When I get my computer up and running in my new place, I'll do some more testing!

XDread

I have asked the very same question earlier, and I doubt they will! And to be honest I really do understand why! There is a LOT of work involved creating this driver and I must admit I would have problems just giving away all the good work just like that. At least when there perhaps is no plan for what to do with the source in the future. (This is not an Open Source project.)

For the last problem, I have been wondering to get a ddk to see if I can manage to create a minimalistic driver just to test if the same problem would occur in that case. But then I would have to learn how the whole driver system works and how the driver communicate with hardware and system, not to say where to find specs for the sound card. Interresting, but I'm not sure if I have that much time for the next couple of months...

And for the randomness: YES, yes and yes! But as stated earlier:
Build 3514 is highly unstable. That is there is about a 50% chance to get a beep or a sound when playing music.
Build 3520+ something is more stable, that is: When playing, it usually is playing for a while, when beeping, it is usually beeping for a while...
Build 3533b has the same problem, but seems to be more stable regarding how often the problem may occure. But I might be wrong.

escaflo

Ok, I have posted a previous thread before XDread pointed me in the direction of this thread. I do have the same problem when I was using 3533b. I have since gone back to using Creat**** drivers since. I would very much like to see this problem solve too. My problem is not far from the one that XDread is facing.

vik

Hi everybody

I write today may be this could help. So :

I experienced few weeks ago something like you describe.

I dont know if it's just the same but it looked like:

1 - I would not call it a beep sound but "a kind of" buffer underrun (or overrun ?) sound.
Like I ve read in this thread, just like "looping a few samples" instead of reading next and randomly producing an annoying strange frozen sound in the wav files. Just like if a resident background software was stealing all system ressources for a second.

2 - My system is multiboot with 98SE and XP pro SP1 and it happened only in XP

3 - It happened with any version of KX drivers and any kind of driver (asio, directx, wdm...)

4 - It happened with any music software. None of them could play a single long wavefile without doing this noise

5 - The frequency of the bug ??? Random. But most often once a song minimum.

I tried and tried and tried everything I could. Suspected anything. No way.
Everything else in the system was working fine.

Doesn't it sound like your problem ?

Today I solved my problem. But I can't tell what exactly solved it.

After weeks of testings, upgrading all I could, I suspected something relative to ACPI and APIC Mode
(It was first installed as PC standart).
I formatted partition and reinstalled XP in "Monoprocessor ACPI mode".
Reactivated APIC Mode in the bios and since, Not a single problem.

Was it PCI irq sharing problem? I didnot changed cards placement. But in ACPI mode irq assignment change.
I've read somewhere sound issues between SBLive and 3Com Ethernet Controller.
I got one integrated and I remember that in PC Standart Mode SBLive was sharing one irq with the 3Com chip.

Now
Win XP reports My Live Cards on Irq 16 and 18
KX driver reports SB060 Port: 9000 Irq: a (funny irq a !!) and
SB102 Port: 9800 Irq: 9


So you see, I ain't got "The Answer" but this is my brainstorm for you and I hope it will help.

Keep the pressure KX team. Perfect!

On The Day When Music Pays A Bit
For Devotion,
Creations
I Will share With You.
Just Like You Do.
Thank you

Serious. Promised
Vik

The machine :
asus a7n8x D, Athlon XP 2600+, 512Mo Ram (Dual Channel Mode), Ati Radeon 7500, Maxtor 40Go, Samsung 40Go,
SB Live SB060 with LiveDrive2 and SB Live SB102 OEM
KX Driver 3533 final +Fix

vik

Talking about Creative

I recently appreciated the KXsetup enhancements on multiple sblive cards install with different drivers.
When you did that, it was just what I was waiting For. Thanks Eugene
But I experienced twice the same problem on a clean reinstall of windows XP with 2 SBLIVEs.
It's not a KXbug but it's also an install f_____g problem.

When you install XP on a formatted HDrive It detects SBLives and install by default creative WDM drivers.
Win XP install crashes when it installs the second SBLive
It sounds that these drivers (creative WDM drivers) can't manage 2 SBLives (Thanks Creative)

The only solution to that?

