I've updated my post with a link to the test exe and source, slightly updated, and added a readme file of explanation as well.

Dont know how much more i can help as my coding skills arent even worth mentioning.

i did record the sound clipping starting from 30 each time i lower the write ahead value i change pitch so i could make it in one file

at .23 sound is just about gone.

sound test audio result - Vista64

Well it's too late to send a new file now anyway, I have attached the first version from Scawen including the source, and uploaded it as a cab file. They don't accept external download links for the files.
So I doubt they will go here and download the newer version... Maybe they will write an email if it's fixed
We'll see...

i doubt VERY much there won't be a way of doing live sound or that if a bug exists they happily release the OS anyway, because for example, there are loads of apps costing thousand of dollars that use pc to generate sound. In music biz etc.
i once tested lfs on a very early vista code and it worked flawlessly.
i dunno what exactly happened inbetween but i find this very weird and i suspect some lack of info issue.
good luck anyway.

Guys - I have access to the latest Vista builds and can file bugs directly without jumping through all the public beta hoops. I'm going to try this on the latest Vista build and will report back on whether or not it works. If it does not work I can talk to one of the sound developers and have him take a look.

Edit: I can confirm that this problem still exists in the Vista build from early this week. I'll file a bug against this and try to get the dev to take a look. Scawen - If there's any info you think needs to be put into the bug, please PM me.

Edit2: A bug has been filed and Scawen's "Sound Test" zip file has been attached to the bug. This should give them everything they need to look at the problem. And, trust me, the audio devs WILL see this bug. Whether or not it gets fixed is another matter entirely.

You can find a good readme in the archive that contains Scawen's tool and the source.

Quote from ORION :

You can find a good readme in the archive that contains Scawen's tool and the source.

Yes. I attached the archive and also put Scawen's comments into the descritption field. Hopefully I'm not stepping on your toes by filing a report, Orion. I just know from experience that customer reported bugs take forever before they actually filter down to the point that they are seen by the developers. I've got a little bit more of a "direct" route. I only wish I had seen this thread sooner.

Scawen - Can we break the audio portion of this thread off into it's own thread, separate from the unlocking issue?

appreciated :up:

Just wanted to give a short update on the sound issue. I filed a bug and it was looked at by the audio developers. In short - they are not going to fix this problem. They seemed to indicate that writing data into an already submitted buffer is a Bad Thing and should not be done.

I don't know what other options you have, Scawen. But, you may have to change the way sound works in LFS if you want it to work properly in Vista.

Sorry I couldn't be more help, but this is straight from the horse's mouth, so to speak. If I can be of any further assistance, please let me know.

Well I haven't got a clue how live generated sound can be possible if you aren't allowed to write ahead into a looping buffer. That sounds crazy to me. So what, we just have to play looping samples that have been pre-recorded, and that's all you can do on Windows Vista?

Thanks for looking into it and getting a reply, Cue-Ball.

if it's still allowed on vista, audio kernel streaming could do the trick, or maybe asio (it was menat for real time audio), but that would mean it would only work for users with expensive audio cards...

i think the real reason why ms won't fix this could rely on "trusted computing" features, maybe a sort like winx sp2 that won't let you see any dvd if you have a graphic card with tv out...

Quote from Honey :

i think the real reason why ms won't fix this could rely on "trusted computing" features, maybe a sort like winx sp2 that won't let you see any dvd if you have a graphic card with tv out...

I think you are probably on the right track here. I don't know any of the details, but a lot of stuff in Vista has been "sandboxed" so that malicious software and misbehaving programs can't destabilize the system. I've got a feeling that audio has been handled similarly and that's what is causing this issue. I don't know this for a fact, but it's my hypothesis.

Scawen - I'm not a programmer and even if I was, I'm bound by NDA. If you have specific questions I can try to get you some info. Otherwise, there's a limit to how much I can help, unfortunately.

