Stereo Tool

No, it`s not GUI problem because .. MPXTool shows the same. It cannot be GUI issue when another (outside ST) program shows same.



Checked even some 7.41 betas - same
Checked 7.50 - same

Huh? I checked it in CoolEdit and there was nothing there! Weird...

Edit: Checked the code and all it can be - if it is indeed an error in Stereo Tool - is a rounding error in the sin() function, or in 32 bit floats. This rounding error is actually very likely - if you change the level of the pilot the disturbance gets bigger and smaller without really following the level of the 19 kHz tone. Which would only be explained by rounding issues. But.... they should be MUCH lower. Maybe at -120 dB or so.

These rounding issues might also occur in Intel's IPP library - both MpxTool and Stereo Tool use these. As I said, I don't see anything in CoolEdit. And I should see it there as well.

Pilot at 100% reading. There is highest peek there. but nothig to worry about.

_________________
control point
control point2

Wooh! As I said, in CoolEdit it looks clean. Bojcha, isn't that a Windows resampler error? It looks so bad...

Cool edit is not monitor as i know well.
Use MPX tool instead. And no, i think it may be a problem . Because this is entering in the RDS (nevermind that is small) . -120db sounds more likely .

i test it in just clean install of Windows , because i was thinking it`s a bad driver, or something else, but no.. it`s a clearly a harmonic. The strange thing is that appears without music

Best you can use is Spectra Plus SC . It will show you everything.

Here is the 57 Khz , but nevermind what level is having the 19 Khz , this peak is not moving ( ok may be a bit ) ! So it`s there permanently. At 3% pilot and at 18% , this peak is the almost same level.

I don`t know if it`s a problem or not , but looks like it can be cleared. For Libraries don`t know too. Then may be this will stay untill libraries are "optimized"

Check what other processors do... some push upto -60 dB of noise into the whole RDS area. That's FAR more disturbance than this signal at around -90. Which, still, I don't see in CoolEdit - I don't know why. It shows the noise level at -108 dB, and that's after dithering to 16 bit audio. With no spikes at all except the one at 19 kHz itself.

Audio piplines like VAC is not bit perfect, Hans I think forget about pilot harmonic.
16 bit resolution cannot give 100% pure clean sin wave with 10% of full scale level.

_________________
Urban Junglist

True, but dithering should fix that (spread it over the entire spectrum instead of causing aliasing bumps everywhere). Anyway, I just added an image of (16 bit wav) data generated by Stereo Tool, pilot only, silent input, to the above post and you can see that the only thing there is the pilot.