Computers & Music:

Odd properties of the volume envelope

by Kai Pihl

Now when we have luckily made our first soundbank with one instrument which has one sample, we can start our experiments. We must be sure that the looping is switched off because it disturbs the results at this stage. It can be switched off from the Loop Marking dialog unchecking the 'Enable looping for this sample' checkbox.

However, we shall forget the loops for a little while, and check first out the other characteristics of the volume envelope. Although we disabled the looping from the Loop Marking dialog we must also take care that the possibility to get a sustained sound is also taken away from the envelope settings. It's done by adjusting the sustain phase level of the volume envelope to zero. Select the sustain phase from the select window in the upper part of Volume Envelope dialog.

Then throw the slider all the way to the left. You can see how the sustain level decreses to the bottom of the graph window. Notice also that the length of the sustain phase remains the same, only the level decreases. This is because of the nature of the sustain phase. It's not a time value att all. The sound stays at this level all the time a key is pressed provided that a loop has been determined and it is enabled. If we now check the sound by pushing the midd C key on the virtual keyboard, we can't hear anything. This is because of that all the phases are at their minimums. No attack, no hold and no decay. The release phase comes after the sustain and because there is no sustain there can't be any release of it. We only hear a 'pop' when the sound tries to get start.

Let's start to check out how the attack, hold and decay phases work. Select the hold phase from the Volume Envelope dialog and adjust the length exactly to 1.000 second. Test the result by pressing the middle C key on the virtual keyboard. You hear the first part of our test sample where the pitch is in middle C. To be honest, you also hear a short part of the next pitch - one octave higher. We are going to learn more about this peculiar phenomenon in the next paragaraphs. How short this tiny "bep" migth be? We can check it out by adjusting the hold time shorter in small steps and test the sound between those steps. This way I got it disappear when the length of the hold time was 0.954 seconds.

Now, adjust the length of the hold phase exactly to 2.000 seconds. Test the sound. You hear the second part - one octave higher - to it's end. But, again, we can be hear an unexpected effect, this time a short part of the third pitch. My inspection revealed that this part disappears if the hold phase is adjusted to 1.909. So, is there something wrong? I recorded the output of EMU8000 with WaveStudio the way I explained before. In the WaveStudio everything seems to be allrigth! The lengths of different pitches are correct. These results confuse me. Before I have an answer from Creative to this question, we are forced to accept these pecularities and regard them as software bugs in Vienna 2 and only in it's user interface - not in the sample it self. Needless to say, but this type bugs makes the enchangements in Vienna 2 unusefull. The precise readings of different parameter values are not usable because they are not precise!

Next: Sustained sounds Prev: Making an own test sample To: The contents