Audiobus: Use your music apps together.
What is Audiobus? — Audiobus is an award-winning music app for iPhone and iPad which lets you use your other music apps together. Chain effects on your favourite synth, run the output of apps or Audio Units into an app like GarageBand or Loopy, or select a different audio interface output for each app. Route MIDI between apps — drive a synth from a MIDI sequencer, or add an arpeggiator to your MIDI keyboard — or sync with your external MIDI gear. And control your entire setup from a MIDI controller.
Download on the App StoreAudiobus is the app that makes the rest of your setup better.
Collected thoughts on sound reproduction and monitoring.
This is going to be a series of short articles based on my research into sound reproduction and audio over the course of many years. I'm not claiming to be an absolute authority on the subject, but it's something that I'm very interested in and that is widely misunderstood IMO. Discussion is welcome.
Comments
What is neutral?
You would think this should be an easy question to answer, after all shouldn't a neutral playback system simply faithfully reproduce sound exactly as it was recorded?
It turns out that physics makes the task pretty much impossible. Why? Because speakers inevitably interact with the room they are placed in. The sound you hear is not from the speaker alone, but from the interaction between the speaker and the room. Rooms have a profound influence on the sound we hear from a speaker, especially in frequencies below 500hz.
This makes it impossible to faithfully reproduce recorded sound, since every room is different and engineers have no control over the rooms that recordings will be played back in.
Back in the 1970s this was a big problem for loudspeaker designers, because this lack of objectivity meant that the industry was drowning in a sea of subjectivity and guesswork. It seemed impossible to make quantifiable decisions on frequency response, since the room was as much a part of this as the speaker.
Manufacturers, critics, and consumers alike all relied on totally subjective metrics to evaluate hi-fi gear. This meant that there were no standards to rely on and the sound of audio components was all over the place, dictated by the personal whims of the people who designed and manufactured them.
So in the 1970s Floyd Toole, a PhD in electrical engineering, began a research program at the National Research Council of Canada to try and create some empirical basis for sound reproduction. Over the course of 20 years he and his team carried out a large selection of double-blind tests with hundreds of subjects to see if it was possible to establish some objectivity to loudspeaker design.
Their principal finding was very interesting: the loudspeakers that most people subjectively preferred in blind tests were also the ones that measured as flat when placed in an anechoic chamber (an anechoic chamber is a room full of diffusing surfaces that eliminate room reflections).
So it seems that the human brain is able to filter out of the influence of the room, and test subjects heavily favoured speakers that sounded flat in anechoic chambers - our brains are able to discern a neutral speaker, even in an imperfect room. We simply filter out the influence of the room, at least when comparing one speaker to another. This doesn't mean that rooms aren't important - they're crucial, but more of that in another post.
The great news about these findings was that it was now possible to establish some empirical basis for speaker design: make sure they are flat in an anechoic chamber. This profoundly affected speaker design, leading to much better and more consistent sounding loudspeakers.
In 1991 Floyd Toole was hired by Harman-Kardon where he continued his research program. He retired in 2007 after which the program was headed by Sean Olive.
Sean Olive undertook some research into headphones starting around 2010, once again based on blind tests. What he found was that most people favoured headphones that had the same sound signature has a good pair of speakers in a typical room. Sounds logical, but this does come with some surprising implications.
Headphones that measure perfectly flat will sound very bright compared to speakers in a room. Why? Because speakers placed in a room will have enhanced bass and attenuated treble. The room will amplify the lower frequencies and absorb the higher ones, leading to a frequency response slope that is higher in the bass and lower in the treble.
This means that headphones that measure perfectly flat will not sound natural, because as humans we expect to hear sound in the context of a room. Flat headphones eliminate the room, and their resulting sound is thin and unnatural.
So the research team at Harman devised a frequency response curve that more closely resembles speakers in a room: slightly boosted bass, and slightly attenuated treble. In blind testing this curve was favoured by a majority of listeners.
(The curve also includes the necessary peak at 3khz to account for the acoustics of the head and ears, which is common to all headphone reference curves)
It's important to note that this finding is more subjective than those from the research into loudspeakers, which was much more conclusive (that the vast majority of listeners prefer speakers that measure as flat in an anechoic chamber).
Here is the breakdown that Sean Olive found:
“Harman curve Lovers”: This group, which constitutes 64% of listeners, includes mostly a broad spectrum of people, although they’re generally under age 50. They prefer headphones tuned close to the Harman curve.
“More Bass Is Better”: This next group, which makes up 15% of listeners, prefers headphones with 3 to 6dB more bass than Harman curve below 300Hz, and 1dB more output above 1kHz. This group is predominantly male and younger.
