Ultra-Pure Sub-Bass Patch (no clicking in attack)

[jheronymo]

Yeah totally. I never thought about it this way, but a gate is just the fastest possible envelope (with no AD controls).

As you add a longer and longer ramp up/down, I’m guessing there is a point where the clicks become inaudible. Where that point is probably depends on the pitch of the audio, and how sensitive you are to clicks (which after sitting in the studio listening to clicks for hours is usually pretty damn sensitive ).

Having some harmonics in your audio can mask the clicks too, which is why it’s the most problematic with sine subs.

[wuff_miggler]

^ im sensitive - i use a declicker on all my vinyl rips. I wont have noise that's not part of the signal, in my chain. if i want noise - i'll add it myself

[jheronymo]

^ im sensitive - i use a declicker on all my vinyl rips. I wont have noise that's not part of the signal, in my chain. if i want noise - i'll add it myself

Agreed

[sir stony]

An easy solution is to filter the envelope a bit if you can't reset the phase of the oscillator to 0 (even then the jump to 0 can make a click). If the attack is shorter than a cycle of the wave, it clicks.

A quarter of the wave cycle will do, I think. You just need to ramp up from zero to a maximum in either direction.
At very low frequencies and especially in lfo range, this may even be faster than a zero-crossing detection VCA, as such may need up to a (near) half wave cycle.

[Peake]

I believe the minimoog's minimum attack time is something like 1.5ms which would help in avoiding non-zero crossing artefacts.

[luketeaford]

An easy solution is to filter the envelope a bit if you can't reset the phase of the oscillator to 0 (even then the jump to 0 can make a click). If the attack is shorter than a cycle of the wave, it clicks.

A quarter of the wave cycle will do, I think. You just need to ramp up from zero to a maximum in either direction.
At very low frequencies and especially in lfo range, this may even be faster than a zero-crossing detection VCA, as such may need up to a (near) half wave cycle.

That makes total sense to me-- nicely reasoned!

[nuromantix]

Maybe try a different eg for controlling the vca. I guess you want a logarithmic curve .... Maths springs to mind...

[NoGuitars]

Haven’t tried myself, but is a LPF after the VCA not the simplest solution? No need for a specialised module and easily tested.

[jheronymo]

In practice, any of these should work. All rely on having an extra module, so it depends on your individual system.

Zero-crossing is the “perfect” solution since you can use the original gate/ADSR, but there’s some latency. Worst case, at 20Hz it could be 25ms between zero-crossings (maybe noticeable). At 50Hz the worst case is 10ms and 100Hz it’s 5ms (almost surely not noticeable).

Using a slower envelope rather than gate (or slewing your gate) into the VCA has potentially less latency. You can probably get away with a constant ~5-15ms attack/decay. This might be the best overall. I bet it’s most likely that people have a spare envelope in their system.

How much of the click is removed by an LPF will depend on the slope of the filter. It will attenuate the click whereas the other two will eliminate it. Filters also mess with phase, so that’s something to keep in mind.

[MARK27]

While not any help to the A-110-4 problem, there are also solutions like the SSL ThunDrum, an analog dual sine VCO that is intended to create non-clicking sine-wave-based percussion.

I feel like this isn't the only analog circuit out there that does this, but my mind is failing to work with me right now.

From https://steamsynth.com/shop/euro-module ... odel-2140/

Synchronous Envelope

The ThunDrum employs a unique circuit we call “synchronous envelope” that deserves some explanation. Normally, if you create a patch that has very short, percussive attack times using a free running oscillator, VCA and envelope generator, it’s likely that you will hear slight clicks when triggered because the signal output could be turned on at ANY point in the VCOs waveform.

To prevent the clicks, our circuit waits for the VCO signal to cross the zero-voltage boundary before beginning the attack phase of the envelope, thus eliminating the click. This has some limitations and drawbacks, however. If the tone oscillator is set to a very low frequency (or stalled), the envelope attack will never start or will start with a slight delay. It may take a bit of control fiddling to get to the sound your looking for, but you’ll have fun experimenting.

[wuff_miggler]

i dont have any "clean" filters in my set up. ideally i'd want the cleanest filter . i have 4 filters, all are noisy/character type.
there's been great suggestions in here for a fix - will be on the modular today. hopefully will sort it out

[Stice]

If your VCO offers hard sync, try multing your gate into the sync input. That will reset the Osc to zero at the start of the envelope*

* doesn’t always work in real life for whatever reason.

[SingleCell]

Hard synching has the potential issue of... if you trigger the next note before the VCA is all the way closed, sync will end up CAUSING a pop where there wouldn't have otherwise been one. Also, some oscillators may not reset to a zero crossing under synch. My Erica Black VCO 2 doesn't... though it may be very unusual that way. I also find that running a filter between the Osc and VCA can change the waveform enough that it still pops.

I think something like the WMD digital VCA might be the best solution. I haven't used one, but I've watched it demo'd on their youtube and it seems to work really well.

