About the Levels/ Which dB?

[Darwin]

Hi,


In the manual reference, it is said that if the "LED" is green, the signal is between -40 and 0 dB, Yellow > between 0 and +11 dB And red when > 11dB.

Which dB are they talking about?

[Darwin]

Hi, I want to provide some further information about my yesterday comments :

As you may know, an audio signal can have some clippings when the wave is reaching 0dBFS ( zero dB full scale).
I wanted to know when my patch would start to have internal clippings / distorsion just by watching the "audio" levels which can be monitored in the mixers or the Input/output levels.

In the manual reference, it is said that if the "LED" is green, the signal is between -40 and 0 dB, Yellow > between 0 and +11 dB And red when > 11dB.

Which dB are they talking about?

So, I tryed to see by myself .
I built a simple patch : Sinwave and LevAmp and Output module.
Up to levAmp = x1 > Green
From >1 to 3,50>yellow
Above 3,50 > Red

Then, I registered each wavefom in Wavelab.
I thought I would find some internal clippings when Yellow or red.
Nothing > perfect sine waves in the 3 cases!
The sine changed when I took another LevAmp > (started to change above x4.00)

Same remarks with Sawtooth wave.

So, the dB which have been taken into account are not dBFS.

What dB are they talking about? Anybody worried about this question?

[Fozzie]

I might be wrong here, but this is how I think the deal is:

Internal signals in the G2 are between -64 and +64 'clavia units'. I believe +64 clavia units relates therefore in G2 terms to 0 db. However, the G2 has headroom up to 256 clavia units before clipping takes place. This correlates well with your experimental results, as this would be a signal of 4x '0 db' in clavia units.

[Darwin]

Fozzie,

Yes, you are wright : it is headroom. 24 bits help

I think calculation "4x0dB" is not correct...( it is not equal to 0...otherwise 5x0dB would have the same effect).

In fact with the sine experience, it leads me to think that we can touch the red level without having any internal clipping or distorsion. Strange.

[blue hell]

Fozzie is right, and with 4 x 0 dB he means 4 times the level equivalent to 0 dB, which corresponds to +12 dB.

The level meters turning red before clipping is somewhat arbitrary. When you have 1 voice of from a patch you indeed have 12 dB left, but when you have 4 voices of that same patch you'll reach the clipping level. When there are more than 4 voices you'll go over it. So it must be seen as an indication for what will happen and it is not exact as it depends on the voice count.

Probably this could have been done differently.

[Darwin]

In fact, I all the patches I made, I tried to stay in the green area and I raised my levels with a preamp.

But that's true there are so many patches which are greater when some module outputs reach the yellow zone.

I am going to use my ears more than the peak levels to build now

[Antimon]

One way to check the combined output of voices in a polyphonic patch is to send the output through the FX area and check the output there.

It would be nice to have some kind of debugging modules, for instance a level meter with higher resolution (say, going horizontally across an entire module width), like the one I see in the GUI for my E-MU soundcard (1820 I think it's called), but bigger.

Another thing that I've daydreamed about is a kind of waveform snapshot/sample image that you could use as a substitute for an oscilloscope. I don't think this would be impossible (the oscilloscope doesn't need to have a great resolution, and the image data need only be sent to the editor on a low framerate).

I'm sure that Clavia uses some kind of debugging modules when developing the Nord Modular software.

/Stefan