How To "Dirty" A Mix?

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
RELATED
PRODUCTS

Post

If the criticism suggests the sounds are perceived as "general MIDI" then I would assume a lack of movement and variation within the sounds themselves. Though, to some extent, this can be achieved in a mix with some envelope followers controlling EQ parameters.

However, when I think of dirtying up a mix, I would immediately consider parallel distortion. With regard to 80s sounds though, I would consider the restrictions of the gear at the time: Lower bit depths and sample rates; Narrower band widths.

Post

The easy way is to just use some amp and tape sims; and not simple waveshapers, mind you: you want something that will react differently to changing input levels. Waveshapers just add the same type of distortion to everything, whereas real amps and tape respond: low levels will add a smaller amount of lower-order harmonics, but higher levels generate greater amounts and higher-order of harmonics. It's a bit subtle, but even the sonic wallpaper consumers with their overcompressed, low-bitrate mp3s and earbuds will tend to notice.

But there is a better way, assuming the style of music you're making calls for funky, low-tech grime. It's heretical, but the world needs more blasphemy. :hihi: First, de-quantize things. (Better yet, record non-electronic instruments, but that's probably just my prejudices showing. I like things that respond to touch and volume and such.) Next, record the mix to tape. Yes, tape. Magnetic, that is. Play that tape back and record to a second generation of tape. Repeat this a few times. This ensures a pleasing dirtiness and warmth. Record that to digital. Release. Hey, if it's good enough for Motown and Portishead, it's good enough for you! (Again, assuming it's appropriate for the style; for a lot of great music, keeping things a bit clinical is completely the right choice.)
Wait... loot _then_ burn? D'oh!

Post

Jafo wrote:The easy way is to just use some amp and tape sims; and not simple waveshapers, mind you: you want something that will react differently to changing input levels. Waveshapers just add the same type of distortion to everything, whereas real amps and tape respond: low levels will add a smaller amount of lower-order harmonics, but higher levels generate greater amounts and higher-order of harmonics. It's a bit subtle, but even the sonic wallpaper consumers with their overcompressed, low-bitrate mp3s and earbuds will tend to notice.

But there is a better way, assuming the style of music you're making calls for funky, low-tech grime. It's heretical, but the world needs more blasphemy. :hihi: First, de-quantize things. (Better yet, record non-electronic instruments, but that's probably just my prejudices showing. I like things that respond to touch and volume and such.) Next, record the mix to tape. Yes, tape. Magnetic, that is. Play that tape back and record to a second generation of tape. Repeat this a few times. This ensures a pleasing dirtiness and warmth. Record that to digital. Release. Hey, if it's good enough for Motown and Portishead, it's good enough for you! (Again, assuming it's appropriate for the style; for a lot of great music, keeping things a bit clinical is completely the right choice.)
Thanks, but all my analog gear, including my Teac A3440 reel to reel, is long gone so all that's left is my PC. It'll have to do, unfortunately.

But you should hear my early stuff. Talk about dirt.

Post

My guess is:
  1. The reverb is all wrong. You need to damp this off quite a bit. You could try applying some EQ to the wet signal. Note the sort of reverb most often used in the 80s was really awful spring, plate or digital reverb. It's just too clean as it is and this was the first thing that really popped out at me.
  2. The hats are too bright/clear and regular. Get some more low/mid range timbre in there like you'd get from a real hi-hat. These just sound way too high-pass/synthetic/sampled. Variation in the pedal position is also great. I absolutely love riding the pedal on odd/spare 8ths rather than closing the hat. This gives an effect similar to rivets in a cymbal but not the same. A little variation goes a long way, here it just sounds like someone went nuts on a drum machine. What I'm hearing here is way too regular. If there were any possibility this were a take of a real kit I'd have to guess each instrument was sampled one at a time and only for a single bar! It sounds very sequenced and repetitive because of this which makes it "clean".
  3. Other percussion elements (kick/snare) don't sound much like a real mic'd set, the whole set doesn't really sit well together as if these are independent samples. If you have anyone who could really play a bar of your percs pattern it might help a lot to get the kit to sit together better. Some compression immediately applied to the sounds including a good drum-room reverb impulse might give you a little bit of realism timbre-wise, although it wouldn't help the performance aspects at all which I think are far more important.
  4. The bass is actually quite free of any bass. Likewise there are no thumps or low-frequency components present in any other parts. This makes it feel like something is missing and likely it is intentionally so: as if you've taken some soap and a brush and scrubbed the nether regions (AKA the balls) right off. They're missing.
In general timbre-wise a little high-cut can get you quite far to add body to any part with minimal effort. The essential thing is not to go wild with complex EQ but just to apply very simple and light filters like a plain 6 dB low-pass at ~16 kHz. If you do that you should find that it sounds very heavily "band-passed" due to many other elements (guitars, bass, kick) lacking their low end components. Most importantly the filter can't be too strong otherwise you get instant mud. It's important to keep in mind that this isn't going to help to replace any elements that simply aren't present at all. It may work to tame the extra-bright hi-hats but won't likely help fill in the missing mid/low components as much as I think may be needed.

I think the reverb is what causes the biggest issue though by far. It's really "in-your-face" extra bright and smooth - AKA clean. Since it's applied to the percussion which runs non-stop throughout the track it colors the whole track. Like having a lightly flavored garden salad and smothering it in a thick layer of thousand island dressing!

I've had a go at post-processing dirty-fication which is something I managed to apply reasonably well. I'll link that in PM along with a reaper FX chain.
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

The bass harmonics are really clean. Not only that, the bass sounds sequenced (obviously). There's no movement to the timbre of the bass and also no dynamic to the drums, the drums always hit at the same strength. These aren't bad qualities, but they are uninteresting.

Post

Real tape, tubes, transformers, discrete class a circuits etc. All can add a nice something without sounding obviously distorted.

Post

I should've mentioned in my post that I do like the effect that reverb has on the vocal elements. Just not on the percussion or bass. For the vocal elements since they are vocoder-like effects it makes sense to have this fizzy/smooth tailed reverb applied there because the sounds are so synthetic the synthetic accent is perfect.

On the percussion and bass though it doesn't help to add realism. It rather makes sampled lines that could potentially sound semi-realistic or in the least not be a distracting draw of attention sound awfully synthetic. Especially in combination with what seem to me to be unnatural spectral characteristics of those parts.
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

aciddose wrote:The hats are too bright/clear and regular. Get some more low/mid range timbre in there like you'd get from a real hi-hat. These just sound way too high-pass/synthetic/sampled. Variation in the pedal position is also great. I absolutely love riding the pedal on odd/spare 8ths rather than closing the hat. This gives an effect similar to rivets in a cymbal but not the same. A little variation goes a long way, here it just sounds like someone went nuts on a drum machine. What I'm hearing here is way too regular. If there were any possibility this were a take of a real kit I'd have to guess each instrument was sampled one at a time and only for a single bar! It sounds very sequenced and repetitive because of this which makes it "clean".
The hihats also immediately jumped out at me as sounding too hi-fi. I partially agree with the above, as I'm a big fan of detailed drum programming - when trying to emulate a real drummer that is. I think in this case, there's an argument to be made for the rhythmic sequenced feel if that's what wagtunes is trying to emulate.

The other thing I noticed was a really perceptible change in the volume and tone of the hihats whenever the voice comes in at a busy section. I'm assuming this is some sort of compression, either sidechained to the hats or a multiband compressor on the drum bus? Not being the world's greatest mix artist myself, I can't be sure, but there's something weird going on there.

Post Reply

Return to “Production Techniques”