Open the machine
take off one soundcard
install XP
Install KXdrvs
Put back the second souncard
Reinstall KX Drivers

Just a bit painful install


Anyone clever enough to find another solution than opening the machine?

Thanks

Vik

Eugene Gavrilov

>> My system is multiboot with 98SE and XP pro SP1 and it happened only in XP

it seems to be the real 'clue' to the problem: winXP memory or PCI management (including IRQ issues)
I'll try to release even more debug-oriented version of kx.sys (3534)

/E

XDread

Hi!

I don't think this sounds like the problem we are experiencing in this thread. (Anyone disagree?)

As you say you are experiencing a buffer underrun problem, but I can't say anything about the reason why? Too small buffers?
Another thing could be that you are experiencing some kind of irq problems or something, but I am not sure!

Rox271

Hard to say. Could be IRQ, I did explore this possibility. I've got a Dell Dimension 8100 and noticed that I have very little choice on IRQ. This depends on the PCI slot. It looks like they are clustered and I must share IRQs depending on card placement. I moved to SB Live in a particular slot in order to make sure that the IRQ is not shared but couldn't see any improvement after that. Also since it is the BIOS that manages IRQ assignment, it shouldn't make any difference if one uses Win98 or WinXP right?
I tend to agree about a buffer issue. No doubt that the memory management is very different between XP and 98.

escaflo

Just did a full reinstall of Windows XP and installed KXDrivers. Everything seems fine so far.

Didn't know how to disable ACPI until 15 mins ago. Just reinstall the "Standard PC" driver. Now reinstalled KXDriver. Just hope that it will be fine also.

Anybody knows how to install as a standard pc when doing a fresh installation?

vik

Hi

That's the way

The mode Standard-PC can be selected during installation.
when you are asked to hit F6 for updated SCSI etc. device drivers (blue screen). Press F5 instead and choose Standard-PC.

VIK

escaflo

Thanks. Would it make a lot of difference if I switch to Standard-PC after installing? Or is it better to just install it as Standard-PC during a fresh installation?

vik

It is not recommanded to switch to pC standard after installation of winXP
Simply because all cards and peripherals will have to be reinstalled after reboot.
If the Windows XP CD has got all the drivers you need it may succeed but if you got recent things (like nforce2 chipset motherboard)or special cards your system won't reboot or endless loop

escaflo

Manage to do a switch without any problems. Phew. Thanks for the tips anyway.

jonasb

Was hoping there would be a breakthru by now since this bug can get pretty annoying when your editing a large sound file and it gets in the way.

I still find it interesting that by using the tabledit program, which generates midi events, the problem can be cleared after a handful of notes. I don't start and stop anything, just start playing a piece of music and after a handful of notes the bug goes away. I don't think 16 is any kind of magic number for retries. I think its pretty random. Why would sending midi commands to the synthesizer affect the output of the sound card?

I do think it is a repeating buffer problem becuase until the sound clears up, each note played by the sound card does change the tone of the beep a little. Its sounds just like a small section of sound selected in Sound Forge and looped continously on playback. The different flavors of silence vs beep and clicks that people are talking about would then be a function of what is in the buffer when the bug kicks in. Would knowing the frequency of the clicks give you any insight into what's stuck?

escaflo

I did a complete reinstallation of Windows XP with Service Pack 1 using Standard-PC. My computer have been on for 1 day (close to 2) with constant switching of applications between dvd player and media center. So far so good I have to say. The only difference between this installation and my previous setup was that I never installed any updates from Windows, never install DirectX 9 and also ACPI. I am not sure whether the problem will crop up again but so far it seems fine. Will update this thread with as it progresses.

XDread

I don't think so either, but when testing I have found the bug to end on a multiple of 16 (32) "notes" played. This may be a general rule, but I don't think it is absolute. Actually it is rather rare that the bug lasts for "only" 16 (or 32) voices, but usually for more.

I would like someone other than me to try the 3514 build. (Yeah, it's old.) Then try to reproduce the bug to see what happens. On my computer the bug or normal play happens at about a 50% rate per voice.

