transient troubles..

[stillshaded]

ya'll like alteration right? (edit: alliteration )

Anyways.. How much transient material in a mix is too much? How much is not enough?

Obviously it's a taste situation to some degree, but... using s(m)exoscope, I've noticed that it's often pretty tough to hear the difference in amplitude of transients within a certain range (if you follow), I reckon it could be due to my monitoring setup.

Any thoughts or rules of thumb on this issue?

[mod edit: Several posts by sock puppets claiming to be professional producers have been deleted. A few posts still quote them, and some posters may be referring to their advice in comments that remain.]

[stillshaded]

just found this thread over at gearslutz...

http://www.gearslutz.com/board/masterin ... ering.html


basically it says, don't worry about it. That's the mastering engineers job. Interesting. Makes since really. Kind of a relief. I think I've been worrying about this stuff too much.

[Trakstar]

Get G clip and boost the levels until there are no transients and the levels are all distorted and squashed flat. Listen to the difference between the two examples and I think you will see quite clearly why a mix needs transient and dynamic range to it.

[cruisy18]

Are the transients managed by the mastering engineer with a limiter, or do they usually use other transient plugins?

[Tricky-Loops]

Are the transients managed by the mastering engineer with a limiter, or do they usually use other transient plugins?

I think the transients should be "mastered" before the mastering... You can't do much on the transients if the whole song is finished because you need the single tracks with the transient stuff (particularly beat and percussion elements) to treat them. The mastering engineer is there to make the track sound better AS A WHOLE, not for single elements.

[stillshaded]

Part of what I'm asking is, what methods do you use to determine if there are enough transients in an element? It seems harder to detect than frequency masking. For example I might use a compressor to enhance the transient in a snare and then realize by using smexoscope that it has added an unaudible 3millisecond spike to the begining of the sound.

How do you go about addressing this type of situation?

[quayquay17]

I think you're confusing body and transient.

Compressors tend to smooth out transients, while boosting body (or sustain).

As far as judging how much, you're really looking at loudness. Loudness is dictated by your genre, and your audience.

If you're making music for yourself, do whatever you want. If you're making electro-house/pop you're going to need to compress and limit quite a bit to get the track very loud, to fit sonically with other tracks of the same style.

If you're making jazz, and it's not intended for the radio, the common practice is to have the relative loudness quite low. This doesn't mean turn the volume down, this means have more transients.

A barely compressed jazz song and the last lady gaga song can both peak at 0db but lady gaga will sound MUCH louder, as the genre and audience require much louder songs.

[quayquay17]

As an extra tip:
Get a song by an artist that inspires you that is stylistically very similar.
Put it in your DAW and play it side-by-side with a track you have made. Make sure the peaks match up (the loudest peak in each song hits the same level).
From here, if the reference track sounds louder than yours, consider using more compression and limiting. If yours is louder, consider scaling back.

I'm sure others will be keen to tell you to do everything except use compression and limiting. There are other techniques for "excitement" such as filling out your mix with more elements, using more strategic programming, etc. But at the end of the day, sometimes all a track needs is a good whack through a compressor!

[Aroused by JarJar]

Most stuff you'll hear and read about "dynamics" and "transients" and all that in "professional mixes" is just lip service, aka horseshit.

Spend a lot of time in the acoustic world- straight documentary-type recording- and you'll realize that pop music has almost *no* dynamic activity in the way it appears in the acoustic world.

Get a decent room and put your mike a fair distance away, and just hit some things- a saucepan, even some pieces of wood. The spectral range, the transients and the dynamic range are enormous. When digital recording was first introduced, it wasn't about just about getting headroom (via high sample rates) to capture this huge range, it was about getting "footroom", which is bit depth (detail and range of amplitude changes) and a low noise floor. So the point was to have far lower average sound level, then simply turn the whole thing up. Voila, loud as hell and you have the steep transients. Precious few have the listening environment and listening habits to experience these kinds of recordings properly.