Doesn't RBR generate sound in a similar way to LFS? If so, does it have the same problem? Anyone know?

somebody should try how rbr and nkp with really low sound latency settings work in vista

i can't say anything about rbr or nkp, but GTL seemed to work without the lag issue. now, i have no idea how their audio system differs. just an observation.

that said, i've since ditched vista, and most likely won't be going back for years to come. i don't see any need or reason to do so. and UAC really really annoyed me. Even with UAC completely disabled, and the control panel "run as admin" the system wouldn't grant me permissions to calibrate my DFP. i blame that issue on beta hiccups, but still, i find UAC to be more of a hiderance than help.

i think gtl has prerecorded sounds, so it is not affected by this vista bug (yes to me IT'S a BUG!)

i will not start a discussion about "trusted computing" or will not stop (), but it's clear that starting from vista, users and developers will be no loger the owners of their own systems...

since vista support is taking so much time from devs, my sugestion is posticipate it after S2 final (if it was for me i will never support vista :nod

I completely agree that it is a bug and i feel that MS is going to far out of their way to protect users from theirselves (themselves?).

how is denying access to something thats not to be accessed in the first place a bug ?

It is a bug, there is no doubt about that.

What I'm doing is very simple, you can read the source code of the test program. Just asking windows where it has got to in the looping sound buffer, and writing new sounds ahead of that point, before the system gets to that point.

There's nothing wrong with it, the documentation doesn't say it's a bad idea - they provide a function "waveOutGetPosition" for this exact reason. It makes sense and it works with all previous versions of Windows. So it is a bug and a very bad one. I don't accept some crazy Microsoft programmers telling me what I'm doing is not a good thing. It would be a "good thing" for them to make their own functions work correctly. And if they want to say that this isn't a "good thing" then they should explain how it should be done.

Yes I am irritated. :mad:

not to mention that before asio came out, that was the most common technique for application dedicated to musicians for real time audio effects, who has ever developed will see it as it is: a damn bug!
who is not a developer, may be mislead from microsoft advertisement...that's just sad that road that the TCG is trying to impose...

Hi all...especially Scawen,

What all I am reading here is very annyoing.I mean how Microsoft currently making position with this issue.I still hope this will be fixed by MS but have some doubts about it.

I dont want to start some discussion about how far Microsoft is going with the users/programmers loosing control over the OS and how MS can control things .This is really not place for that.

Scawen,
sorry to say that but it might happend that in this case you are too small "boss" to get the M$ asses moving so you will probably have to do with this something in future(I still hope you will have not to really bother with this issue).We all know that you more then a good programmer and that some guys in MS should wake up and not bother with their "truth".Wish you the luck in this case.Fingers crossed

P.S:we äll know that MS sometimes call "bug" a new featureillepall

Maybe some of the ppl at MS have heard the name "Peter Molyneux" though

Somehow I have the feeling they dont even know "Bill Gates" though...

I can guarantee you that this is not going to be changed for Vista. It's too late in the product cycle for such a change, and the function that allowed this system to work (allowed for inserting data in an already streaming buffer) is deemed as risky. Simple as that.

Scawen will have to find some other way to generate the sound. Luckily, he's one of the most talented programmers I've come across, so I'm sure he can do it. With a little luck not only will he work around this new limitation, but he'll improve the end product as well. Who knows, maybe this will be the catalyst that gets an improved sound generation system put in the game?

Note: I'm not a representative of Microsoft and make no statements on their behalf. You'll just have to trust that I know what I'm talking about.

maybe i didn't remarked enough in my first intervetion...solution exists: Kernel Streaming, but that means scawen will have (if he wants to support vista) to reimplement any part of the lfs sound engine...let's say a third of lfs should be reimplemented...
for this a many other reasons i thing that the gaming industry, will be the last to support vista...if that happens, ms maybe will be forced to step back with a service pack for "backward" compatibility.

Quote from Honey :

for this a many other reasons i thing that the gaming industry, will be the last to support vista...if that happens, ms maybe will be forced to step back with a service pack for "backward" compatibility.

I have some feeling that Vista might end like Windows Millenium.The biggest mistake in MS history.

Quote from Honey :

.. it would only work for users with expensive audio cards

Not really, ASIO has had DirectX drivers for over 7 years. With reasonably simple content such as LFS' engine sounds a budget card/chip should easily be able to run with latency well under 20ms.

There's also http://www.asio4all.com/ (WDM).