I'm a softsynth and DAW newbee. So there might be an easy solution to the following problem I'm not aware of. My DAW is Studio One. But I'll switch to another DAW if that's the only way to solve the following problem:
By playing a polyphone softsynth live (both hands playing multiple notes at the same time) I realize that for my ears and musically needs most of my softsynths play notes on the lower end too quite and the higher notes too loud. So I like to adjust the overall balance of low vs higher note levels along the keyboard range. I suppose this problem is quite common and hopefully the solution also.
Because beginning at the lowest key each higher note gets a bit louder a linear level correction of the whole note range would be ideal. For example using zero level correction on the "Middle C", increasing level of the lower keys and decreasing level of the higher ones continuously.
But unfortunately I don't know how to achieve that. I can think of three solutions:
1. Internal virtual instrument option: "volume key tracking"
Drawback: I was told that some hardware synthesizers have this option. But none of my softsynths provide it. Only some of them got "filter key tracking" but that is not what I'm searching for. Ableton Live "Operator" was the only synth I found that got "volume key tracking". Look the first two minutes of this video:
2. Realtime velocity transforming in DAW
Drawback: Depending on the synthesizer patch manipulating the velocity level can change the sound. The same if a patch consists of multiple layers which are velocity sensitive. And finally I even don't know how (or even if) realtime velocity transforming is working in any DAW.
3. EQ or multiband compressor
Drawback: High pass filters also affecting overtones of the lower keys and low pass filters do affect the sound of higher notes. The same problem with multiband compressors.
So what is the best way to go? With minimal compromise on sound quality and because it should work in realtime playing live without adding additional latency?
How adjust softsynth low/high keys level
-
- KVRist
- 144 posts since 1 Jul, 2015
I think this is what you need. Use in such a way that the higher the pitch of the note, the lower the offset pitch of built-in filter - it moves more to the left. And filter type in the softsynth is "low pass 1" for example, that doesn't cut high frequencies too much.mireiner wrote: Tue Mar 26, 2019 8:38 pm Only some of them got "filter key tracking" but that is not what I'm searching for.
-
- KVRian
- 626 posts since 15 Jun, 2015
So... are the notes on the low end quieter? Or just perceived quieter?
I can't remember if Studio One has a native spectrum analyzer, but if not, download a free copy of Voxengo's Span. Are the bass notes actually quieter or do they just sound quieter? Is it your speaker's/headphone's response? Is it the acoustics of the room? Is it just the natural response of the instrument? Is it your ears? In orchestral music, there are fewer woodwinds in the upper scale because the audience perceives the higher pitches of a single piccolo as louder than several bassoons.
On a lot of synths, both hardware and software, key tracking is used to simulate the subtle timbre change heard on real instruments. For example, a piano or guitar sounds slightly differently across the full scale of its keyboard or strings. Higher notes on a piano are "brighter", lower keys are more muted--filter key tracking allows you to simulate this real-world effect. Yes, as you've noticed, the timbre is also different depending on the velocity of the note, so trying to solve this with velocity is like always plucking the lower notes on the guitar a lot harder than the higher notes.
Take a look at your sound levels on a spectrum display and you'll have a better idea or what to do, if anything.
I can't remember if Studio One has a native spectrum analyzer, but if not, download a free copy of Voxengo's Span. Are the bass notes actually quieter or do they just sound quieter? Is it your speaker's/headphone's response? Is it the acoustics of the room? Is it just the natural response of the instrument? Is it your ears? In orchestral music, there are fewer woodwinds in the upper scale because the audience perceives the higher pitches of a single piccolo as louder than several bassoons.
On a lot of synths, both hardware and software, key tracking is used to simulate the subtle timbre change heard on real instruments. For example, a piano or guitar sounds slightly differently across the full scale of its keyboard or strings. Higher notes on a piano are "brighter", lower keys are more muted--filter key tracking allows you to simulate this real-world effect. Yes, as you've noticed, the timbre is also different depending on the velocity of the note, so trying to solve this with velocity is like always plucking the lower notes on the guitar a lot harder than the higher notes.
Take a look at your sound levels on a spectrum display and you'll have a better idea or what to do, if anything.
-
- KVRer
- Topic Starter
- 28 posts since 4 Feb, 2019
Hi RichieWitch,
I look at the frequency spectrum in Fabfilter Pro Q3 all the time.
It's just as I wrote before. On a 49 key keyboard the lowest keys C1, D1, E1, F1 and G1 got a maximum level of +3db. The midrange key C3 maximum level is +6db and the highest key C5 maximum level is +9b. So from lowest to highest key there's a difference of 6db. The higher notes sound even loader because they are more cutting through.
Most soft synths I tested behave like this. Especially the low keys C1-G1 are weak in the tonal range below 80Hz. That is no problem if you edit your songs on the computer. Then you can record lower and higher tones on different tracks and mix them seperatly. But if you perform live like me you can't do that.
So again what's the best way to solve this problem? Is using an EQ and a multiband compressor already the best thing I can do about it? Even if it's a suboptimal solution?