Pop music creates an illusion of dynamics first and foremost with arrangement. Listen critically to "loud" music, EDM, modern hip-hop, whatever. It's not like an ochestra (which be so loud without amplification that conductors can get hearing damage like rock and rollers do) at all. In an orchestra masses of air get moving not just by blasting on horns fortissimo, but even in quiet passages, because it's many sounds blending together. The loudness really is "volume"; it is mostly "fullness".

Now listen to a modern pop production. There is very little "blending together" at all. It is to a large degree basically discrete sounds near 0dBfs (maxed in digital) alternating with one another. Even when sounds have theoretically the same point of attack, in the score so to speak, they're sidechained so that the peak of one follows the peak of the other. Instead of "fullness", there is a rapid succession of "blasts". Where there are different sounds happening at the same time, they are almost always far out of each other's way.

And of course everything is, compared to the acoustic world, absolutely smashed flat with compression, maintaining RMS wherever the peaks are not happening.

If you're doing EDM or some other "loud" pop style, the answer is simple: first and formost get everything out of everything else's way. That's basically the nature of pop arrangement today.

But what if you are not recording an orchestra, which always sounds loud and punchy whenever it wants even if your peaks and RMS are way down in the recording, and you're not recording a solo acoustic instrument (which can be made louder than *anything* especially if it's a high thin sound), and you're not doing one of the styles in which the compositional approach itself has evolved around "loudness", like EDM?

What if you want sharp transients and dynamics but you don't want your recording to be appreciated only by those with Hi-Fi systems or the exceptionally high IQ required to turn a volume knob "up" or "down"? Now that is hard, but it can be done by examining how things work in the other approaches.

[stillshaded]

If you're doing EDM or some other "loud" pop style, the answer is simple: first and formost get everything out of everything else's way. That's basically the nature of pop arrangement today.

But what if you are not recording an orchestra, which always sounds loud and punchy whenever it wants even if your peaks and RMS are way down in the recording, and you're not recording a solo acoustic instrument (which can be made louder than *anything* especially if it's a high thin sound), and you're not doing one of the styles in which the compositional approach itself has evolved around "loudness", like EDM?

What if you want sharp transients and dynamics but you don't want your recording to be appreciated only by those with Hi-Fi systems or the exceptionally high IQ required to turn a volume knob "up" or "down"? Now that is hard, but it can be done by examining how things work in the other approaches.

that last paragraph basically sums up what I'm attempting to do. I want a kind of middle way. Not really after super loud mixes, but I would like them to sound decent on mediocre stereo's. So I'm after compromise I reckon. I create strange electronic music generally. experimental stuff. aphex twin is a hero.

Thanks again for the thoughtful reply jarjar. You've got a cool attitude, I appreciate it.

[quayquay17]

Worst advice I have heard in a while my friend. You clearly don't know the implications and the nature of limiting and compression, specially when you try to imitate heavily squashed material.

Thanks for putting me on blast my friend, but contrary to your hi-fi teachings, there is no way you can make a 2012 pop-dance hit single without some compression. Name one track that reaches today's standards of loudness that doesn't rely on gain reduction to get there.

I'm not saying it is the best thing for those poor transients, I'm saying this: the intended audience of your music will often dictate the loudness. Is it an inherently good thing that pop is so bloody loud? I never said that. But if you want your track to sit next to those without sounding quiet or out of place, you're going to have to do some of that nasty nasty GR

And to the comment on mastering and working retroactively to bring back more dynamic range, I can certainly sympathize, the last mix I finalized for someone was incredibly squashed in the low end and gave the track no punch. Certainly brickwalling and straight compressing everything is NOT the answer to a powerful mix, but for certain genres/sounds/situations more loudness is required than what you would get from arrangement work.

EDIT: maybe I should have clarified my first point, I didn't imply adding more and more compression and limiting at the master, rather, increasing the loudness of your tracks/busses.

