What's new

Wet libraries

Syneast

Active Member
I love wet libraries.

I have learned that putting ER and tail reverbs on a dry recording or even a sound stage recording will not make it sound like it was recorded in a large hall or a church. A large space like a symphony hall does more to the sound than simply add reverbation. Without getting into the science of it, a large space seems to smooth out harsh frequencies that a smaller space would instead accentuate.

Don't get me wrong, I love a good studio or sound stage sound just as much as the next guy, but for some things it's just not appropriate. For this reason I invite you to list your favorite wet (orchestral) libraries with long tails baked into the recordings. I'll start with the obvious ones that I know of:

- East West Quantum Leap Symphonic Orchestra
- Anything by Spitfire Audio that's recorded at Air Lyndhurst
- Orchestral Tools Berlin Series
- Native Instruments Symphony Series Brass

P.S. I know some libraries can have a sort of distant sound even if they are recorded in a dry space, like Red Room Audio Palette series for instance, but this post is all about the libraries recorded in big spaces with tails around 2 seconds or longer.
 
Last edited:
I'm not a big fan of wet libraries in general, but I do love the CS2 hall.

I also really like the ewqlso hall.

I know the Lyndhurst hall is very popular, but I've never been able to jump on that bandwagon. In fact, i own a lot of Spitfire stuff, but am instantly reminded of that large hall sound as soon as a play one of their samples, so i usually avoid it.
 
Without getting into the science of it, a large space seems to smooth out harsh frequencies that a smaller space would instead accentuate.

This can happen when instruments are recorded from a mic position farther back than the "critical distance," which is the distance beyond which the reverberant sound is louder than the direct sound of the instrument reaching the microphone. Beyond that point, the sound of the space dominates over the sound of the instrument itself. Of course, it is a continuum from near to far so it's not just like the sound instantly switches once you put the mics a certain distance back.

If there are "harsh frequencies" in a dry/direct sample, it's because those frequencies really exist in the instrument itself. Depending on mic technique and equipment chosen they may get boosted even more. But a nice space can hide flaws in recording, playing, or equipment, especially beyond the critical distance, because the space itself becomes the dominant sound.

This can actually be emulated in mixing, though, too: take a dry sample, and instead of trying to add reverb by the usual method of sending to a post-fader effect, insert the reverb on the channel instead, and use the reverb's Dry/Wet control. 50% Wet is like the sound of a recording being made at precisely the critical distance, and anything above 50% is like moving the mics back even farther, allowing the room sound of the reverb to dominate what you hear. Not possible with an aux buss unless you're boosting that aux buss beyond 0db.

With good recordings and really good reverbs (I would argue IR reverbs of real spaces are better for this) it should have the same effect of "smoothing things over" that you like.
 
With good recordings and really good reverbs (I would argue IR reverbs of real spaces are better for this) it should have the same effect of "smoothing things over" that you like.

That is rubbish, at least if you're trying to create an orchestral concert-listening experience. [edit: a recorded orchestral concert / listening / soundtrack experience]

The sound of any instrument recorded from further away just isn't the same signature as one recorded close-in. You can put whatever reverb on it you like but it will never sound the same as an instrument (or orchestra) recorded at a normal concert listening distance.

For what -- 20 years? -- I've listened to demos and mixes from people who have tried to coax a nice "concert orchestra" sound out of close-mic'd orchestral samples. It never sounds the same, and often, alas, sounds tortured.

It's completely different if you're trying to make a pop song. In a studio setting, those instruments always are recorded pretty close, actually often with the mic practically touching the speaker on the amp. So that does work fine.
 
Last edited:
. . . The sound of any instrument recorded from further away just isn't the same signature as one recorded close-in. You can put whatever reverb on it you like but it will never sound the same as an instrument (or orchestra) recorded at a normal concert listening distance . . .
Ditto! Been saying this for years regarding sampled pianos.
 
Every library should have a demo on the website of the developper, showing how it sounds with the driest mic position all by itself.

Libraries that are too wet are completely unusable, and are my first source of frustration.
 
That is rubbish, at least if you're trying to create an orchestral concert-listening experience.
I think he's saying that IR/ convolution reverbs work better than algorithmic reverbs for trying to emulate a real concert space but I don't think he's necessarily saying it's better or the same as capturing the real space with a mic. Personally I think the LASS demos show that you can create nice space with a dry recording but maybe not quite a concert space? I take it you don't like dry recorded strings.
 
That is rubbish, at least if you're trying to create an orchestral concert-listening experience.

