recording vibraphone

[goose134]

I'm curious as to how those transients manifest themselves. Did it ring out too much? Get cloudy? I totally respect your opinion but it seems counter to virtually everything I've read. Just trying to educate myself here.

[Nick Sevilla]

I'm curious as to how those transients manifest themselves. Did it ring out too much? Get cloudy? I totally respect your opinion but it seems counter to virtually everything I've read. Just trying to educate myself here.

Hi goose,

Transients don't "ring out", they are very fast in the vibraphone because you are striking a metal object, and those tend to create very fast spikes of sound, followed by a much less loud ringing. A lot of inexperienced engineers mistakenly try to make the tail of the sound louder, and end up overloading their signal with the huge transient.

Transients also don't "get cloudy" at least in my book, cloudy means undefined, foggy, smeared sound. Usually associated with a longer waveform, like on a synth pad, or crappily recorded cymbals.

Usually, vibraphones, claves, cymbals, tambourines, bells, crotal bells, timbales, certain highly tuned congas, bongos, drums, et al, all are instruments that give out much louder transients, compared with their sound decays.

I hear distorted transients in all these type of instruments ALL THE TIME. Typically because of this problem I mentioned of the engineer simply not listening to the transient, and believing they need to make the decay louder than it can be.

Many modern converters simply cannot indicate these transients because their LED or other signal meters are too slow even for those. Usually percussive transients can be as short as a few picoseconds, going to as slow as half a millisecond, in the more soft materials. such as the woods in some marimbas for example. But the problem is that they are VERY LOUD and extremely FAST, and mostly, in the course of a modern recording session where everything needs to be recorded ASAP, they do not get the proper attention.

On these signals, I typically only allow the LED indicators on whatever converter I am using to go to about half scale. Then on inspection of the actual recorded waveform, I sometimes find transients in reality that get to -4dBFS or so, even though the meters did not get that high. Because they are too slow.

Now, why the super long dissertation on these transients? Because in a mix, they acre CRUCIAL in determining the beat of the rhythm, and also they are important in distinguishing percussion instruments from others such as electric guitar, which also can have very fast transients, if properly recorded.

Now, IF you have properly recorded a vibraphone or other percussive instrument, you can always turn it up, without causing unwanted distortion. They way I typically do this, is by using one adjustable limiter to reduce the transient by 4-6 dB, and a second limiter/compressor to "grab" the decay, and alter its shape, by grabbing it, without touching the transient, and giving more gain makeup than compression, in essence making the entire decay louder. This method in the end, reduces the original dynamic range of the recorded instrument, and can easily be adjusted to fit any mix whatsoever, keeping the percussion instrument distinct and easily distinguishable in a dense mix.

I hope this clears up my way of dealing with percussive instruments like the vibraphone.

Cheers

[Nick Sevilla]

Now,
As to address an improperly recorded fast transient...

Depending on the microphone and preamplifier being used, your transient will not look nor sound like a nice sharp sine wave modified with very little internal waveforms depending on the instrument.

A badly recorded transient can look like a small series of waves at the beginning of the sound, followed by a GIANT huge wave, and sometimes the top of this large wave may be clipped entirely, if not, it is usually very long, sometimes 4-5 times longer than the actual transient, and then a longer decay is seen, also of long waveforms.

This indicates an overload either in the mic circuit and/or the preamplifier.

That sort of waveform is the op amp overloading, and the resulting overload giving you that shape.

A well recorded fast transient starts with a very small amount of very small waveforms, which can be the electronics of the microphone reacting to the air coming from the instrument , follows very very quickly by the actual strike, which is huge, and then the decay.

If you can picture a wave of water, you can picture that there is a rise in the undisturbed water in front of the wave, and then at a certain threshold, the actual wave can be said to begin. This happens the same in air, except in a much shorter (faster) scale. A good signal chain will allow you to capture all of that air disturbance accurately, while an inferior signal chain will just poop all over that in sometimes unpredictable ways.

Thus why it is important to have a very good pair of microphones and mic preamps and converters that are well designed for these kinds of instruments.

One example of a mic preamp that typically cannot deal with super fast transients, is the Neve mic preamp, the classic design from the late 60s and early 70s. Because of the sheer amount of components, you are almost guaranteed a "smear" of transients. And sometimes this sound is actually desirable. Like in a more sparse arrangement, where you want softer transients. If you want super accurate transient recordings, you should look to solid state mics, like AKG 414s, 451s, Neumann KM 84 series, and stuff like that. As to mic preamps, you want something like the John Hardy M1, based on the API type of circuit, but much better designed.
Never forget to alw the meters to go only about half scale, so you do not let the transients to get screwed up by the electronics.

And if you are going to real tape, don't let the VU meter get much past -20dB, as that will indicate the right level to tape of the transient. The tape recording will still smear and distort the transient, but if done right, it will still sound good. In fact this was one of the nightmares of us engineers who started on tape... trying to get an accurate transient recorded. JEEBUS!!!

Cheers

[MoreSpaceEcho]

lots of good stuff in those last two posts.

[drumsound]

lots of good stuff in those last two posts.

I agree. Really great stuff from Nick, as usual.

[Nick Sevilla]

Thanks guys.
I edited to spellcheck.
Writing in the morning on an iPad2 is horrible.

Cheers,and Happy Labor Day!

[goose134]

Wow, that was pretty great. I kind of pictured you looking over a pair of reading glasses while delivering it. Tons of great stuff that is way out my league of recording prowess. Still, it's one of the things I love dearly about this forum: picking the brains of those much smarter than I and much better at their craft. It'll be something I'll be listening for when I try this out. Thanks again, and Happy Labor Day!

[L?Andratt]

Wow.
Totally in awe.
I?m not going to record vibraphone and I?m just a homerecording guy with shitty gear, but always looking for brainfood and getting it delivered here for free.

Thanks a lot. Tastes very good!

http://youtu.be/usEkQ2UjfS8

[biasvoltage]

The only time I recorded vibes, I used a pair of SDCs (Octava MC012's) in a coincident pair right over the middle of the thing, about 3' up.

Sounded great, with a nice stereo spread. However, also caught the wheezing and sniffles of the player, who had a cold at the time. So maybe add some Claritin to the list of equipment needed for this.

[ott0bot]

poo poo transients

[drumsound]

The only time I recorded vibes, I used a pair of SDCs (Octava MC012's) in a coincident pair right over the middle of the thing, about 3' up.

Sounded great, with a nice stereo spread. However, also caught the wheezing and sniffles of the player, who had a cold at the time. So maybe add some Claritin to the list of equipment needed for this.

If you want to save some money, the generic Claritin, with the Michael Joly mod works wonders.