Just now I finished listening to an album we did in the Studio with heavy metal artists Ammunition. The music was heavy indeed but at the same time was very nice to the ears. Myko and Allan tracked this project in 2003, while I mixed it much much later. This project really was a quick low-budget but fully multitracked production, with rarely a second take done during tracking. Thus a typical song was completely recorded in an hour or less, and on most of them, there was only one rhythm guitar track present. Meanwhile, the mix for the whole album was completed in about four hours, all ten songs of it, and because I put a limiter (plus some other mojo stuff) on the two-buss, it was quasi-mastered already during mixdown. I did what I could because I wanted the songs to sound like it was a major release, and if you get to hear the album, I think you will notice that the fine musicianship of the band, the intensity of their performance, and the quality of the recording reflect none of the constraints I just described at all…

The guitar sound on this album reminds me of an audio “secret” that I’m going to share with you today. While we usually multi-mic the guitar amp on most sessions (even for demo recordings), for some reason, Myko used only one mic on many songs here. I was going for a dry and massive guitar sound, primarily because the arrangements were very sparse, and so I wanted to really push it as far as it can go. And so the main problem I had was to create a good, convincing stereo sound from this single source, and because I really hate these pseudo-stereo processors (because they all sound like a bad ice-cream headache, and collapses really bad in mono), I went old-school and used the Haas effect while I crafted most of the guitar sounds.

Here is the song “Against the Wall” off the album for your listening pleasure. Try to focus on the guitar sound and how its stereo image is presented. (With special thanks and permission from Rodney, Sidhart, and Barry of Ammunition.)

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

The sounds of silence

Let’s set aside the song for a while… Close your eyes for a minute and listen intently to the sounds you hear all around you. There could be a lawnmower coming from the outside the window, to your right, while you could hear your dog barking somewhat to your left, at 10 o’clock, and your TV is airing the news dead center. How do you know exactly where all these sounds are coming from? And how do you know how far they are from you?

Well, we have two ears separated by our head by about 18 cm facing opposite directions. And when sound comes from a certain direction, it reaches our ears at different times, and with different intensities. If the sound was coming from the median plane, like the TV, then it would arrive in both ears simultaneously, at almost the same loudness. If it was off to the right, as with the lawnmower, the right ear would receive the sound before the left (and vice-versa as with the dog barking), with the louder side obviously coming from the the ear facing the sound source. It really gets a lot more complicated that this, but this knowledge should be sufficient for our topic today.

In the perception of sound, we can think of our ear-brain system as some kind of audio processor. And in the 1940’s Helmut Haas found out that when two identical sounds originate from two sources at different distances from the listener, the ear will only recognize the sound that arrives first, and will combine the second source as long as the delay between them is around 40 ms or less. Any other sources of the same sound arriving within this time window (for instance, the reflected TV sounds from the room walls surrounding you as in the example above) will fuse with the main initial sound, and thus we essentially hear just “one” sound, and only a loudness enhancement occurs. Only reflections arriving later than 40 ms will be heard as distinct echoes.

Down deep into the pain

In the audio recording world, depth, or the three-dimensional image of a mix, is one of the hallmarks of a great mixer. It is one of these things that separate the seasoned professional from the amateurs. Depth is this wonderful listening experience where each instrument seem to occupy a distinct location within the sound field, all the while realizing that this feat is done in stereo, that is, with two speakers only.

And here is where the notion of sonic depth comes in… The Haas effect allows echoes within 40 ms to enhance the original sound without confusion as to where it’s coming from. While all these sounds and reflections coming in from different directions are fused, our brain continues to recognize the location of the initial sound as *the* source without confusing its directionality.

Dolby Labs use this trick to create what they call the “magic surround”. They found out that when a stereo source was duplicated in the rear speakers, and delayed by a certain amount, the ambience within the stereo recording became more prominent. It seems as if the natural reverberation from the original source was extracted into the rear speakers when the delay was introduced. They use some other mid-side tricks as well, but this is the essential gist of the technique.

The sound of one hand clapping

When a snare drum is hit, the sound picked up by a mic placed over it will be the combination of the direct sound and all the reflected sounds from the room (i.e., the walls, ceiling, floor, and other furniture and objects within the room). All these reflections, within the Haas limit, are directly correlated with the original sound, and supports the initial direct hit, without interfering with its left-to-right orientation. The longer reflected sounds (e.g., reflections if the room was sufficiently large, or secondary reflections, and so on) beyond the Haas limit are uncorrelated, and this is what we call the natural ambience of the room. This group of early reflections as it’s also called is what gives us the perception of distance, a sense of space, but they don’t help us locate the original source in space.

This is the reason why the sense of location of the original sound is usually destroyed when synthetic reverb is added to a single-miked source. Compare it to a stereo-recording of the same source, where early reflections are now binaurally captured, and you’ll see that when the same reverb is added, the sound remains convincingly natural and the spatial location is preserved.

Frequency response also affects our sense of distance. Air absorbs high frequencies, such that the further down a sound source is, the more its high-frequency content is reduced. And our ears have been trained to interpret distance like this. As with our lawnmower example above, the direct sound coming into our right ear contains more high-frequency content, than the softer sound coming into our left ear. The combined Haas effect is thus interpreted in our brain as a lawnmower that is as far right of our sound field as can be.

