MSF - Moog Style Filter / MTurboFilter

[jmg8]

Operator, could I ask you to do something please (or anyone else that is interested)
Please would you run a basic test so I can hear the difference between the Melda filters and the others that you speak of. This way we can directly compare how the filters sound.

If you, myself or anyone else is to add to this test, I think we will need some basic rules to keep things fair:
Input should be a basic SAW wave at A0 55hz (this will give a low frequency to hear what is going on in the low end plus plenty of harmonics to clearly hear what the filter is doing in the highs. Hopefully it should sound mostly the same in any synth)
Filter frequency should modulate between 20hz and 20khz (make sure any key tracking is off, as this could change the range depending on which key is used)
Modulator should be a sine wave LFO at 1hz (record a few second of this to give chance to hear the sweep a few times)
Any info should be mentioned (such as the name of the filter, name of synth, how much resonance, filter slope 24db etc)

Is this a fair test? Is there anything else to add here? I'll start by adding the first one (Melda) hopefully others will add too. I am really interested to hear how much difference there is.

[jmg8]

Does anyone know the best way to share an audio file here? (in such a way that it can be clicked on and played right here in the browser) I saw someone did this with Dropbox, but when I took a look there I couldn't work it out.

[Faiky]

You can copy the link in dropbox and change the dl=0 to dl=1. The rest is done by the forum.
Test: https://www.dropbox.com/s/fzv7bb0ve8t1k ... f.mp3?dl=0
Very nice idea jmg8! I wonder what will come of it.

edit: Works with dl=0, too.

[jmg8]

Melda LP12 25%Res
https://www.dropbox.com/s/e6slsty0jufh2 ... s.wav?dl=0

Edit: This doesnt seem to play back for me. Does it work for anyone else? Not sure what I did wrong.

[operator]

Operator, could I ask you to do something please (or anyone else that is interested)
Please would you run a basic test so I can hear the difference between the Melda filters and the others that you speak of. This way we can directly compare how the filters sound.

If you, myself or anyone else is to add to this test, I think we will need some basic rules to keep things fair:
Input should be a basic SAW wave at A0 55hz (this will give a low frequency to hear what is going on in the low end plus plenty of harmonics to clearly hear what the filter is doing in the highs. Hopefully it should sound mostly the same in any synth)
Filter frequency should modulate between 20hz and 20khz (make sure any key tracking is off, as this could change the range depending on which key is used)
Modulator should be a sine wave LFO at 1hz (record a few second of this to give chance to hear the sweep a few times)
Any info should be mentioned (such as the name of the filter, name of synth, how much resonance, filter slope 24db etc)

Is this a fair test? Is there anything else to add here? I'll start by adding the first one (Melda) hopefully others will add too. I am really interested to hear how much difference there is.

Not sure if this is such a good idea... THEN such standardized/technical tests probablly won´t bring some results in relation to musicallity. Also there is to much risk of wrong testing and when we all compare our results, who knows who did it right and who made a little mistake that curops the results...

I found comparing filter-sweeps doesn´t always bring out the strenght and weaknesses of a filter and it´s capabilities to create musical sounds. In a neutral state almost all filters will sound similar. But creating e.g. a TB-303 stlye acid-lead (a Lead with very high resonance) in a Synth and playing with the resonance will bring out the character. Playing with the resonance at such an extreme state will show if the filter is muscal or not... and with sylenth for example i get much better results than in MSF...

This is the thing I´m refering to when I say it is sounding "cold"... not that the filter slope is weird or some missing distortion inside the Filteralgorithm isn´t creating some noise which fakes anolog sound. It is the behaviour throughout all it possible states of a filter and all of its combinations. MSF´s Filters aren`t bad, they just dont behave as musical constant as others. This is best observed when playing with the resonance at high settings...

Sometimes (when testing musicality of a piece of software) it´s better to test the borders/extremes/ the upper and lower limits of a synth, then most of the time this is the area where the magic happens. For example when using Bazille and using the Filter at it´s almost self-resonating position... feeding some signals into it will create a very pleasing sound compared to MSF´s Resonance.

One Major Problem that will occur in testing will be: When comparing MSF filter to others, the different behaviour of the Resonance from MSF Filter. I used VST Analyzer to completly match the curve of the ZEbra XMF Filters... but a slight change in either MSF or Zebra Filter will shift the slopes ---> especially with resonance added.

