DirectSound Test [VISTA]

[Scawen]

[ EDIT : XP users please no more reports unless you find a problem - we need Vista reports now ]

OK... some of you may remember the sound test program which we used to identify the Windows Vista problem. I've changed that test program to use DirectSound instead of waveOut. It's very small, only 52 K.

Instructions are on screen but anyway, run the program and press S to start. You will then find, if you use A and Z to change the frequency, there is quite a lag in the changing of the sound. You use N and M to reduce or increase this lag (known as "Write Ahead"). Try to go as low as possible. At some point, you will go too low, and you will get sound corruption either continually, or when changing the frequency.

What we need to find out is :

- can this eliminate sound lag in Vista?
- how low can you go with "Write ahead"
- any problems with it in XP or 98?

http://www.liveforspeed.co.uk/SoundTest_DSOUND.zip

[three_jump]

win xp machine here:
- write ahead working till 0,02sec, 0,01 causes constant "scratches"

[nesrulz]

win xp machine here:
- write ahead working till 0,02sec, 0,01 causes constant "scratches"

Same here.

[Vain]

Xp Sp2 with an intel onboard soundchip: Work ahead works perfectly with 0.02 while 0.01 causes major distortions. Does so with all audible frequencies.

[Edit] Update
On my win2k SP4 the write ahead even works with 0.01ms. This is on a Creative Coundblaster 128 PCI.

Vain

[Scawen]

Thanks for the reports. Same here in Windows XP. 0.02 is a fine result, so all OK so far.

Let's stop the thread getting long with XP reports now, avoid them please unless an XP user finds a problem with it.

The main concern now is... Vista! :)

[Highsider9]

Win Xp Home SP2

Works fine, even with 0.01 sec delay.

[askoff]

Same here.
And here.

[BurnOut69]

Did you guys read what Scawen just posted? Vista is the problem, not XP :)

[Scawen]

Yes, the XP reports up to now are appreciated. But no more please unless you have a problem in XP! :)

[askoff]

Did you guys read what Scawen just posted? Vista is the problem, not XP :)
Yes. There where only three post's when I opened this thread and started to reply...

[Viper29]

XP SP2, Creative Audigy 2 ZS. No problems at all, even at 0.01.

Edit: Sorry, just saw Scawen's post about not submitting more XP related info...

[Scawen]

I've made that quite clear now in the first post. :D

[bbman]

I don't know if you're calling this a problem, but here it is anyway:

I can go as low as 0.02 sec without any problem, but with 0.01 sec I get skipping/distortions in short intervals, changing every time I start the program (but mostly 3 sec. interval)...

I'm on Windows XP Prof. SP2 with USB-Headset...

Don't know if that helps... :shrug:

[Scawen]

0.02 is fine, a very acceptable (and imperceptible) lag.

[t1ger]

OK just tested on Vista (Ultimate Edition) Beta 2 (5308). This is not the latest version, I know - but I havent installed the latest one yet on a "gaming" PC, I will bring the DVD home with me tomorrow and test it this next week.

Test results:

0.01 - get a click every second, and every 3-4 seconds get two or more in quick succession clicks.
0.02 - get a click every second, sometimes misses a click
0.03 - usually OK, but every now and then, if I switch to another window and then back again, the click every second starts. If I move up to 0.04 and then back down again to 0.03 is settles again and works fine. This is very intermittent and doesn't always click.
0.04 - all OK even when switching windows.

During this test, the PC was not doing anything other than the test and running IE to type this in.

PC spec:
Intel P4 3.0GHz
1GB RAM
SoundBlaster PCI Soundcard (can't remember exact model - can find it if neccessary).

EDIT: when on 0.03 hovering the cursor over the File menu causes it to start clicking, on anything higher (ie 0.04 or above) this does not happen.

EDIT2: Don't know if it is relevant, but when the SoundTest window is not in focus the sound stops, and when the Help >> About window is open, the sound misses for about half a second, every second.

EDIT3: My Vista RC1 DVD is bust, so I have been delayed in testing on this version.

Thanks

Tim

[ORION]

Vista RC1 build 5600 works perfect at 0.02s :D :up:
[=no delay aswell]

(nForce2 onboard sound)

[wjw]

Ok

[DanneDA]

WinXP SP2, Realtek HD Audio, Okay till 0.02s, not okay at 0.01s.

Will update with message with Win Vista Build 5744 (RC2) in a few minutes.

Vista Build 5744 (RC 2), Realtek HD Audio, Okay till 0.03s, not okay at 0.02-0.01s.

Scratches in both cases.

[St4Lk3R]

wow... reading all these reports makes me sorta happy :)

the last time this soundcheck test program was run, We were talking about a sound lag of 0.3 up to .7 seconds. now (on my XP machine) it works extremely well even with soundlag of 0.02! great work you did there Scawen!

[Scawen]

Good results! :)

Even though Vista is slightly lagged compared with XP, 0.05 really isn't a problem at all. Using waveOut, you needed 0.25 to 0.30 to eliminate constant glitching with Vista.