I look at the frequency spectrum in Fabfilter Pro Q3 all the time.
It's just as I wrote before. On a 49 key keyboard the lowest keys C1, D1, E1, F1 and G1 got a maximum level of +3db. The midrange key C3 maximum level is +6db and the highest key C5 maximum level is +9b. So from lowest to highest key there's a difference of 6db. The higher notes sound even loader because they are more cutting through.
Most soft synths I tested behave like this. Especially the low keys C1-G1 are weak in the tonal range below 80Hz. That is no problem if you edit your songs on the computer. Then you can record lower and higher tones on different tracks and mix them seperatly. But if you perform live like me you can't do that.
So again what's the best way to solve this problem? Is using an EQ and a multiband compressor already the best thing I can do about it? Even if it's a suboptimal solution?
-
- KVRAF
- 7540 posts since 7 Aug, 2003 from San Francisco Bay Area
What you are describing is not a typical problem which most people experience, notice, or have to correct for. Perhaps it is an issue with your specific keyboard, your playing technique, or monitoring environment. It isn’t accurate to characterize this as a common problem which many or most synths suffer from.
Incomplete list of my gear: 1/8" audio input jack.
-
- KVRian
- 626 posts since 15 Jun, 2015
Sorry I couldn't be more helpful. In my studio, I have the exact opposite problem. Whether I use softsynths or hardware, the low end is always much louder than it sounds on my headphones or monitors. When having my tracks professionally mastered, I've had to use a spectrum analyzer to find and trim the low end because the engineers I've worked with pointed out the problems that I couldn't hear.
I should probably just buy a subwoofer...
The only other solution I've found is constantly running out to my car and testing there. 
I should probably just buy a subwoofer...
The only other solution I've found is constantly running out to my car and testing there. -
- KVRAF
- 9571 posts since 6 Jan, 2017 from Outer Space
It depends on your DAW as well. In Bitwig you could define a modulater which is the Midi note, and that could simply adjust the output level of the synth according to a curve of your taste...
Other hosts might have other options...
But you should definitely do some research why that behaviour is there. Usually a synth played with velocity 127 should output the same peak level on any key. Unless the patch applies some difference via key tracking. To test just use a single oscillator and bypass all filters and treatments. If there is still a difference check your keyboard if it really pulls the full velocity on all keys...
Other hosts might have other options...
But you should definitely do some research why that behaviour is there. Usually a synth played with velocity 127 should output the same peak level on any key. Unless the patch applies some difference via key tracking. To test just use a single oscillator and bypass all filters and treatments. If there is still a difference check your keyboard if it really pulls the full velocity on all keys...
-
- KVRer
- Topic Starter
- 28 posts since 4 Feb, 2019
Thank you deastman and Tj Shredder.
After reading your posts I realized that my statements might lead to misunderstanding.
When I play a single note at a time across the whole range of my 49 key midi-keyboard (from C1-C5) they got all the same level on the stereo level meter.
But when I watch the same single notes while playing in a spectrum analyser (Fabfilter Pro Q3) then the overtone curves of C1 reaches a maximum level of +3dB, C5 +6dB and C5 +9dB maximum level. I did test this also with Voxengo SPAN as suggested and the difference between the overtone curves of low notes and high notes is the same.
Theses maximum levels of overtone curves in the spectrum analyser correspond to what I'm hearing when listening to the sound that come out of my speakers or my Sennheiser HD 600 headphones. The latter got a frequency range of 12-40000Hz +-3dB.
My midi-keyboard outputs correct velocity values as I can see in the midi track. But to be sure I even bypassed my midi-keyboard and did manually edit the single notes in the midi track editor using my pc keyboard. The result is the same.
My USB audio interface is RME Fireface UCX with external power supply running newest ASIO drivers. Never had any trouble with this device.
After reading your posts I realized that my statements might lead to misunderstanding.
When I play a single note at a time across the whole range of my 49 key midi-keyboard (from C1-C5) they got all the same level on the stereo level meter.
But when I watch the same single notes while playing in a spectrum analyser (Fabfilter Pro Q3) then the overtone curves of C1 reaches a maximum level of +3dB, C5 +6dB and C5 +9dB maximum level. I did test this also with Voxengo SPAN as suggested and the difference between the overtone curves of low notes and high notes is the same.
Theses maximum levels of overtone curves in the spectrum analyser correspond to what I'm hearing when listening to the sound that come out of my speakers or my Sennheiser HD 600 headphones. The latter got a frequency range of 12-40000Hz +-3dB.
My midi-keyboard outputs correct velocity values as I can see in the midi track. But to be sure I even bypassed my midi-keyboard and did manually edit the single notes in the midi track editor using my pc keyboard. The result is the same.
My USB audio interface is RME Fireface UCX with external power supply running newest ASIO drivers. Never had any trouble with this device.
Last edited by mireiner on Wed Mar 27, 2019 4:11 pm, edited 2 times in total.
