Applying compression to a sound several times

Question

I would like to know what happens, objectively speaking, when you apply compression to a sound 2, 3, or 4 times. I would beleive its useless to do this. Or counterproductive. I generally use a compression to process a signal in real time. Then I normalize again or compresss (or normalize and compress) the resulting audio. I do this in order to add gain to the sound, but still don't about the quality loose. What would be your suggestion for enhancing an entire audio signal, besides compression, or changing the ssd card of the computer, or besides mastering?

Thanks

Answer 1

It can actually be a very good idea to put two or more compressors (with different parameter settings) in sequence. For instance, limiting¹ can allow you to make drums louder, but it tends to make transients a bit murky. One way to counteract this is to put a slower compressor (possibly with sidechain low-cut) before the limiter, to “pump” these transients.

In particular, it's very effective to cascade multiple stages of compressors / nonlinears with intermittent EQs. There are countless possibilities to this, it depends on what you're trying to achieve.

As for quality loss – modern oversampled digital compressors really don't have this, unless you set them up wrongly. And multiple gentle compression stages will generally preserve more of the sound than a single compressor with the same total gain-reduction scope.

---

¹_{I rather dislike limiters as such, prefer soft-clipping (e.g. by tube / tape simulations).}

Answer 2

Pretty much everything you hear in popular music has been through more than one stage of compression. During mixing, it's pretty common for vocals to be run through two compression stages, with one being a low ratio, longer time constant compression to smooth things out, and the other being a higher ratio, shorter attack and release compressor (or a limiter - the line between the two is blurry) to reign in peaks. Vocals are usually highly dynamic sound sources. And it's common in mastering for a compressor (maybe variable mu or some other "glue" type compressor) and a limiter to be used, gently in both cases.

There are two aspects of compression that affect the quality of the results: the algorithm and the settings. While some algorithms have advantages and disadvantages in certain situations, as leftaroundabout says, it's more likely the settings that are to blame if something sounds worse after compression. Learning to use compression in a way that makes things sound better instead of worse can take years.

Regarding your exact situation, I suggest you stop normalizing. That's rarely a process that makes things better. I'm not sure what you mean by changing your SSD card, but it doesn't sound like anything that would affect the quality of your audio at all either way. Having been in your shoes I recommend looking for a look-ahead peak limiter plug-in. Waves L1, L2, or L3, Izotope Ozone, and Massey 2007 are popular examples. These will automatically help you increase the volume in the correct way, which is by increasing the input gain on a compression/limiting stage, not by normalizing.

Answer 3

I believe the compression ratios at each stage of compression multiply. So applying 3:1 compression twice results in a 9:1 ratio. Adding multiple stages of compression is common when mastering tracks though it's generally recommended that compression ratios are kept to a minimum at each stage. Sometimes compression can be used to increase the volume of the sound once it crosses the threshold (I think this is called expansion) instead of decrease the volume. I've used multiply stages of this effect from time to time to turn tiny peaks in the waveform into huge peaks.

Compression doesn't really have an effect on audio quality as far as I'm aware. It can be used to remove dynamic range and make every second of the sound as loud as possible, which should be frowned upon. If you want to increase sound quality in your stereo, get a better sound card or look into other effects to make the sound sound better.