Are speakers and headphones not protected against DC? When building in an highpass filter to eliminate possible DC, don't you risk loosing the bottom end of the sound?Safety warning: Do not let this signal go to your speakers/headphones. DC voltage will damage them. Always use a High Pass before you leave the digital world. Mulab is a clean application with no such HP installed by default
Amplitude Modulation
-
- KVRAF
- 13865 posts since 24 Jun, 2008 from Europe
Nice trick Andreas. Still i'm wondering when you need audio offset.
-
- KVRAF
- 13865 posts since 24 Jun, 2008 from Europe
-
- KVRian
- 855 posts since 3 Mar, 2009
Jo, finally back at my system to give some proper feedback.mutools wrote:Nice trick Andreas. Still i'm wondering when you need audio offset.
Are speakers and headphones not protected against DC? When building in an highpass filter to eliminate possible DC, don't you risk loosing the bottom end of the sound?Safety warning: Do not let this signal go to your speakers/headphones. DC voltage will damage them. Always use a High Pass before you leave the digital world. Mulab is a clean application with no such HP installed by default
About the DC power send to speakers (all idealised speaking):
Say you run a 100W speaker with a 100W amp and the DC part would be set so that 50W of DC power is going into the speaker. That means the membrane would get constantly 50W, will heat up and maybe not destroyed but it's potentially not healthy.
Further on the speaker will have it's zero resting point at something like 50% forward. The speaker will have now 25% of it's natural excursion to the front and 75% to the back. So the positive halve wave will be distorted (compressed) while the negative is still clean.
As far as I know this will result in 2nd order harmonic distortion. In M6 this can be easier achieved with the pos/neg splitter module, but in M5 this was the only way.
About the PWM of a square wave made with DC voltage, see the attached MUX patch.
The New <-> Standard knob let you listen to my "the old analog way" and the standard clean factory module way. They do sound different.
You do not have the required permissions to view the files attached to this post.
-
- KVRAF
- 7412 posts since 8 Feb, 2003 from London, UK
Just thought I'd take a look for educational purposes as I've been reading the thread with interest. The two source saw wave oscillators have some differences -- could you explain the choices?
(Also, must keep an eye out for this: I dropped the .Mux into my user presets but it didn't get picked up; I cleared the cache and restarted MuLab, still didn't get picked up; I had to drop it onto a rackslot to load it.)
(Also, must keep an eye out for this: I dropped the .Mux into my user presets but it didn't get picked up; I cleared the cache and restarted MuLab, still didn't get picked up; I had to drop it onto a rackslot to load it.)
-
- KVRian
- 855 posts since 3 Mar, 2009
With the standard factory osc I use the AIPS to make a square wave out of a saw wave. That is a pure math function. Phase shift a saw wave e.g. by 180 degrees and add original and shifted, you get a square with 50% duty cycle.
You can use a square wave in the first place, but you can modulate the AIPS parameter and get PWM.
The other way to make a square, the one where i use the DC signal, takes a saw wave and boost and saturate (soft limit) it big time. That results in all positive saw values become positive max value and all negative saw values become negative max value. This again is a square wave. If you add a (positive) DC signal, you shift the saw up and more parts of the saw are above the zero line, which means the positive part is longer and the negative part is shorter = pulse wave. Modulate the DC amount and you have PWM.
If i remember correctly i once read that this was used in a HW synth. The different sound comes form, i guess, the saturation and the DC remover HP plus an extra HP that emulates the various other HPs (capacitor coupling) in an analog circuit.
About recognising a MUX, i usually drop it into a rack slot or i (manually) delete the MuxPresetFileCache.MuData file is the relevant folder. Usually the Refresh List in the browser, Mux Preset, Options menu does this as well.
You can use a square wave in the first place, but you can modulate the AIPS parameter and get PWM.
The other way to make a square, the one where i use the DC signal, takes a saw wave and boost and saturate (soft limit) it big time. That results in all positive saw values become positive max value and all negative saw values become negative max value. This again is a square wave. If you add a (positive) DC signal, you shift the saw up and more parts of the saw are above the zero line, which means the positive part is longer and the negative part is shorter = pulse wave. Modulate the DC amount and you have PWM.
If i remember correctly i once read that this was used in a HW synth. The different sound comes form, i guess, the saturation and the DC remover HP plus an extra HP that emulates the various other HPs (capacitor coupling) in an analog circuit.
About recognising a MUX, i usually drop it into a rack slot or i (manually) delete the MuxPresetFileCache.MuData file is the relevant folder. Usually the Refresh List in the browser, Mux Preset, Options menu does this as well.
-
- KVRAF
- 7412 posts since 8 Feb, 2003 from London, UK
OK, both have AIPS 50%. The "new" also adds 6.02dB (against -8.36dB for old) and some different tuning values. The new then gets "hammered" further into shape. I was wondering if the tuning differences were relevant, though, or you'd just been tweaking one and not the other at some point.
-
- KVRian
- 855 posts since 3 Mar, 2009
Both have AIPS but only the Standard has it activated.
The -8dB is to have equal level at both paths.
The other saw wave is at max level to make boosting easier - which explains why i had to boost the DC by an extra 6dB, I was wondering.
The level of this part is set by the Out Level of the "Make Square" module, which is a tanh and acts as a saturator/limiter. The tanh usually starts latest at 50% drive to alias dramatically. So, the drive is high but not too high. Play with the drive and you see what i mean.
I made this thing with M5, where the osc was able to create DC signals. At one point this was gone, and it's kind of back again. So i replaced the osc module from M5 patch with the DC source for M6.3. See the link for the Vintage Square original file. Those patches are clicking in M6 and play fine in M5.
https://www.dropbox.com/sh/lztrpkl7q129 ... reasD?dl=0
The -8dB is to have equal level at both paths.
The other saw wave is at max level to make boosting easier - which explains why i had to boost the DC by an extra 6dB, I was wondering.
The level of this part is set by the Out Level of the "Make Square" module, which is a tanh and acts as a saturator/limiter. The tanh usually starts latest at 50% drive to alias dramatically. So, the drive is high but not too high. Play with the drive and you see what i mean.
I made this thing with M5, where the osc was able to create DC signals. At one point this was gone, and it's kind of back again. So i replaced the osc module from M5 patch with the DC source for M6.3. See the link for the Vintage Square original file. Those patches are clicking in M6 and play fine in M5.
https://www.dropbox.com/sh/lztrpkl7q129 ... reasD?dl=0
-
- KVRAF
- 13865 posts since 24 Jun, 2008 from Europe
Thanks for the info, yes i knew that.AndreasD wrote:About the DC power send to speakers (all idealised speaking):
Say you run a 100W speaker with a 100W amp and the DC part would be set so that 50W of DC power is going into the speaker. That means the membrane would get constantly 50W, will heat up and maybe not destroyed but it's potentially not healthy.
Interesting aspect, didn't think of that. It depends on the speaker though, right? Maybe there are (future) speakers with kind of headroom and/or linear curves?Further on the speaker will have it's zero resting point at something like 50% forward. The speaker will have now 25% of it's natural excursion to the front and 75% to the back. So the positive halve wave will be distorted (compressed) while the negative is still clean.
I would not take that into account when creating patches cause it's something that depends on the speaker system, and you don't know the physical rendering device's (eg speaker) characteristics when creating a MUX patch.As far as I know this will result in 2nd order harmonic distortion.
More generally: I would really try to avoid creating patches that generate DC. And thus also avoid the need to filter DC (and thereby eventually also touching the real audio in an unwanted way)
If you want PWM please use the Multi-Form Oscillator. That avoids DC.