Windows 7 64-bit wave driver recording: Very strange noise

[elfan]

Hi everyone,

I've got a real mysterious problem that I hope that someone might be able to help me with.

The thing is, I've been using my Pulsar cards and Scope 4.5 since 2008/2009 and Scope 4.5 & 5.1 have been working just great for me in Windows XP ever since then (and I've even made sure that it runs well in Windows 7 32-bit). Playback and recording via ASIO works just fine without any cracks or pops and the wave driver likewise (sometimes I've had to increase the number of recording buffers or the buffer size in my DAW and audio recording program but that has always been related to performance so it's ok and can be explained that way).

However, I've been using Windows 7 64-bit since January and I thought everything was perfect. I was using playback and recording via ASIO mostly and the wave driver playback was just fine too. BUT! A few days ago I tried to record from the wave driver and that was just TERRIBLE (I just wanted to record the audio output from the system media player and also do a crude master dump by recording everything in the system). I'm not talking cracks and pops, I'm talking about constant trash data in the buffers. Periodic noise that is not related to any playback and not performance related at all. I have a fresh Windows 7 Ultimate 64-bit install and even when the machine is idling this noise is ever present. I've read through the Scope 5.1 Windows 7 64-bit tweak thread here on the forum and turned off everything performance intensive (fancy GUI stuff, services, stuff in the registry) - no change. I've turned off the things in the BIOS that I'm supposed to turn off (all according to that thread)- no change. It doesn't make any difference if I do those things or not, because I guess those target performance-related noise rather than what I'm seeing here.

The nature of the problem is that in Scope there is no trace of any noise. The Scope mixer does not pick up any noise from the wave source device but in the Windows 7 sound settings it's a constant level offset in the VU meter on the recording device in the sound recording settings. There is no problem when recording from ASIO in a DAW or anything, just when doing recording from the wave driver. The playback is just fine, both from wave and ASIO drivers - not a single pop or crack.

Here's the kicker: The problem GOES AWAY when I limit my system RAM to 2 sticks of 2GB each. These sticks of RAM can be distributed over any of my 4 slots and I can pick 2 out of any of my 4 RAM modules and put them in any slots and it will work but as soon as I've filled up 2 of the slots and pop in a third stick of RAM the problem appears. It's interesting because in Windows XP and Windows 7 32-bit this problem is not present and those OS:es only use less than 4 GB of RAM (does that mean that those OS:es disregard any installed RAM outside of the first 2 sticks of RAM that amount to 4GB?), but the problem appears in Windows 7 64-bit when more than 4 GB is used. So there is some kind of correlation there. I have run a full RAM check using the built-in mdsched tool in Windows 7 and there are no functional errors in the RAM that can be detected. And I've run the extended test too - took all day - no errors detected. The system works just fine besides from this and if there would've been some functional RAM glitches I expect I would've seen many more BSODs than I've seen lately. Sure, I see one from time to time but most of the time it can been attributed to application problems.
What could cause this? Does Windows 7 have some problem addressing more than 2 modules @ 2GB simultaneously under some settings? Does the DP35DP mobo have some setting that might affect the addressing of more than 2 modules? Is there some 64-bit, >4 GB mode than I should know about? Everything in the system is working fine... except for WHEN _RECORDING_ from the WAVE DRIVER... when using more than 4 GB of RAM? I don't even understand if this problem is related to Windows or to Scope. It's so strange! :/

My system is as follows:
OS: Windows 7 Ultimate 64-bit, fully upgraded to SP1 + all updates
CPU: Intel Core2Quad Q9450 @ 2.66 GHz
Mobo: Intel DP35DP (all the latest Mobo drivers installed)
Sound card: Creamware Pulsar 2 (I've got a Pulsar 1 that I run too, but the problem appears either way)
Graphics card: NVidia NVS-440
Memory: 4 x Corsair CM2X2048-6400C5DHX 1.80 V, rev. 4.1 (2048 MB, DDR2, 800 MHz, 5-5-5-18)

Note 1: The problem is not related to if I connect 1 or 4 screens or run single screen or multiscreen mode, also the problem is not related to if I run the NVidia driver or the driver that comes with Windows
Note 2: I was first running an ancient BIOS version from 2008 (02xx) and upgraded to the latest one (0517 from 2009 I think) - no change!
Note 3: The problem is not affected by whether or not you have installed the Intel dp35dp drivers.