Yeah, it does, doesn't it? On my computer the frequency is all the same, but the "shape" or whatever does change. I suppose that is because random data is played in the short repeating loop. Random MAY be actual sound data, but only enough to fill the small buffer. It is impossible to hear whether it is the actual data or just a non-cleared buffer.

Please do! I am pretty sure your problem will show up again! Regarding Direct X 9, I must say that I have run DX8 minus before having the same bug. Therfore I don't think that is the problem.

escaflo

Com's been up and running for 2 days and 13hrs 45 mins. So far so good still. Have watched DVD, play Mp3s, stop using audio for almost 12 hrs. Let's hope it stay this way.

Eugene Gavrilov

>> Why would sending midi commands to the synthesizer affect the output of the sound card?

because (a) the synth uses the same voices as Wave (well, the same -set- of voices) -- watch analyzer window for details (b) because it seems that the 'beeping' wav channel isn't accessing PCI bus correctly (probably due to chipset-PCI-processor-Win2k/XP issue).

>> don't start and stop anything, just start playing a piece of music and after a handful of notes the bug
>> goes away. I don't think 16 is any kind of magic number for retries

this is important
do you mean the following:
start audio playback -> noise -> start (some) MIDI notes (w/o stopping wave) -> bug is solved?
if yes, then the problem is definitely PCI-related...
(or, probably, you mean: start -> noise -> stop -> MIDI notes -> start wave -> fixed?)

open kX Mixer's analyzer and watch the changes when you play MIDI --- note channel allocations

>> I would like someone other than me to try the 3514 build

I have already mentioned many times, that there were no significant change between 3514 and 3534 related to this issue! what is different is channel allocation strategy. playing MIDI or starting/stopping wav simply re-allocates PCI channels. 3514 used to allocate previously used channels, while 3534 allocates all the 64 hardware channels one-by-one

/Eugene

escaflo

I think it is ok for me to say that I've fixed the problem. How? I don't know. The problem haven't appear yet so far and I believe that it will not appear again. (I hope).

XDread

I would just put in that if there is a beep, then there is a beep for ALL audio channels currently playing.

(I haven't got to test anything more till now, I'll try to see if I get the time for it soon.)
Ok, I'm sorry, but there IS a difference on my computer... All I am saying is that this perhaps might give a hint. (I'll try reinstalling the 3514 driver to triple-check...)

Update:
I have triple-checked and yes there is a significant difference between those versions even though there are no significant change between 3514 and 3534.

(But you are (of course) rigth in that the first 2 or 4 channel are reused!)

I noticed one thing that might be of interest:
While playing a song in WinAmp, I tried to start and stop a movie played in MediaPlayer. What I noticed is that the beep starts when pressing the STOP button. (Beep because of a song played in WA.) Then the usual beep is for both WA and MP. Beep is also removed on a click on STOP. It seems that the beep is not appearing when starting a song, but when stopping it.

(I'll post an update on spy later now that I easily can reproduce the bug.)

escaflo

Just to drop a note. No problem encountered so far. I think my problem is solved. 9 days of uptime with no reoccurence

XDread

No reboot? I bet that you will encounter the problem again when you have to do your monthly reboot of your Windows system!

But for your sake, let's hope not! I think I'll stick with my 3514 build. There at least this bug is predictable.

escaflo

I have done several time reboot before this (installing softwares and drivers and stuff like that). So far it has proven to be fairly stable for now. But of course, I am not hoping for it to happen or anything. Hope you can solve ur problem soon too.

XDread

Update on SPY test: (Tested on version 3514 since the bug is very easy to reproduce there.)

Things I have noticed:
While playing, information about different other channels than the two currently playing is displayed in a third row. I assume this is a former bug that is corrected in the latest versions.

Nothing strange with the other registers except:
Register 0x0c:
While playing correct, values for voice 0,1 are 0x00036002, 0x03577fff.
While beeping, values for voice 0,1 are 0x03577fff, 0x03577fff.

Register 0x0d:
While playing correct, values for voice 0,1 are 0x00034001, 0x03577fff.
While beeping, values for voice 0,1 are 0x03577fff, 0x03577fff.

XDread

Hey! Don't touch the system now that is is working!