“Less Bass Is Better”: This group, 21% of listeners, prefers 2 to 3dB less bass than the Harman curve and 1dB more output above 1kHz. This group is disproportionately female and older than 50.
What does this mean in practice? Well having a standardised curve allows us to make useful comparisons. Even if you don't personally like the sound of the Harman Curve, having a standardised reference means that you can still evaluate whether a headphone will be right for you by looking at the frequency response. If you favour less bass than the Harman Curve you can tell straight away which headphones are likely to appeal to you. Again it's a matter of removing guesswork and subjectivity.
And since Sean Olive first published his findings in 2012 there has been a proliferation of graphs and measurements on various sites, most of which use the Harman Curve or something similar. This makes it much easier for people to choose the right headphone for them.
Some useful references:
Floyd Toole: Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms
https://www.soundstagesolo.com/index.php/features/217-where-are-we-at-with-the-harman-curve
Oratory1990 - what does harman and diffuse field equalization mean?
Next week: Bass, the final frontier
very interesting. thanks!
Thanks , interesting read!
Fascinating, and I learned some useful new things. Looking forward to what comes next!
Thanks for the excellent public service work, mate. 👍
Sold. Had me from the thread title.
Really interesting article too. Thanks.
Next instalment:
Bass: The Final Frontier
Bass is the hardest thing to monitor accurately. Especially with speakers. Yes that's right, despite what the conventional wisdom says, if you want to monitor bass accurately you're more likely to succeed with headphones. Headphones however have their own issues, but more on those later. Let's start with speakers.
The average studio monitor in the average room is likely to have absolutely terrible bass accuracy. Not because the average studio monitor is bad, but because of the way speakers, even very expensive speakers, interact with the room they are placed in.
Rooms will mess with your bass response, and there's not much you can do about it without spending a lot of money.
Firstly the room will skew the frequency response of your speakers, by emphasizing the bass and absorbing the treble, and no two rooms will do this in the same way.
Bass, due to its lower wavelength, will radiate out from your speakers in every direction. This in turn means that it will be reflected by every surface, including the wall directly behind the speaker.
Treble, on the other hand, is much more directional. It will radiate forward from the tweeter, and so will not be reflected by the back wall. Treble in fact is more likely to be absorbed if it encounters a soft surface such as a curtain or a sofa. Bass frequencies are not so easily absorbed.
What this means in practice is that rooms will emphasize the bass, and attenuate treble.
Bass radiates out in all directions from the speaker, whereas treble radiates forwards
This is not as much of a problem as you might think though, because our brains are good at compensating for the effect of the room. If you were to listen to a pair of perfectly neutral speakers in an anechoic chamber, it would sound really weird. So perfect neutrality is not what we are looking for here, even if such a thing were achievable.
The issue is in fact that no two rooms will sound the same. Some rooms have lots of soft furnishings and furniture to absorb and diffuse sound. Others have wooden floors and bare walls that will reflect sound. The same set of speakers in those two rooms will sound completely different, with a radically different frequency response.
You can't trust the room, even if you have taken the trouble to install acoustic panels and bass traps. The best you can hope for is to get used to the room.
And that's before we even mention the biggest problem: standing waves.
When the wavelength of a sound wave matches the dimensions of the room you will get interference - as waves are reflected back off the walls they will amplify or cancel out incoming waves. You get boomy bass or bass that disappears.
A sound wave at 85hz measures roughly 4 meters, a soundwave of 115hz measures roughly 3 meters. If your room is 4m x 3m you will have standing waves at those two frequencies, and boy will they mess up your mixes.
Trying to fix the bass in your mix in an untreated room is like trying to perform brain surgery with a hammer and a nail: you'll be doing more harm than good. Forget it.
Can this be fixed? Yes, to some extent.
The absolute most cost-effective method is to invest in a calibrated measurement microphone and use it to analyse your room (in conjunction with software such as Room Equalisation Wizard). You can then create some correction curves that will attenuate the standing waves at your listening position and create a more reliable low-end response.
If you have the money you can also invest in bass traps and acoustic treatment. This is a much more expensive and time-consuming route, but is essential for anyone hoping to monitor bass accurately.
Without room treatment simply assume that you can't trust the bass you are hearing in your monitors.
OK, so rooms suck. What about headphones? Well actually headphones still suck, but they suck a lot less. Getting reliable bass from headphones is easier to achieve, and at a much lower cost.