I've attached 3 pictures.
One picture is of a recording of the noise. It's supposed to be completely silent and instead this appears. The noise is not consistent. Sometimes it's more silent, sometimes it's unevenly distributed between the stereo channels. Sometimes I actually manage to make a recording of what I want to record but then it's broken up by these spikes that you can see in the picture.

Another picture is of how the main mixer of the Scope 5.1 routing setup looks like. No noise except for some tiny DC input from the analog input. The noise is still there if I just have a wave source and a wave dest so it's not related to that. Besides, that DC signal is there in Windows XP and Windows 7 32-bit too and in that case there is no problem.
In this picture you can also see how the Windows sound recording settings look like and the VU meter for the recording input. It's constantly offset and correlates to the noise level that I record.

The final picture is of the Scope 5.1 routing environment just to show that I'm running the default settings. It's a fresh Scope 5.1 install.

Is there anything that I'm supposed to set up in regards to the memory? I've tried tweaking the RAM timing in the BIOS to like 667 MHz 4-5-5-13, 800 MHz 4-5-5-13 and 667 MHz 5-5-5-18, but no change.

I hope that someone can help me with this problem and point me in the right direction?

Thank you,
Tomas

[djmicron]

never used the wave dest before and didn't know about this.
I have the same problem on win 7 64 bits, i can record without noise with 16 bit wave dest, but the sound is fragmented.
i think it's a driver problem.

[garyb]

there's definitely a 64bit wave driver issue. afaik, playback is most affected, but....

[elfan]

Thanks for replying ! If someone could verify that downgrading to less than 4 GB of RAM causes the problem to go away that would be most interesting. Or if different RAM configurations (2/4 vs 3/4 sticks of RAM) affects the problem. Also, it's very strange that the problem is so different for different people. I've never had any playback issues of any kind with this driver and not that many people seem to have reported recording issues. I wonder why the sympoms are so fragmented and different. That is, if it's the same issue... there might be a number of different issues that just seem to be the same issue I guess.

[djmicron]

the symptoms are the same on my 64 bit system.

If you try to record with 24 bit wave dest, you'll get noisy audio.
If you try to record with 16 bit wave dest, you'll get fragmented audio.

Playback is fine on my side.

Reducing ram to 4 GB solves the problem, true.

Using only one stick of ram, it depends on your motherboard, mine is dual channel, so, no less than 2 sticks on the primary slots.

Since most of audio applications are asio based, this is not a big deal.

I have a a boot configuration used as mediacenter and the playback over wave source is fine.

If you have to work with wave destination because the asio driver cannot be shared, you can try something as rearoute, else i don't know why you need the wave destination.

[garyb]

wow, reduced ram fixes it!

well, the drivers are being rebuilt anyway. they should be released with v6. v5 users should also benefit, or this was the plan the last time i was informed....
the fact that reduced amounts or ram work, means that the problem can be in the bios, windows ram management, or the driver/programs themselves...it can also be a combination of all of those elements.

[elfan]

Aha, fascinating! Thanks for trying this out and getting back to me . So they have communicated some kind of plan to fix this in Scope 6? I guess I shouldn't speculate when the planned release date is for v6 considering how long I waited to upgrade to 5.1 ;D.

Should I mail a report of the whole RAM downgrade thing to Sonic Core customer support or does the Sonic Core customer support read these threads or how do you think it works ? Whatever helps Sonic Core release faster and produce more reliable software .