[annode]

I don't have time to read all the replies, so excuse me if i'm repeating what was said.
Some may disagree, but I see people confusing a transient with an attack peak level.
This differentiation is import to realize when your working with limiters and compressors when dealing with peak levels.
First off, the attack is the time it takes for the sound to reach it's highest level. In respect to this thread that time would be very quick and mechanical in nature...such as a drum hit or bringing any acoustic sound, (or oscillator) into resonance for that matter.
The transient(s) are what remnants of this mechanical excitement left sounding after the excitement has ceased. The transient(s) are after the peak...are short lived, (fall off quickly) and inharmonious to the resonant tone of the instrument.
The combinations of the attack, transient(s) and tonal resonance give the instrument it's particular sound characteristic called timbre. Without these characteristics the sound would begin to lose...well...it's character.

Attenuating very fast attacks with comps and limiters not designed to respond to such quick attacks will generate distortion and other shit. If your plug can't 'release' itself quickly enough, you'll lose those transients.

So in my opinion, clamp down on fast peaks with a properly designed comp/lim able to respond to these speeds without breaking up.

[Aroused by JarJar]

If you're doing EDM or some other "loud" pop style, the answer is simple: first and formost get everything out of everything else's way. That's basically the nature of pop arrangement today.

But what if you are not recording an orchestra, which always sounds loud and punchy whenever it wants even if your peaks and RMS are way down in the recording, and you're not recording a solo acoustic instrument (which can be made louder than *anything* especially if it's a high thin sound), and you're not doing one of the styles in which the compositional approach itself has evolved around "loudness", like EDM?

What if you want sharp transients and dynamics but you don't want your recording to be appreciated only by those with Hi-Fi systems or the exceptionally high IQ required to turn a volume knob "up" or "down"? Now that is hard, but it can be done by examining how things work in the other approaches.

that last paragraph basically sums up what I'm attempting to do. I want a kind of middle way. Not really after super loud mixes, but I would like them to sound decent on mediocre stereo's. So I'm after compromise I reckon. I create strange electronic music generally. experimental stuff. aphex twin is a hero.

Thanks again for the thoughtful reply jarjar. You've got a cool attitude, I appreciate it.

Hey thanks, I'm glad to know not everybody's doing EDM or Dubstep, hehe. I've got some "things I've found" about getting levels appropriate to music played in a changer (where you will notice most if you're too loud or too quiet) but I don't have the time at the moment to make anything more than brief comments, so I'll get to it in the next couple of days.

[Aroused by JarJar]

One thing I've found is that if you are doing your own synthesis and sound design, you can "pre-produce" in the sounds themselves, and so avoid a lot of having to dick around with stuff later.

In other words, you can get sounds out of each other's way, or blending together if that's what you're after, not just in orchestration, but in the envelopes and spectra of the instruments. This is exactly what people do when they layer killer kicks, using the attack from one, the depth from another, the higher tones from yet another, etc., but it can be done with all the sounds.

For example, let's say you have snappy kick and a snappy plucked-type bass both with rapid attacks and sudden decays. Each on their own sounds great. When the kick is on a beat and the bass on an offbeat, it sounds great. But when both fall on the same beat, you've got a massive amplitude spike, an attack transient that clips unless you turn everything way down. You could smash it down with a compressor or limiter. Or you could use a side-chained compressor, so the bass attack, is pushed down when the kick attack is at its max. Effectively, this makes the peak of the bass attack come in immediately after the kick attack. There's also some kind of pumping feeling in this technique, it's great for certain kinds of music.

But if you don't want this pumping, you can program your bass envelope to be "pre-sidechained". Using an exponential rise, or even better a custom envelope, you can simply make a longer attack on the bass. So, the amplitude is very low during the intial smack of the kick, then it leaps up to peak immediately after the kick "click". And the bass envelope can then hold (or drop or rise a tiny bit) for some milliseconds. So the bass will be pre-compressed and pre-sidechained, with the added benefit that you can put filter modulation, FM modulations, etc. where you want them instead of getting them nice and then doing extra work trying to get the compressor to rewrite the envelope while not eliminating the peak of resonance or whatever.

A Fender P-bass, and other electric basses played by expert players, do this kind of thing "automagically" by the way- a round body of sound "floats out right after" so to speak, so you can have an extremely tight rhythm section that is full sounding and not spike! empty... spike! empty...