-
- KVRian
- 626 posts since 15 Jun, 2015
It's just the overtone curves? So you're talking about the sum of the gain of the harmonics above the fundamental frequency?
If I'm understanding you correctly, you're asking to suspend the laws of physics.
Sorry if I sound flippant, no offense intended.
Across any given timbre, the lower notes have fewer audible harmonics because the waveform is longer/slower, and so fewer harmonics exist to add their gain to the overtone curve. Take any fundamental frequency and calculate the harmonics yourself and you'll see.
Given a constant velocity, all the fundamental frequencies should have approximately the same gain, depending on key tracking and such. After that, the physics of sound take over.
If I'm understanding you correctly, you're asking to suspend the laws of physics.
Sorry if I sound flippant, no offense intended.
Across any given timbre, the lower notes have fewer audible harmonics because the waveform is longer/slower, and so fewer harmonics exist to add their gain to the overtone curve. Take any fundamental frequency and calculate the harmonics yourself and you'll see.
Given a constant velocity, all the fundamental frequencies should have approximately the same gain, depending on key tracking and such. After that, the physics of sound take over.
-
- KVRer
- Topic Starter
- 28 posts since 4 Feb, 2019
Hi RichieWitch,
forgive me - I'm not a native english speaker and maybe "overtone curves" isn't the correct term?
With "overtone curves" I mean the curves that are shown in a spectrum analyser. The frequency spectrum that is shown there got many small peak curves on the upper side. And the maximum level of one of these curves on the upper side of the frequency spectrum are at note C1 +3dB and on C5 +9dB. About this difference between low and hight notes I'm speaking.
Do you really wanna say that doesn't matter?
forgive me - I'm not a native english speaker and maybe "overtone curves" isn't the correct term?
With "overtone curves" I mean the curves that are shown in a spectrum analyser. The frequency spectrum that is shown there got many small peak curves on the upper side. And the maximum level of one of these curves on the upper side of the frequency spectrum are at note C1 +3dB and on C5 +9dB. About this difference between low and hight notes I'm speaking.
Do you really wanna say that doesn't matter?
-
- KVRian
- 626 posts since 15 Jun, 2015
Okay, apparently, I *am* confused about what you mean. So.... I opened my DAW and tested what I thought you were talking about. The following two screenshots are the harmonic spectrum of a single sawtooth wave, the first at C5, the other at C1. Notice that the meters on the left are nearly identical for both notes. So there goes my theory. Can you provide screenshots of what you're seeing on your system? I still don't get it.
You do not have the required permissions to view the files attached to this post.
-
- addled muppet weed
- 111299 posts since 26 Jan, 2003 from through the looking glass
yeah.RichieWitch wrote: Wed Mar 27, 2019 4:09 pm
If I'm understanding you correctly, you're asking to suspend the laws of physics.
and?
you saying you cant do it?
-
- KVRian
- 626 posts since 15 Jun, 2015
I'm weak!!!vurt wrote: Wed Mar 27, 2019 5:11 pmyeah.RichieWitch wrote: Wed Mar 27, 2019 4:09 pm
If I'm understanding you correctly, you're asking to suspend the laws of physics.
and?
you saying you cant do it?
-
- KVRer
- Topic Starter
- 28 posts since 4 Feb, 2019
So here is the spectrum analyser view of the free soft synth:
"u-he Tyrell N6 V3" with factory preset "3ee_SawK" playing the single note C1, C3 and C5.
As you can see peek level of spectrum curves are:
C1 = +3dB
C3 = +6dB
C5 = +9dB
And the stereo level meter on the right side of the screen shows that overall volume of these three notes are the same about +10dB. That is the typical situation I get with the majority of synths I tested. The notes are played with the same midi velocity.
"u-he Tyrell N6 V3" with factory preset "3ee_SawK" playing the single note C1, C3 and C5.
As you can see peek level of spectrum curves are:
C1 = +3dB
C3 = +6dB
C5 = +9dB
And the stereo level meter on the right side of the screen shows that overall volume of these three notes are the same about +10dB. That is the typical situation I get with the majority of synths I tested. The notes are played with the same midi velocity.
You do not have the required permissions to view the files attached to this post.
Last edited by mireiner on Wed Mar 27, 2019 8:05 pm, edited 1 time in total.
-
- KVRian
- 626 posts since 15 Jun, 2015
Hmmm.... I'm suspicious of the amount of sound information below the fundamental. Notice how there is nothing below the fundamental frequency in my examples?
That says to me that your synth patch might have a lot to do with the behavior, the shape and type of the filter, any FX processing like reverb/delay/EQ on an FX screen in the synth maybe. Can you do it with a single oscillator/simple waveform and all FX turned off? A square wave or sawtooth perhaps?
That says to me that your synth patch might have a lot to do with the behavior, the shape and type of the filter, any FX processing like reverb/delay/EQ on an FX screen in the synth maybe. Can you do it with a single oscillator/simple waveform and all FX turned off? A square wave or sawtooth perhaps?