EDIT:
Another thing, just to make sure: Has anyone tried running more than 4 GB of RAM but with 2 sticks only? (like 2 sticks of 4 GB each?) My original configuration was 4 sticks of 2 GB each and because of that I couldn't conclude whether it's the number of sticks or the amount of installed RAM that changes things since I'm limited to 2GB per stick. I've been thinking of getting a pair of 4GB sticks to try it out but if someone has already tried it I won't bother. Thanks!

//Tomas

[babaorum]

well, the drivers are being rebuilt anyway. they should be released with v6. v5 users should also benefit, or this was the plan the last time i was informed....

good news !

[elfan]

UPDATE:

Not that anyone but me had any doubts about this but it's definitely the total amount of RAM and not the configuration (like number of RAM sticks or the timing) that is the culprit. I just changed from 4 x 2GB DDR2 800MHz (5-5-5-5-18) to 2 x 4 GB DDR 667 MHz (5-5-5-15) and in neither case it works. It doesn't matter if I use two slots in one RAM channel or one slot in each channel, the result is the same. And in both cases it works when I down grade to 2 x 2GB or 1 x 4GB respectively.

Sigh.

Oh well...

[dest4b]

Hi .. im in the same situation .. Wave Dest is not working in win 64bit.

is there any workaround or fix out now ?

[babaorum]

scope 6 ...

[petal]

scope 6 ...

It's the holy grail.

[jksuperstar]

It sounds like memory management in windows and the driver (or the driver isn't suited for addressing space above 4gb, even though it is a 64bit driver.). Which also explains why it's an intermittent issue across users...where the buffer is placed in memory, and how large a space is allocated for buffering. Still, I'm glad it's all known and being worked.

And I'm really glad to hear zo many things are being reworked and debugged for scope 6. It makes it even more worth the wait.

[dest4b]

Is there also a know bug with the MIDI device ?
Because i had massive problems with midi clock.
With a USB device it works without problems.

[SSJ]

Long live planetz !

I'm glad I stumbled across this thread. I have exactly the same problem, even down to the same noise pattern in cooledit on one channel.

I was very happy using sonic core 4.5 on XP but some of my VST soft synths were no longer supported on XP. So I finallly upgraded to sonic core 5.1 on Win7 64 bit and put in the extra 4gb stick of memory I had that I couldn't use on XP. I often make weird and wonderfull sounds in cubase with various synths and then sample them in cooledit going via sonic core and the wave drivers. But when I upgraded to win 7 all I got was noise (the same as in Elfan's screen shot). After reading this thread, I pulled out the extra stick of memory (I now have one 4gb stick) and it works fine !!!!!

So my questions are :-
1) is this a known issue with the wave drivers, or are there some more tests I can do ?
2) do I need to raise this with support ?

[garyb]

known issue mentioned in the letter that accompanies v5.1 keys....

[medway]

Reviving an old thread here but has this been fixed in v7 to work in Win 7 x64?

Cheers

[t_tangent]

Hi Medway,

Not yet, but I believe SC are aware of it and hopefully this will be one of the bug fixes in the next update, whenever that may be. Soon with any luck

[medway]

Hi Medway,

Not yet, but I believe SC are aware of it and hopefully this will be one of the bug fixes in the next update, whenever that may be. Soon with any luck

Thanks! I got around it for now by ticking the option to release ASIO in Wavelab when Cubase is selected. For some reason Cubase doesn't need to be set to release its driver for Wavelab to work (i have that option off).

Does v7 include multi client ASIO? Or do I just need to set them to release individually?

[t_tangent]

Scope ASIO driver is not multi client as far as I know, so just keep Release in background ticked in Wavelab and Cubase, or alternatively use something like VoicMeeter, some interesting info on this link, https://www.gearslutz.com/board/music-c ... ndows.html

Or Reaper has something called Rearoute or something like that, I can't remember what it's called exactly but should be easy enough to find out. Sorry I can't be of more help, but hope that gives some ideas.