I recently got the Cosmotronic Delta-V and while I love how plucky and LPGish it can sound, it is TOO snappy for me most of the time. I've ended up running it thru a slew limiter. When set fast enough to not make the envelopes sound sluggish, it will still leave an audible pop on pure low frequency stuff... but it's more subdued, less bright and snappy. It's much more pleasant on my ears. I happened to have the WMD Time Warp handy, which is stupid overkill (features AND price), but it has an on/off switch so I can disable it without re-patching, when I do want super snappy.

[Ichor&Gore]

i dont have any "clean" filters in my set up. ideally i'd want the cleanest filter . i have 4 filters, all are noisy/character type.
there's been great suggestions in here for a fix - will be on the modular today. hopefully will sort it out

Were you able to remedy the click? Do you have a passive LPG? I do and will try this next time I set up as you and I have some similar Doepfer modules based on your posts. Clicks bug me too and I completely understand the power of the A-110-4 for bass. Wish that module would progress to an MKII or further advancement.

[wuff_miggler]

^ i can only fix this with filtering. I would really love to use this oscillator without the filter.
Using doepfer LPG softens the clicks - but is "coloured" the oscillator starts sounding too fat, looses the purity of the 110-4 running through the clean doepfer VCA. i had an idea to fix it - which was to tune 2 oscillators to the same note - very accurately checking on a scope...the non 110-4 oscillator would have its rect output running into a S&H module, and my gate signal driving the ADSR would be running into the S&H as well - opening the ADSR only when the gate was fired at the same phase of the rect/110-4 combo.

not sure if dieter has a recommendation - but the above is my only solution for this - and its a bit annoying as it ties up 2 x oscilaltors.

[LunaticSound]

I still feel, there is something wrong here.

Have you tried an envelope with a log/exp curve setting? Does it click on every attack? Or just some of them? Does it click, when no note was playing before?

At a certain degree of slapiness, the click will be there, but the click itself IS the slapiness then. If you want extremely fast attack, you want a click! (actually, clicking subs can be great when tightly combined with a kick)

A filter does the same as slewing the envelope, because it slews the waveform, which is the result of a multiplication of your envelopes voltage with the waveforms voltage. Just slewing the envelope right will produce the same wave and spare you the colorations.

[wuff_miggler]

I still feel, there is something wrong here.

Have you tried an envelope with a log/exp curve setting? Does it click on every attack? Or just some of them? Does it click, when no note was playing before?

i'll try again tomorrow - ive got time for it im fairly sure clicks happen all the time on sub bass material...i'll report back

i want what i can hear the WMD Digital VCA videos. I've tried every attack setting on the doepfer ADSR. using x1, x10, x100 very slowly changing attack settings. the click isnt slappiness - the click is a wave cycle after or before a zero crossing. It's common when people first get into audio editing, and wonder why their edits sound like ass. i want fast attack and i want the envelope to start with a zero crossing.

i dont seem to the only one though, so im assuming others with different envelopes trying the same thing arent getting far without a zero crossing adsr solution - i dont have an issue with higher pitched sounds Its subs mainly

[LunaticSound]

I get, where the click comes from, but consider this:

The VCA is a multiplier, which multiplies with values from 0 to 1. So if your note starts with an envelope sitting at zero, the result will be zero. Your resulting waveform will start at zero.

You are getting the clicks with subs, as the envelope can be faster than half a wavecycle, so if it ramps up while the subs cycle is at peak, you will basically just get the envelopes shape as resulting waveform. BUT if your envelope has an exponential shape, you will also get an exponential result, which is less squarey and also spends longer time lingering around zero, if you get what I mean.

I know that this is a common issue, but I still attribute zero crossing VCAs to cargo cult

So, I might have to eat my own words and apologize, if this isn´t helpful, but before you buy something you could still try changing the shape of your adsr by feeding back it´s output into it´s attack phse CV input. Attenuvert this fed back signal for making the resukting envelope shape exp/log and who knows, you might find out that there is a fine line between a click and a slap

[wuff_miggler]

So, I might have to eat my own words and apologize, if this isn´t helpful, but before you buy something you could still try changing the shape of your adsr by feeding back it´s output into it´s attack phse CV input. Attenuvert this fed back signal for making the resukting envelope shape exp/log and who knows, you might find out that there is a fine line between a click and a slap

thank you for mentioning this - as part of tomorrows trial - i will do this - but also check the manual

[wuff_miggler]

So, I might have to eat my own words and apologize, if this isn´t helpful, but before you buy something you could still try changing the shape of your adsr by feeding back it´s output into it´s attack phse CV input....

good news - you only have to eat a few of your words

so, the only time i have actually tried feeding the inverted out of my A-141-2 ADSR into the CV Attack input was when i tried it as an oscillator.
messing with the the log/exp really helps improve it. But - unfortunately to get it click free completely - a bit more impact than i'd like gets chewed up and to do that - I needed to use A-170 slew limiter (lower section on L setting) with the knobs turned up all the way.

good news is i can actually enjoy it more now. and i can get the clicks way less bitey. so over music - it should actually sound sweet. as with all doepfer - you need to really turn the knobs slowly - there's a definate sweet spot. i'm still learning this....

edit: i know a few users have mentioned slew. i was hoping not to have to use it - but without buying more modules...its the only thing i can do.