Basslines - Opening Filter Raises Volume

[Kriminal]

Disclaimer: i dont actually do that myself

so what do you do?

Partially deafens himself ever time he does a filter sweep

Actually, i only open the res up on my acidlines, and i like them to get loud and nasty. I usually keep a flat type of bassline going at the same time.

[Quaternion]

When the filter is open, just reduce the hi-freqs with an EQ to a desireable level, without altering the low frequencies.

A compressor/limiter (unless multiband) is likely to introduce changes in the volume of the low frequencies when you open the filter of your bassline.

That's why I think an EQ is the solution you are looking for in the scenario you described.

[C00kie]

Opening the filter, and then attenuate high freqs.. Wouldn't that defeat the purpose??

[BONES]

The problem I have is that opening the filter raises the volume (as the high frequencies are let thru the filter I guess).

I often have the opposite problem, especially with WaspXT. As I close the filter, the resonance pushes the level through the roof. It doesn't really sound any louder but the level gets very high. I fuckin' love it!

[Rozzer]

A simple compressor should solve 90% of these problems. I always compress my bass channel... and my drums... and my drum groups... and....

ah f**k it, I just love compressors

[phatty]

put a nasty brickwall limiter on each channel.

[Kim Lajoie]

Ok, just to play devils advocate:

Single-band compression won't sound good because as the filter opens, the level rises, and the compressor lowers the gain. The overall gain for the channel. This mean that that bass sound gets quieter. The highs get quieter (which is fine), but the lows get quieter as well. Unless you've got some heavy multiband compression further down the DSP process tree (such as a group or on the master bus), then your low bass will actually get softer as you open the filter. Maybe good for jazz or country/western... but bad news for club/dance music.

Multiband compression and EQ will be slightly better, becuase they both will be configured to keep the lower bass level (close to) constant while selectively attenuating the highs. This will reduce the perceived level rise somewhat, but then you have a cruel trade-off - the more constant you want the level, the less pronounced the filter opening can be. Likewise, the more pronounced you want the filter opening, the less constant the level will be.

I hope you appreciate the technical conundrum we have here.

The real problem is that of perception. When you open the filter of a sound, that sound naturally gets bigger. I'm not just talking about level - I'm talking about the perceived size of the sound. In the case of the filter opening, the cues we have are the increased level of the sound, and the increased harmonics as well. Spectrally, the sound gets bigger. So let's talk about what you're trying to do ...

If indeed you're trying to make the sound bigger, you may also find it useful to approach it as a composition element, not a sound design element. This is, let the other instruments move with the bassline. Here you'll have to refer back to your larger-scale plan for the piece. You can have several other instruments become larger simultaneous with the bassline, to create an effect of the whole music growing. Alternatively, you could have some other instruments get smaller as the bassline gets bigger... that way the smaller instruments will "make space" for the bassline. If you do this well, they will not only make space spectrally (technically), but also psychologically. It's very important to be mindful of the psychology of the listener.

If you don't want the bassline to get bigger, then perhaps you have to reconsider why you are opening the filter. If you want directionless movement, then there are other signal processing techniques that can achieve this. Using a phaser can be cool. If your bassline does not have enough upper harmonic content to make the phasing audible enough, you can consider layering the main bassline with another one, richer in harmonics, but further back in the mix. Apply the phaser to the harmonically-rich layer.

Another signal processing technique you can consider is waveshaping distortion. If the level is kept reasonable (via an appropriate algorithm, or by careful compression downstream), waveshaping can be applied in such a way that it changes the sound without changing its size. Let me know if you want more hints about that.

I hope that gives you some food for thought.

Apologies for any incoherence. I'm really tired.

Forever,




Kim.

[Quaternion]

Another signal processing technique you can consider is waveshaping distortion. If the level is kept reasonable (via an appropriate algorithm, or by careful compression downstream), waveshaping can be applied in such a way that it changes the sound without changing its size. Let me know if you want more hints about that.

Yes, I do, please!

[Sascha Franck]

(Multiband) compression might be one way to go, but let me ask: What are you using for your basslines?

In case you're using any modern synth it could most likely be set up in a way that the volume is modulated along with the filter cutoff simultaneously (= using the same CC).

Still Jeez' comment would be valid: By lowering the overall volume the low frequencies would be lowered as well.

When thinking about it, a plain compressor should perhaps just do the job fine. By compressing the complete range the output volume of the "non-offensive" frequencies (i.e. the bass frequencies) should stay around the same.