There are 3 main categories of headphones: open-backed, closed, and IEMs (In-Ear Monitors).
Open-backed headphones, due to their lack of enclosure around the driver, will have a steep drop-off below 100hz. This means that sub-bass frequencies are essentially lost. This applies to pretty much every model available, from the Sennheiser HD650 to the $4000 Focal Utopia.
You can mitigate this with correction software such as Toneboosters Morphit though, which I definitely recommend for any headphone user.
Closed-back headphones can handle sub-bass better, but due to their closed design they can suffer from resonances in the earcups which are difficult to correct since they vary from person to person.
IEMs probably have the best and most accurate bass response of any monitoring option, as long as you can get a good fit and seal for your ears. A poor seal will result in terrible bass, and some people can struggle to get IEMs that fit their ears. Personally I don't enjoy using them even though I respect their accuracy, because I find them too fiddly and uncomfortable to want to wear them for extended sessions.
So headphones, especially relatively neutral headphones coupled with corrective software such as Morphit, will give you the most reliable bass accuracy on a reasonable budget. A really good pair of IEMs is going to be the best option, but even open-backed headphones with corrective EQ should be acceptable.
You probably want headphones that are tuned to something like the Harman Curve, because the Harman Curve is somewhat similar to speakers in a room. For some people the Harman Curve might sound quite bass-heavy, but the danger of mixing on headphones that are lacking in bass (or that have unrealistically light bass) is that you could be creating mixes that sound fine to you, but will be really boomy when played back on speakers. You can't mix what you can't hear.
So when buying a pair of headphones for mixing, check the frequency response graphs, and check how these compare to the Harman Curve. Even if you don't like the Harman Curve (and not everyone does), be sure to buy headphones that are tuned to some kind of reference curve, not just some random sound signature. And then use Morphit to make them more accurate as monitors.
Thank you again, Richard. And an idea: given the results of that poll from earlier this year, which revealed that the average age of us forum members is about 132, will you be doing an article on what to do when you just can't hear hi hats any more? 🤷😊
I can't help with the hi-hat issue, as I'm slowly going deaf myself 🙉
There is an app that you can use the test how much high-frequency loss you've suffered:
https://apps.apple.com/us/app/audio-function-generator-pro/id895074703
My hearing cuts out just before 13khz, so the hi-hats are already getting a little bit less fizzy.
I’m loving this series. It got me to reconsider my IEMs, which for some reason I’d written off as not suitable, when they’re really quite awesome.
Testing my hearing with that app was depressing though. I can’t hear anything past 11k, and it’s nearly silent from 9k onward.
The free version of the app worked just fine for me:
https://apps.apple.com/us/app/audio-function-generator/id768229610
The numbers are misleading: that missing 10khz bandwidth is just 1 single octave.
And there isn‘t much going on there anyway - it‘s sometimes called the air-band.
(you may have the doctor check for deep stuck dirt in your ears, though)
Yes you've probably seen this chart that's been posted in a few places - a decent reference:
(Edit) and since every octave is a doubling of frequency as @Telefunky says, there's only one octave between 10khz and 20khz
I'm glad you found it useful
It's perfectly OK to mix on headphones and/or IEMs, you just need to also check the mix on other systems, including speakers.
This is a pretty good video I posted in a previous thread:
Awesome work, thanks Richard.
I’m sure most of us have seen this, but…
F**K SECRECY: Hearing Loss and Music Production. Let's talk.
These articles are great! Very interesting. Scared to test my hearing though…
I’m in big time need of an ear cleaning, my left ear is pretty muffled
Nothing to be scared of. Loss of high frequency hearing is an unavoidable aspect of aging.
Too many nights DJing where the club system monitors scream at you has left me with lifelong tinnitus. On the upside, my ears cancel out the piercing high strings towards the end of Barber's Adagio for Strings...
I had my ears cleaned by my ear doctor recently, made a world of difference. I can hear about 2K higher now. Sub bass is more audible now too. I've had to adjust my mixes as a result, but it was worth it since everything is much clearer now.
Random thoughts -
@richardyot Let me know if that info above isn't appropriate and I'll delete it.
It's all good
I had my ears syringed a few months ago, which got rid of the low rumble that was constantly in my left ear.
Thanks and very well written, understandable articles, Richard. I was considering IEMs for a long time, but on rare occasions I will get some super loud volume from an error or a glitch… so afraid of the damage that would cause.
At 49 I can't hear 'clearly'(ie. it becomes 'guesswork') past 11k (sine wave) and I guess the rest of the 'harmonics' are created by my brain as I know partially what to expect and I'm ok with it... (I can always use a spectrum analyzer to sort out 'problems' in case they get disturbing).