The differences are largely quantifiable though. Whether anyone does a good job of applying those differences to a mix is debatable (I agree it rarely sounds "right" in the context of sample libraries) but they are still quantifiable.

It is easier to have more success coaxing a reasonably-convincing space onto an actual performance recording that is miked close, than to a sample library that is. I think this speaks to weaknesses in consistency and transitions in the current state of the sample library world more than anything. Yes, a room covers these weaknesses somewhat. But it is just a covering.

The other thing that happens at distance, other than the room sound dominating the tonal profile, is air absorption. High frequencies fall off a cliff over distance (depending on altitude, temperature, and humidity). So you also have to apply a pretty severe EQ profile that has a similar effect. But air absorption is has been precisely modelled, and again, is quantifiable. Again, this is really just removing information, which covers over weaknesses... whether done by processing or by air itself in a real hall.

If directional microphones are used, there is also proximity effect to contend with. Thankfully with omnidirectional mics typically used in orchestral recording this isn't a factor.

You can put whatever reverb on it you like but it will never sound the same as an instrument (or orchestra) recorded at a normal concert listening distance.

Orchestras are almost never recorded from a "normal concert listening distance," but rather much closer (typically just behind the conductor for the main mics). Mics placed in the audience are usually well beyond critical distance and waaaay too wet to use in normal commercial releases.

Our brains are quite good at honing in at an actual live performance, aided by the visual cues of the performers playing, and also by acoustic cues in subtle shifts of our heads etc -- we can actually sort of "process out" a lot of wetness when we're actually in a space, but a recording doesn't have the same benefit.

Anyway, on the whole I find it at least possible to coax something usable out of a too-dry sample than a too-wet one. Much easier to go in that direction... and many libraries (like Spitfire's Air-recorded libraries) are just so wet that they really only work well in a single context (a washy concert hall mix). Just my take on things.
 
I prefer my library to be dry. Water and books just don't go together. At all.

Sorry, couldn't help myself. I really love my SSO libraries. But you do have to mix the mic a bit so it sounds not too close but not too far. But it just sounds good without a lot of fussing with reverbs. Not to say I don't like a good reverb. Hmmm. Wonder what Blackhole and ambient mics would sound like... Getting some ideas.
 
Orchestras are almost never recorded from a "normal concert listening distance," but rather much closer (typically just behind the conductor for the main mics).

That also is rubbish.

They are recorded from many different distances and in many different settings. The recordings from Royce Hall, or from the Conzertgebouw in Amsterdam, sound quite different from what you hear on a sound stage, and mic positioning varies, sometimes radically, from engineer to engineer and producer to producer.

I'm not just being mean, or contrary -- I've had orchestras record in churches, concert halls, and recording studios. Sometimes they sound great, sometimes not, but there is enormous variety.

One thing for sure, we do not just use the mics over the conductor's head, or indeed even all that close to him.
 
Yes, a sample recorded in space sounds better than a dry one with EQ and early reflections and rev tail and whatever. But only one sample.
The problem is that wet libraries are not as playable as dry ones. An added real recorded release tail crossfaded at the end of each sample note doesn't behave at all like a real musical performance in a room/hall. The benefit of the wet recording goes into reverse the more variation happens in the performance. No problem with only one shot shorts or slow motion big lines. But legatos cut the rev tail of previous note and jumps from legatos to short notes or vice versa in the same musical phrase are always a problem. Really playable virtual instruments like Samplemodeling, Audiomodeling, WIVI etc. can't be created with wet samples but they literally outperform all wet libraries by far.
So there's always one death to die. You have to decide between music and sound. I get mad when I have a musical idea in mind and can't perform it. Happens to me with (i.e.) Spitfire stuff every time... oh, that sounds beautiful, lets play... and a frustrating half an hour later the Spitfire stuff went out of the template. Back to music and EQ/early reflections/tail...
 
Yes, a sample recorded in space sounds better than a dry one with EQ and early reflections and rev tail and whatever. But only one sample.
The problem is that wet libraries are not as playable as dry ones. An added real recorded release tail crossfaded at the end of each sample note doesn't behave at all like a real musical performance in a room/hall. The benefit of the wet recording goes into reverse the more variation happens in the performance. No problem with only one shot shorts or slow motion big lines. But legatos cut the rev tail of previous note and jumps from legatos to short notes or vice versa in the same musical phrase are always a problem. Really playable virtual instruments like Samplemodeling, Audiomodeling, WIVI etc. can't be created with wet samples but they literally outperform all wet libraries by far.
So there's always one death to die. You have to decide between music and sound. I get mad when I have a musical idea in mind and can't perform it. Happens to me with (i.e.) Spitfire stuff every time... oh, that sounds beautiful, lets play... and a frustrating half an hour later the Spitfire stuff went out of the template. Back to music and EQ/early reflections/tail...

