Chiptune song sounds a bit "empty"

Share and discuss your LMMS music projects here, and see what people think!
With the spectrometer, if you observe the frequency spectrum of a snare, in general, you will see that it is quite full between 20Hz and 20000Hz, with a peak intensity around 200Hz.

Like the snare, many instruments have a broad spectrum.

On the contrary, if you observe the spectrum of a sine wave, you will only see a peak of intensity corresponding to the fundamental frequency of the note played.

If the majority of instruments have a broad spectrum, it must be understood that when several instruments play at the same time, their spectrum overlaps with each other. In other words, each instrument loses some of its definition. Maybe you have already noticed that when your track is well loaded, when there are many instruments playing at the same time in the same register, sometimes you do not hear your kick, or your pad, or any other instrument.

Filters and EQ can be used in many cases. But here, the filters and EQ are used in this context of spectral overlap.

The idea is to identify in which part of the spectrum such instrument expresses the most, and to reduce (with EQ) or filter frequencies that are not useful or that impede the intelligibility of the track.

Example: place a high-pass filter on your snare around 160 Hz and play your track. Do you hear an unpleasant alteration of your snare drum? If so, decrease the cutoff frequency. If not, increase it a little. As soon as you hear that your snare drum is altered in a way that bothers you, go down a little bit the cutoff frequency. That's it, you removed or very strongly attenuated the frequencies that are below 160Hz. This allows instruments that are below this threshold, such as kick, to occupy this space without being disturbed by your snare.

Now the problem is that in a normal track, you have several instruments, so possibly a lot of overlap. The EQ and the fitres serve in particular for that: to assign a place to your instruments in the mix, to circumscribe each instrument in a determined frequency range. In this context, you do not really perceive, you do not really hear what you do with the EQ and filters. But in the end, by treating each instrument, each instrument group, it makes you gain in definition and headroom.

And all this is valid for instrument layering. For the design of such a pad, you can take the bass frequencies of a synth, the medium frequencies of another, and the high frequencies of a third, and you get a custom pad from these three layers. There is plenty of way to do it.

What I say is far from complete, EQ and filters can do a lot of things, there are many proven and documented techniques. Just try again and again, read, search, gain experience there too.
Regarding filters, I advise you to read or reread https://lmms.io/wiki/index.php?title=Wo ... ts#Filters

Concerning the EQ, I advise you to start with C*Eq or C*Eq2x2 and Multiband EQ. They are complementary and they are easy to use because they allow to use only one parameter: the gain (in dB) of fixed frequency bands (in Hz).

A detail that could be important:
- When you add C*Eq and C*Eq2x2, a notch filter is automatically applied to the signal at +/- 22 kHz (you can't ear that, so it's not really a problem...).
- When you add Multiband EQ, it automatically increases the signal volume by a few dB and it delays the signal by a few ms. The fact that it delays the signal by a few ms may be a problem depending on what you want to do.

Once you get to use these two EQs, you can try the TAP Equalizer/BW. This plugin allows you to manipulate three parameters:
- the gain (in dB) assigned to a frequency band,
- The frequency (in Hz) on which the gain will act,
- The width of the band (in octaves) at the center of which is the chosen frequency, and on which will act the gain.
D.Ipsum wrote:
Sat Nov 25, 2017 12:31 am
noba wrote:
Thu Nov 23, 2017 10:09 pm
I still can't find out how to do what you suggested, and I did look at some of the guide (starting from the beginning and including the EQ page.)
I advise you to use a spectrometer.

Personally, I use EasySSP that you can download here: https://au.tomatl.org/
It's a VSTe, VST effect, that you place in the same folder as your VST.
All VST effects are opened via the FX tab of the instruments or via the Effect Chain in the mixer.

EasySSP contains a goniometer and a spectrometer. The spectrometer allows you to visualize the frequency spectrum of your sound. This is very useful when working with EQ and filter (because these act on the sound spectrum).

I advise you to visualize the spectrum of each of your instruments and the spectrum of your entire track. To get an overview, you can place the spectrogram in the master chanel of the mixer.

Similarly, when you use EQ, filters or any effect acting on the spectrum, I advise you to place a spectrometer after these effects so that you can visualize what you do.


The principle of layering is simple. Consider the TripleOscillator. This synth works according to this principle. Indeed, we can consider that Oscillator1 = Layer1, Oscillator2 = Layer2 and Oscillator3 = Layer3. You only have one instrument, this instrument consists of 3 layers, and when you use these 3 oscillators, they play the same score. The listener does not feel like listening to 3 instruments, but only one.

You can do the same thing with samples. For example, you can choose 2 snare samples and have them play the same score. If done well, the listener will have the impression of listening to one snare, not two.

Why use layers? To enrich the timbre, to fill the spectrum, to get a more complex sound, or for other reasons.


Regarding EQ and filters, there is a lot to say, I do not know where to start, I do not know if I can.
I think the best thing is that you look for information on the net and watch videos, videos tutorials ("how to use EQ", "how to EQ", "EQ filters", "EQ layering", etc. .), etc. Once you have acquired the basics, I might be able to answer more specific and circumscribed questions.

If someone can help ...
That was great! Thanks!
I've tried layering, I'm not really using samples aside from the drums however, but from what I understood I assume you just make a duplicate of the existing synth. So I did that, then panned each one slightly to the left and right. Here's the changed one (I'll probably change up the second synth somehow, but what I should change? I'm not sure yet.)
https://soundcloud.com/jole-challinor/c ... ing-thingy
noba wrote:
Sun Nov 26, 2017 9:32 pm
That was great! Thanks!
I've tried layering, I'm not really using samples aside from the drums however, but from what I understood I assume you just make a duplicate of the existing synth. So I did that, then panned each one slightly to the left and right. Here's the changed one (I'll probably change up the second synth somehow, but what I should change? I'm not sure yet.)
https://soundcloud.com/jole-challinor/c ... ing-thingy
I listened to your last version, it sounds good, I do not know what I could propose.

In fact, there are several ways to use layers (and there are plenty of tutorials on the net). The only limit is that of your imagination.

For example, if you use a snare sample in one of your tracks, and you find that it lacks some noise, rather than trying to crush your snare drum with a ton of effects, you could add filtered or equalized white noise by making sure its volume envelope matches the sample's volume envelope. Then, to better glue your sample and your white noise, you can assign them to the same channel in the mixer, and treat them together with some effects. In the end, what you listen is only one snare, even if you had to use two, three, four or more sources (synth, sample, VSTi, sf2, etc.).

You can try this principle in your next track for your bass line:
- selects two instruments that sound good in the bass,
- make them play the same score,
- assign them to the same channel in the mixer,
- equalize or filters these two basses differently from each other via the FX tab of these instruments,
- use some effects in the channel they share so that they glue together.
- In the end, you only have one bass, even if you used two instruments.

These are just examples I give. I advise you to experience everything you can think of, everything that crosses your mind. With experience, you will see what works and what works less well.
I just listened to the first version, you posted on page one and the latest version.
And after listening to them twice, I can kinda understand, why you're not satisfied with the track.

This one is a toughy. I have no idea what advice to give you, on this one. :P

Maybe a reverbed piano melody, or a steel pan melody, in the back ground might work.
Or maybe, make the piano or steel pan, play the same main melody, in the background
but played at a lower volume.

I'll listen to the track again. And see if I can come up, with another idea. :)
But overall, the track sounds good though. :)