[useruseruser]

Hello,

Kinda self explanatory really. A lot of my tracks involve a bassline growing by opening the filter. The problem I have is that opening the filter raises the volume (as the high frequencies are let thru the filter I guess).

I usually overcome this by fading the wave file slightly. Is there a better solution, would compression make any difference?

I would try this myself but I'm at work.

That's a problem with many filters, but not all.
Whatever, compressing the track smartly can make it better but contra-automate the volume brings very often better results. Sometimes you can design modulation routing internally at the synthesizer stage (cutoff control increases and at same time decreases volume). The last one prevents you from tweakin arround with automation tracks and you can save this patch for further usage.

.

[BONES]

Excuse my ignorance but isn't the whole idea of automating the filter to create a dynamic effect? If so, the percpetion of increased volume will aid said idea, will it not?

[Sascha Franck]

Excuse my ignorance but isn't the whole idea of automating the filter to create a dynamic effect? If so, the percpetion of increased volume will aid said idea, will it not?

In a way you're completely right.
Just sometimes you may only want to alter the sonic quality, not the overall level.

[Kim Lajoie]

Another signal processing technique you can consider is waveshaping distortion. If the level is kept reasonable (via an appropriate algorithm, or by careful compression downstream), waveshaping can be applied in such a way that it changes the sound without changing its size. Let me know if you want more hints about that.

Yes, I do, please!

Waveshaping distortion can be a good way of modifying the harmonic content of a sound. Coupled with pre and post eq/filtering, a great deal of control can be exercised. Ok, now for some specifics.

Pre eq/filtering
This is used to modify the character of the waveshaping. More specifically, this stage can be used to emphasise and de-emphasise particular frequency ranges. This causes the waveshaping stage to respond differently to different frequency areas. For example, boosting low frequencies at this stage will make the waveshaper response more to bass content. It will become more sensitive to lower frequencies. Likewise, reducing the bass content at this stage will make the waveshaper less sensitive to bass content.

At this stage, filtering is just an extreme form of EQ - instead of merely reducing a frequency area, filtering will completely remove that component... perhaps with a frequency emphasis as well.

Waveshape
Most waveshaping distortion is expansive - that is, it increases the "spacial" area between the waveform and DC, making the sound louder. Without going into details about transfer graphs and the like... I'll just say that this is not the only form of waveshaping. There are transfer types that make the sound "thinner", as well as any bizarre in-between. Bram's Cyanide is good fun to play with.

Post eq/filtering
This stage is where you'd typically compensate for any frequency irregularities caused by the previous two stages. For example, if the pre eq stage de-emphasised the bass and emphasised the treble (this is a typical pattern for making waveshaping sound pleasing - especially on complex material), then this (post eq) stage could be used to increase the bass and reduce (or even filter) the treble, to make the resulting sound closer to the original. This stage can also be use to control the size of the resulting sound very effectively, by using filtering to remove upper harmonics created by the waveshaping.


So, if we want to apply this to a bassline to create movement without changing the size of the sound, I'd take this approach:

1) Set the waveshaper to something close to what sort of sound I want. Set it to have a pretty strong effect.
2) Configure the post eq/filter stage to keep the sound where I want it in the mix (the waveshaping will probably add too much high frequency content).
3) Automate the pre-eq stage, probably by adjusting the centre frequency of a mid/high-mid boost. This will change the character of the distortion, but without having a significant impact on the overall size of the sound.

Forever,




Kim.

[Muon Software Ltd]

When designing a filter you pretty much have two options regarding the overall output level.

You can normalise the filter so that the passband is at unity gain and the frequencies in the resonant peak are boosted above unity gain...or you can normalise the filter so that the resonance amount is at unity gain and for every dB you increase the resonance the passband gain drops a dB.

The filter in the 303 bassline follows the latter design and overall output gain of the filter falls off as the resonance increases, keeping the resonant peak around unity. Basically the overall output gets quieter as you increase the resonance. The 303 has quite some interesting dynamics for such a simple instrument, so if you're using a good emulation you might be needing some kind of compression or even (as another has suggested) a bit of distortion (which normally decreases the dynamic range anyway).

Cheers
Dave
Muon Software Ltd
www.muon-software.com

[kritikon]

I never knew that. Thanks Dave...useful information. I always wondered why a 303 didn't blow up my monitors, but other synths would. I learnt the hard way with a 202 on high res Now I also know why the MS10 was so good at basslines - it always retained the bass content.