It is better to test by creating simple patchs in MSF and e.g. Sylenth and compare the behaviour of the paramters when being dialed in.

DIALING IN is the Buzzword here... because this is where the musician and the synth/sound meet... Dialing in sounds, hearing looking for sweetspots and noises... and the better the Filter is the better this dialing in will work. Sorry for this esoteric explanation but this is trully the point where others really shine... e.g. Hive or Bazille. Good Filters will augment the sound and will help you carve out the frequencies you really need... with MSF this is sometimes very hard. Without my experince and the knowledge that a certain sound needs to be at a certain setting in a synth, I often would have over-heard the sweetsport in MSF.

For me it doesn´t need any testing... in the last months I came across this observation that the Filters don´t act in such a way, that I can use it without thinkning too much about them... (gain compensation, constant spectrum, etc..) They act a lot differently than others e.g. u-he, sylenth... the resonance behaviour is one example. I could get used to the resonance behaviour, if it would benefit a musical results.

Time is limited and before i would do intense testing (AGAIN )... I rather would like to test some new Analog Filters that Vojtech maybe finds the time to code ... because like I said, I heard the differnce and for me it doesn´t need no further testing to get to a conclusion I already know.

And I think everybody else can observe the same by just playing and listening to the comparison of MSF to e.g Sylenth... Dialing in some sounds, play some sequence and play with resonance and cutoff. No need for complicated test... I think we all hear the same thing, if we try to be honest to ourselves...

[Faiky]

Hm, melda always gives us the opportunity to exaggerate, to go even further than other synthesizers and I like that. It only sounds good in a certain range, that's clear and it's more difficult to get musical results. But I think it's good that I have control over it myself.
The resonance is exactly in the frequency range where I set it, keytracking works perfectly. The resonance is constant, but I can modulate it for ACID, if i want.

But I also want to have analog simulations that mimic the sound and behavior. You can read a lot about this on the internet. But as a starting point I think such a test is good!
In the end it is important which sound comes out in which setting, not how something is constructed.

[operator]

Self-resonating Filters.... This would be another reason to implement Analog Filters. It is an extremely usefull method to play into an almost self-resonating filter which for a short time will add a sound wich decays away dependend on the input amplitude.

[operator]

sry

[operator]

If y'all want to make tests, I would strongly recommend to try to test/compare only the filters... not the whole synth.

My test-scenario was: MSF as Oscillator (no filter). Followed by MTruboFilter, Zebrify and/or SerumFX... then I switched the later ones on and off one after another.
This way I could eliminate differences in the OSC of a Synth and more important the Input gain into the Filter... so I really could only compare the Filter.

[jmg8]

Operator the test isn't for you specifically. You may have found the conclusion you are looking for, that's good for you. The test is for everyone to hear how filter types differ in a controlled way. If you remove the rules then the test will not directly compare them in a fair way. If you don't want to join in, that's fine, I just figured your were interested in this and would want to.
You say that the difference between the filters is more heard when in a musical context and when pushed to the limits. So how about this...make a very simple patch that can be replicated in MSF so we can hear the difference of the filters.
I'm not trying to target you here BTW, I am genuinely interested to find out about this "good filter" vs "bad filter" difference.

[operator]

Operator the test isn't for you specifically. You may have found the conclusion you are looking for, that's good for you. The test is for everyone to hear how filter types differ in a controlled way. If you remove the rules then the test will not directly compare them in a fair way. If you don't want to join in, that's fine, I just figured your were interested in this and would want to.
You say that the difference between the filters is more heard when in a musical context and when pushed to the limits. So how about this...make a very simple patch that can be replicated in MSF so we can hear the difference of the filters.
I'm not trying to target you here BTW, I am genuinely interested to find out about this "good filter" vs "bad filter" difference.

Ohhh, I sincerly apologize... didn´t want to sound so rude. Sometimes I say/write things without thinking about how others will perceive me. SORRY Wasn´t ment this way. Maybe I´m just a little bit too furstrated/dissapointed when it comes to the Filters in MSF.
And yes you are right... this would serve a greater good. And yes, I´m always curious about stuff like that and will try to make a comparison/example. I hope i can capture it in a good way, so it will highlight my concern.

