Stereo Tool

Hello back to all.
Here I am with question about ST bug maybe.
Here is video to watch and Hans to check what happens when no signal and Stereo widening is showing activity without input signal !

http://www19.zippyshare.com/v/60723341/file.html

I think the graph is working on the noise from the soundcard that it receives.

Oh no...
don`t think that way , because virtual cables don`t have that noise you think about
It`s not an input of sound card - they are all MUTED
Input is virtual cable from Winamp, and Winamp stopped, closed.
Never mind if it`s opened - it`s the same.

Ok then it's probably some rounding errors. But i'll leave answer over to Hans.

@Hans About the tilt display can you change color of graph to red when it hits the upper and lower edges? Because that makes it easier to see if it reaches the limits of the soundcard.

Yep.. good suggestion for the color !

I add a (very) low amount of noise (below -200 dB) to avoid problems in calculations with 0's - Intel CPU's don't like the number 0 and absolute silence causes the CPU load to more than double. So that's why you see something here even if there's absolute silence playing.

It changes when you disable/enable reverb protection because if that's disabled it uses 10 bands (it lifts on the existing bad splitting from Multiband), if it's disabled it uses 5.


I'll add a bar that displays the maximum output level to the sound card so you can always see if there are (or have been) problems.

WOW...!
Is Intel really do that with the 0`s...
O god...can`t realize why.
Turn to AMD then....I always think that they are something little more.Even i have the E version of CPU ( and previous CPU was E version too ) . Suggest You to try with AMD libraries to build one or maybe two betas/versions .
Then we all will see the difference may be.
See eldorado ..... problem is Intel .. don`t expect that this may be the problem... really. Confused...


But Hans ... ok i understand .
Is this mean that you have to build an internal noise generator to have the intel satisfied about that.
I think it`s stupid from intel`s side to pass this problem trough and continue to build CPU`s without solve this.
Ok maybe solved in the latest ones...but...

And in the VST version too. Thanks!

Not sure if AMD doesn't have this issue. Actually it probably does - by calculating with 0's you risk getting 'denormals' in your output and they cause problems. And there's an interrupt for that in the CPU - as you can imaging things don't get faster from this.

Yep , may be ....but this isn`t cost anything to try compiling with AMD.. if You have any.
I`m not saying anything but only to try if you like to.
Simply i can`t believe that the problem is in the processor . But you say so. Machines are not perfect yet... thats what i understand from Hans today

Between .. yeah this new beta looks/sounds very good .
I found a bug in the buffer display in previous beta but it looks fixed.