Well said!

Also want to point out that mathematically, convolution with an impulse responses is a precise way to characterize a system. The problem is that this "system" must also hard code the frequency response, position, and directivity pattern of the source speaker and the microphones. The virtual instrument being placed in the reverb is almost guaranteed not to match the source speaker's directivity pattern or position, and that's the underlying reason for all the problems people are talking about here.

There is a solution to this but it will probably take time for any company to have the resources to do it: instead of dry sample instruments in stereo, the actual 3d sound wave in all directions need to be captured. This can be done with high order ambisonics. and then they will need to develop convolution reverb software that can apply special impulse responses to that ambisonics audio. Realistically, it requires a lot of channels and a lot of convolutions per instrument. So I wouldn't expect this to happen any time soon.

So meanwhile yes it seems like a trade-off between the problems of simulating reverb versus the problems of tweaking performances with baked in room..
 
I think that ambience in samples is incredibly overrated. It's far, far, far more important that plausible and musical performances can be coaxed out of a virtual instrument. Many very "wet" libraries are severely lacking in this department.

On the other side, even a completely dry sample can be made to sound "good enough" with the right reverberation technique.

IMO, a musically interesting and believable VST performance in a "good enough" virtual space will sound much less jarring and problematic than a production with a really reach, "real" sounding baked in space, but poor performance that sounds fake and dilletantish no matter how much much deplorable struggle you put up with because the samples and their scripting is so unwieldy. I believe that if the music itself is good, the programming is sound and diligent and production is good, the musical interest will take over and nobody's gonna waste a thought on the concept of "big space".

I don't even know why everyone obsessses so much over the big, grand, wide, washy tails. It only makes sense for certain types of music and production aesthetics anyway. I think it often sounds ridiculous. In sampled re-productions of the supposedly universally desirable "Hollywood" sound ideal even more so than the real deal itself (which, quite honestly, often itself sounds kitschy and cheap).

I'm not in militant "bone dry" camp exclusively, and do like libraries with a good sense of real space, but which are still "dry" and tight enough. Like CSS. Huge reverb wash in samples is mostly nonsense. I do use some what you would call "wet" libraries as well - but rarely on their own. If the music isn't simplistic or background noodle, it more often than not falls apart with just the wet stuff. The important musical content is always bolstered with something more dry, malleable, reliable. The wet stuff is more for dimension and "sheen".
 
Jimmy - that’s demonstrably false by just listening to one or two of Andy Blaney’s demos for the Symphonic series on our website.

I’m not sure where this idea comes from that you can’t program musical sounding music with so called “wet” libraries!

Although I dislike that division. One significant difference when you are recording a musician is how they react to the space they are playing in. I could write pages on all this but I’m sure I shouldn’t!

Anyway - just dipped into this thread as it looked interesting.

Happy Sunday all!

Paul
 
Interesting stuff in this thread.
I really like the room sound from my wet libraries, especially Metropolis Ark (therefore Berlin Series / Teldex). Excluding noise. It just feels right to combine MA1 and MA2 if I want a big dynamic range. Same stage, same ambience.
If I want them to sound a bit dryer and put my own reverb on, besides using closes mics (not always!), cutting the release time (to about 500ms) can be usefull, even though it's not exactly the same as a dry library, as reverb of course isn't only in the tail of a sample.
 
Last edited:
I'd like to point out that people shouldn't confuse "wet" with "long tail". For example, Spitfire Symphobic Brass and CSB sound like sad, emptying balloons on dry mics alone, but clearly there is a night and day difference with the room mic sound. The room of say, CSB or any Cinesamples stuff (Sony Scoring Stage) is way different than in the Air studio, but technically speaking it's all wet as long as you use the room mics.

That being said, I feel that the only Spitfire SSO library which is hugely dependant on the room is that brass, and you can do pretty dryish stuff with the close mics of the rest without sounding bad. Sure, there is still that tail in the dry mics, but it's so quiet that it really shouldn't bother. So at least in my opinion, you can treat most of the SSO libraries either dry or roomy, depending on your needs.
 
I love wet libraries but approach them with a view to writing around the strengths of the libraries rather than trying to get them to bend to my will.

Generally, if I'm writing for real instruments then mocking up using sample libraries is a waste of time. Sibelius and Noteperformer work fine. I might render it to audio and play some sample modeling or wivi libraries over the top for fun though.
 
Top Bottom