[jmg8]

Thank you so much I appreciate your time.
I have downloaded Zebrify and Cytomic-The Drop and started to compare them with MTurboFilter.
I hope our tests will reveal more than I have found so far. I actually was able to make MTF sound the same in most cases and when it was different, I'm not sure I could say if one sounded better than the other, just slightly different.
When it comes to MTF time, there will be people (myself being one of them) creating filter algorithms for it. These tests will hopefully be useful to show how the filters can differ and how to recreate them.

I am only at a basic level at understanding the internal stuff that makes up a filter. However, from how I see it now, it seems to me that, filters usually sound different due to slope and resonance changing across the frequency range. For example, a filter might have higher resonance in the upper frequencies and a steeper slope in the lower frequencies. This can easily be replicated in MTF.

But what am I missing? What do these other filters do that can not be replicated in MTF?
So far I only notice 1 thing.....MTF can not self resonate.

I hope we find some interesting results from this. Operator is not the only person to mention that Melda filters don't sound as good as some of the competition. What I wish to achieve is, not IF they do sound different but HOW and WHY.........

[operator]

Thank you so much I appreciate your time.
I have downloaded Zebrify and Cytomic-The Drop and started to compare them with MTurboFilter.
I hope our tests will reveal more than I have found so far. I actually was able to make MTF sound the same in most cases and when it was different, I'm not sure I could say if one sounded better than the other, just slightly different.
When it comes to MTF time, there will be people (myself being one of them) creating filter algorithms for it. These tests will hopefully be useful to show how the filters can differ and how to recreate them.

I am only at a basic level at understanding the internal stuff that makes up a filter. However, from how I see it now, it seems to me that, filters usually sound different due to slope and resonance changing across the frequency range. For example, a filter might have higher resonance in the upper frequencies and a steeper slope in the lower frequencies. This can easily be replicated in MTF.

But what am I missing? What do these other filters do that can not be replicated in MTF?
So far I only notice 1 thing.....MTF can not self resonate.

I hope we find some interesting results from this. Operator is not the only person to mention that Melda filters don't sound as good as some of the competition. What I wish to achieve is, not IF they do sound different but HOW and WHY.........

This is exactly what i mean... at a certain setting the filter sounds the same or similar to all the others (hive, Zebrify, etc) And in a VST Analyzer you can even match them 100%... BUT as soon as you tweak the MTF setting you always need to tweek input gain, resonance and gain for every setting to match again with Zebrify Filters. It is not as well balanced as e.g u-he filters or Sylenth. (not as musically)
BUT I think for people to call a filter a "good filter" it is also important that it behaves in a pleasing in a balanced way...

It is the same as a guitar... 6 strings which sound good on their own, but as soon as you play them toghether they need to be always in a certain relationship. The Parameters in a Filter probably need the same finetuning including all parameters over the whole spectrum.

What is missing in the MTurboFilter? It is a musically consistent tuning of all those parameters in relation to each other... this is what I wrote above. They are not bad, only they could be better optimized... like you said --> e.g resonance adapts, gain adapts to resonance chages, etc...

As I tried to make a comparison example for you... the only thing I came up with is, that I would have to make a video comparing Zebrify and MTF and their relation between gain, cutoff and resonance when dialing in the sound.

Maybe you could now try that for yourself:
When you installed Zebrify --> try to send a low Saw (Bass range) into the Zebrify Filter (MS2035) and then use your mouse to move the cutoff with some wobble movements in the low end area. Listen to the low Thump it creates when sweeping fast through the bass area... then try this in MTF --> no such thick low Thump and as more you increase the Analog Knob the less of this thump you get... I don´t use the Analog Knob very often for this very reason... it seams to make the filter worse most of the time.

Keep me updated, if you could hear also some slight fuller, more coherent, less watery sound in the Zebrify Filters compared to MTF.

[werzel]

Do some synths have default or built in (coded) envelopes that dictate how the filter or its Q react to incoming signals? I'm asking because that is how I understand that early analogue synths operated. (I reserve the right to be both corrected and wrong!)

[DarkStar]

Have a look at
https://www.soundonsound.com/techniques ... ationships
https://www.soundonsound.com/techniques/further-filters
You might even like the whole series: https://www.soundonsound.com/series/synth-secrets

Also
https://ask.audio/articles/synthesis-es ... ur-filters
https://reverb.com/uk/news/a-guide-to-s ... cale=en-GB