Hello, thanks for the video ! I have a question : why do you combine the two amplified channels into [abs] object ? They will add one to each other no ? So the amplitude will be closed to twice the single value and will disturb the threshiold detection no ? Why not connecting a [max] object to the amplified input level and make a threshold detection on the largest input selected by [max] ?
Hi, thank you :) that's correct and worth to point out! I'm treating this signal as if it was mono, although the signal itself is stereo and being sent to a stereo output, the compression is acting as a single channel compressor. To make a true stereo compressor, you should definitely measure the loudness of each individual channel and send the gain correction to each individual output. In this video I show the most basic gain computing method, you can however extract better loudness representations from after the input gain (for example, channels (1+2)/2, or the root mean square of a given portion of the signal. Bottom line is that in reality the gain computing stage can be stretched to a far more complicated (and creative) thing, to be covered in this particular video. While the concept of compressors is fairly the same for any compressor, the way they operate can be very different (it's a big part of what giving them unique colors)
@@Tune4media I noticed that also and was glad someone mentioned the merging of signed values before taking their absolute value - it means both channels could be peaking (+1, -1) but they'd resolve to 0 if you sum them and then abs them, whereas if you individually abs'd them, you'd have 2 (+1 and +1). fun fact - jongly.aif is mono
how do you get gen~ to be stereo? for some reason when I try to type it in it gives me two inputs but only one output. How do I configure it so I can have two outputs?
It should be simple as adding 2 'out' object for each channel and adding a number for each ('out 1' and 'out 2') Maybe, make sure that the number of the channel is written with space before the word 'out'.
hmm I copy the workflow you have here, but when I go to set the input label under @ in the menu, there's nothing there. Any idea what else needs to be done? This is a screenshot of within the gen~ object and then back out when trying to link the parameter. imgur.com/a/WmuHdTq
you find making a GUI hard ? its fun and easy esp in this tut if you follow it..its very easy.. add dials to all the number boxes so put them on top and route them into the number box outside of Gen like he has- and change the parameters to the dials accordingly- all you do is choose what you want to be shown in presentation mode and then go to presentation mode and arrange the dial how you want them to look in the device window in ableton
I can't find a reason why not :) gen~ process is by single samples (unlike MSP) so I'd actually prefer gen~ over MSP to create granular process (at least at it core sound), Check out also my video about Karplus-Strong synthesis. The power of gen~ makes the synth sound better and more capable.
No, MaxMSP has a dedicated domain/object to do FFT process. Look for the 'pfft~' object (I think the official TH-cam channel of Cycling 74 has a few videos about FFT in Max) You can however incorporate gen~ processes inside the pfft~ object
![Max | Designing your own Reverb (with gen~) [Full Lesson]](http://i.ytimg.com/vi/F_FhvxT5mLA/mqdefault.jpg)
![Max | Designing your own Reverb (with gen~) [Full Lesson]](/img/tr.png)
you, sir (tune4media), have saved the internet (my grades) today (4 years ago).
saved our groupproject. thank you
Hiii
Great Video!
thank you very much
Hello, thanks for the video ! I have a question : why do you combine the two amplified channels into [abs] object ? They will add one to each other no ? So the amplitude will be closed to twice the single value and will disturb the threshiold detection no ? Why not connecting a [max] object to the amplified input level and make a threshold detection on the largest input selected by [max] ?
Hi, thank you :) that's correct and worth to point out! I'm treating this signal as if it was mono, although the signal itself is stereo and being sent to a stereo output, the compression is acting as a single channel compressor. To make a true stereo compressor, you should definitely measure the loudness of each individual channel and send the gain correction to each individual output. In this video I show the most basic gain computing method, you can however extract better loudness representations from after the input gain (for example, channels (1+2)/2, or the root mean square of a given portion of the signal. Bottom line is that in reality the gain computing stage can be stretched to a far more complicated (and creative) thing, to be covered in this particular video. While the concept of compressors is fairly the same for any compressor, the way they operate can be very different (it's a big part of what giving them unique colors)
@@Tune4media I noticed that also and was glad someone mentioned the merging of signed values before taking their absolute value - it means both channels could be peaking (+1, -1) but they'd resolve to 0 if you sum them and then abs them, whereas if you individually abs'd them, you'd have 2 (+1 and +1). fun fact - jongly.aif is mono
Well, you should travel to Bulgaria more often :)
whithout the envelop, isn't it a distortion patch ?
thanks
how do you get gen~ to be stereo? for some reason when I try to type it in it gives me two inputs but only one output. How do I configure it so I can have two outputs?
It should be simple as adding 2 'out' object for each channel and adding a number for each ('out 1' and 'out 2') Maybe, make sure that the number of the channel is written with space before the word 'out'.
hmm I copy the workflow you have here, but when I go to set the input label under @ in the menu, there's nothing there. Any idea what else needs to be done?
This is a screenshot of within the gen~ object and then back out when trying to link the parameter.
imgur.com/a/WmuHdTq
@KairiRiku nope...
what about the GUI? no one shows this!
you find making a GUI hard ? its fun and easy esp in this tut if you follow it..its very easy.. add dials to all the number boxes so put them on top and route them into the number box outside of Gen like he has- and change the parameters to the dials accordingly- all you do is choose what you want to be shown in presentation mode and then go to presentation mode and arrange the dial how you want them to look in the device window in ableton
granular in gen? what do you think? :D
I can't find a reason why not :)
gen~ process is by single samples (unlike MSP) so I'd actually prefer gen~ over MSP to create granular process (at least at it core sound), Check out also my video about Karplus-Strong synthesis. The power of gen~ makes the synth sound better and more capable.
what microphone are you using for the vo?
Recorded with AKG Perception 220
Where can I have infos about this X[n] stuff ?
If you Google: Julius Orion Smith you'll find his remarkable papers over CCRMA. Great information about DSP (required some math)
Hello, its possible to create FFT in gen~ ?
No, MaxMSP has a dedicated domain/object to do FFT process. Look for the 'pfft~' object (I think the official TH-cam channel of Cycling 74 has a few videos about FFT in Max)
You can however incorporate gen~ processes inside the pfft~ object
Haha funny intro