Vista people, please verify that the delay is "true". What I mean is, when you set to that safe value of 0.05, and you then change the frequency with A and Z, is the delay really that small? 0.05 should be unnoticeable.

You'll know if there's a problem, i.e. a "hidden lag" because the changing of the frequency would be noticeably delayed after pressing the A or Z key (sort of like it is when you set to 0.30).

What I'm trying to establish, is 0.05 in Vista the same as it is in XP (i.e. too small to perceive).

[t1ger]

Good results! :)

Vista people, please verify that the delay is "true". What I mean is, when you set to that safe value of 0.05, and you then change the frequency with A and Z, is the delay really that small? 0.05 should be unnoticeable.

You'll know if there's a problem, i.e. a "hidden lag" because the changing of the frequency would be noticeably delayed after pressing the A or Z key (sort of like it is when you set to 0.30).

What I'm trying to establish, is 0.05 in Vista the same as it is in XP (i.e. too small to perceive).

No noticeable delay here at 0.05. :)

Thanks
Tim

[DanneDA]

Good results! :)

Even though Vista is slightly lagged compared with XP, 0.05 really isn't a problem at all. Using waveOut, you needed 0.25 to 0.30 to eliminate constant glitching with Vista.

Vista people, please verify that the delay is "true". What I mean is, when you set to that safe value of 0.05, and you then change the frequency with A and Z, is the delay really that small? 0.05 should be unnoticeable.

You'll know if there's a problem, i.e. a "hidden lag" because the changing of the frequency would be noticeably delayed after pressing the A or Z key (sort of like it is when you set to 0.30).

What I'm trying to establish, is 0.05 in Vista the same as it is in XP (i.e. too small to perceive).

Did not hear any delay in Vista.

[Zero7]

Vista32 RC1 Build 5600 on hp NC8000 laptop / AC97 Audio

Setting the write ahead to 0.05 and changing frequency doesn't produce any noticable lag on my installation (even rapid frequency change).

Tried at 0.02 and rapidly changing the frequency (keep finger pressed on A/Z) does result in some clicking, but it's very infrequent.

[wien]

Vista RC2 32bit, Realtek AC'97 (Vista built-in driver). The occational blip when changing frequency at 0.03s, but everything's good at 0.04s. Can't say I can notice any lag when changing frequency.

EDIT: Same computer with a Terratec DMX6Fire (driver v5.40.3.130) seems to work fine at 0.03s. :)

[Dredd]

What we need to find out is :

- can this eliminate sound lag in Vista? yes
- how low can you go with "Write ahead" 0.03 <- still happens once or twice when changing from up to down (a & z) with any write ahead
- any problems with it in XP or 98? vista 32bit

[hardcoreobscure]

RC1 64 bit
AC97

ok right down to 0.02, and its not noticable at 0.05 (to me anyways) :thumb:

*edit

if I hammer the A & Z buttons at 0.02 I get occasional scratching, however at 0.03 I didnt get any scratches at all

[AndroidXP]

Apparently every system so far seems to work at 0.05, I'd say hardcode it to that value. This would be even faster than the current minimum sound lag we have (0.1s), so there should be no complaints. I'm glad to see that you have finally found a solution :)

Just for reference, I'm on XP with an Audigy 2 ZS and have no problems even at 0.01s.

[KiDCoDEa]

Vista rc2 5744 , nothing running except OS stuff + IE.

0.03 is lowest flawless i could get. 0.02 has a skip every 5 secs or so. 0.01 is skipfest.

cmi 8738 c3dx onboard.

amd@1.71 on asusV333deluxe.

[daveb948576]

ok im gonna try it with vista 32bit build 5600 and with a 3700+ cpu onboard sound and a 16mb graphics card. im unsure if the graphics card makes and difference but bear in mind my graphics is all being done by my cpu. "ill be back"

0.03 is good for me.

[Ge-Off]

Vista RC2 build 5744 x86
Default nForce2 drivers (6.14.462.0 = 4.62)

0.02 sec = very few scratches
0.03 sec = perfect

[trackah123]

Vista RC2
latest x-fi vista drivers

0.01 i get pops.
works fine from 0.02 and higher.

also all frequencies seems to work fine

[hotmail]

win xp 0.02

but the program doent work .

he say:

Sound Test: DIRECTSOUND

start/stop(s): Stopped

i allso have sound problems ,
scawen i hope u can do some think with this information

[K--]

Vista RC1 : 5600

0.02 is fine, 0.01 is mostly fine but not 100% perfect all the time.

Using Realtek 882 HDA chip on ICH7R of Intel 975x chipset (asus p5w-dh)

[Scawen]

win xp 0.02

but the program doent work .

he say:

Sound Test: DIRECTSOUND

start/stop(s): StoppedThat's what it says when you first run the test program.

You must press S to start the sound. Does it work then?

[Scawen]

TEST Patch U23. DirectSound now implemented into LFS!

http://www.lfsforum.net/showthread.php?p=243881#post243881

Thanks for your help! I'll close this thread now.