Back to the egg

And thus with this newfound understanding of this particular aspect of the psychoacoustics of hearing, and its myriad possibilities, we come back to our mono guitar example. What things were done to the original guitar sound that resulted into the mix you heard above?

Ok, for starters, there were a few things, such as adaptive noise reduction, EQ (to sculpt the basic sound) and compression, that are totally unrelated to our topic (but had to mention as well, so there!). As far as creating that stereo sound, the original track was hard-panned on one side, and duplicated onto the other side, which was hard-panned as well. This duplicate track was then delayed by a certain number of samples, corresponding to a certain time delay within the Haas specification. How much exactly depends on the resulting spread and tone in stereo, and how the stereo sound behaves (or misbehaves) when collapsed to mono. I find that there is no general rule we can follow here, because the nasty comb filtering effects vary with the program material, and in the case of guitars, these things actually change the timbre and tone as well.

A common mixing trick you can further try out (prior to delaying the duplicate track) is to detune the duplicate track (using a pitch-shifter) by a certain amount, usually no more than 10 cents. This increases the stereo perception even more. I sometimes do this with vocals, but in this case, this technique was not employed.

The two guitar tracks are then fed to a stereo-buss where further EQ (this time to complement the vocal track) and compression are added. A tiny amount of plate reverb was added at the very end, just enough to glue the elements, but whose amount you can barely hear. It is also important that the reverb doesn’t wash away the attack transients of the guitar.

And so there you have it folks. A mono guitar track turned into rich stereo. Have fun everybody!

For additional reading, please check out the following:

• The article Haas effect – Wikipedia, the free encyclopedia discusses the basics.
• Haas Effect (Precedence Effect) is another good introduction to the topic.
• In Stereo Perception, the various listening perceptions related to our anatomy are dissected. Things such as front-to-back and elevation perceptions are also explained.
• In Depth and Dimension, mastering engineer Bob Katz discusses the various psychoacoustic phenomena (including the Haas effect) as it relates to our sensation of depth.
• The paper ASC, The World’s Best Studio Acoustics – SOUND FUSION AND THE ACOUSTIC PRESENCE EFFECT, discusses the complex differences between the Haas effect, early reflections, comb filtering, and the phenomenon of masking.
• In The Birth of the Array: Part 5, knowledge of the Haas effect is used extensively in speaker placement and delay compensation in big concert venues that use multi-array loudspeakers.
• An excellent Haas effect mixing tutorial described in step-by-step detail may be found in the article Turn a mono track into rich stereo « benvesco.com.
• Another ultra-cool mixing tutorial may be found in this article Pro Mixing Series: Episode two: The Haas Effect | EMusicTips.
• TAP-plugins hosts a nice freeware Linux plugin Tap Stereo Echo that achieves the Haas effect (among other delay capabilities). Too bad I’m using Windows!
• Another great freeware plugin HaasCheezburger may be found in this article Native Instruments Kore, Komplete, Reaktor @ Create Digital Music » I Can Haas Stereo_ Lolcat Reaktor Delay FX.

This old Caltech paper reminds me of the older Oohashi paper that deals with the same subject (which I couldn’t find on the web anymore!). As you may know, in this day and age, digital audio is brick-walled at around 20kHz, since we believe that the human ear is incapable of hearing above this frequency, and thus there is no use to include this information in the audio signal (unless you are recording for bats or dogs). That is why no matter what sample rate is used for any audio project, like with the prevalence of 96kHz today, we never get to hear any audio information above 20kHz during reproduction. (Well, I simplified too much. I also need to mention that very few microphones and speakers go beyond 20kHz. And analog tape can certainly record higher frequencies.)

The paper says, “A discussion of the significance of these results describes others’ work on perception of air- and bone-conducted ultrasound; and points out that even if ultrasound be taken as having no effect on perception of live sound, yet its presence may still pose a problem to the audio equipment designer and recording engineer.”

There’s actually a lot of hot debate going on regarding the benefits of high-sampling rates for recording audio, and also a lot of marketing bullsh*t for new products that are capable of recording at such frequencies. But really, higher is not better, and Dan Lavry tells you why. The old Oohashi paper actually concluded that while audio is surely inaudible beyond 20kHz, the ultrasonic frequencies trigger some electrical brain activity in the listener’s head that changes the overall listening perception, or something along those  lines if I remember correctly.

Here’s the link to the Caltech paper… http://www.cco.caltech.edu/~boyk/spectra/spectra.htm. Have a read if this interests you.

Let’s admit it… Some people just cannot sing. While most of them struggle with identifying the correct notes to a melody, a few people I’ve heard cannot even sing in time, which is even weirder. Now, maybe, while an individual can identify the right frequencies, he cannot deliver it with his voice. It could be so just because he didn’t have the experience or the facility to do it early on in life, and thus because the skill wasn’t learned, it now becomes impossible to do it.

Is music a purely human experience? What exactly makes us human beings experience music the way we do, intellectually, emotionally, spiritually? And now, since we are able to immerse ourselves into the music, we also usually like to participate, and the simplest way to do this is through singing. But what is it inside us that lets us do this properly, and what is missing in less fortunate people that eludes them from singing in tune?

A very interesting read… http://www.newscientist.com/article/mg20327223.300-tonedeaf-people-have-fewer-brain-connections.html?DCMP=OTC-rss&nsref=online-news