I am however quite sensitive to 'sub harmonics' (ie. I can easily spot a malfunctioning switched transformers, ground loops, and other 'disturbances', low-level clicks, pops etc.).
Sometimes I wish I could 'see' as clearly as I can hear as my eyes-sight is degrading faster than what I had hoped for...
I really expected this to happen to me as I was DJing for 10 years from the late 80s, pop pickers! Actually, let's just clear that up before it gets out of hand. Reggae and then House.
Anyway, I am probably tempting fate but I just checked my hearing and it is above 11000hz still so I'm happy with that. But yeah, those were the days that there were as many places with a table on a stage, looking at you Sheffield Uni, as an actual booth or whatever.
I have a slight theory that as people grow out of the sharp other-offensive music of their youth (punk, or in my case 70s glam) and proceed headlong into getting older and settling down, the liking of that oh-so-boring soporific wallpaper such as Dire Straits, Phil Collins, Steely Dan etc and indeed the making of exactly that sort of AOR (any old rubbish, I mean adult oriented rock) is guided by what we can no longer hear at those ages. The makers of such AOR can’t hear the top end, and their consumers can’t either, so it matches. I’ve noticed I’m only putting into my own music what I can actually hear comfortably, so my stuff probably will appeal to people my age (60).
Veering crazily back toward topic, I was thinking of buying some other near field monitors recently but I decided to stick with my JBL Control Ones (with a new Nobsound stereo amp that fits in the palm of my hand, costs little, but sounds superb!) basically because all the new active speakers I was perusing the main advantages were in their top-end flatness or lack of. Well, I can’t hear up there anyway, and consequently I don’t make music to fill up there either, so I can get away with the Control Ones easily, they’re already more toppy than I can hear.
WTF is HRTF?
HRTF, or Head-Related Transfer Function, is a measurement of the frequency response at the eardrum. Typically an HRTF curve would look like this:
As you can see there is a gigantic spike at around 3khz, right in the main frequency range for human speech.
So what the typical human eardrum hears is not "flat" or "neutral". The anatomy of our outer ear, our inner ear, our head, and even our shoulders will affect the frequency response of the sound that reaches our ears.
We are quite literally tuned to be more sensitive to the frequency of the human voice above all other sounds. Our anatomy has evolved to amplify the region at around 3khz, so that we can hone in on speech.
Of course we are not consciously aware of this, since our brain interprets this 3khz spike as "neutral". But if you are going to make a pair of headphones this becomes important. A pair of headphones without this 3khz spike will not sound neutral, it will sound weird.
So manufacturers of headphones have to take the HRTF curve into account when tuning their products, so that they sound natural.
And to achieve this they will use a dummy head and take measurements at the eardrum.
There's just one small problem: humans come in many different sizes and shapes. There are enormous variations in the shape of the outer ear, and these will profoundly affect the frequency response that individual people hear. Your ears are not the same as mine, quite literally, because they have a different shape, and are at a different angle to mine. This means that each individual has their own HRTF curve.
The closer your particular anatomy matches the dummy head used by the headphone manufacturer, the more accurate the tuning will sound to you. But if your anatomy is very different from that of the dummy head, there is a strong chance that the tuning won't sound right to you. Maybe your ears have a bigger spike at 3khz, or maybe the spike is at 2 khz, or 4 khz.
Variations in human anatomy make it impossible to design headphones to suit every individual. This means that there is no such thing as a "neutral" headphone.
And as we know from the previous articles there is also no such thing as a neutral speaker, since speakers are always affected by the room. And rooms vary just as much as human anatomy does, so perfect sound reproduction is physically impossible, whether you are using speakers or headphones.
This article adds a little background to the frequency range of human speech. While the fundamentals are in the 200hz region, consonants and the overtones of vowels are found in the 2-5khz region that our ears are tuned to amplify:
https://blog.accusonus.com/pro-audio-production/human-voice-frequency-range/
Speakers vs monitors
Are mixing monitors more accurate and neutral than consumer speakers? Good speakers should be designed to have a flat on-axis response (and so measure as neutral in an anechoic chamber), and this applies to any speaker, whether it is marketed as a monitor or a consumer loudspeaker.
In blind tests the vast majority of listeners express a strong preference for speakers with a flat on-axis response. This applies to trained experts such as mixing engineers as well as regular people. A good consumer loudspeaker should be just as accurate as a monitor, and many are.
Of course there are consumer loudspeakers that are highly coloured, with excessive bass and/or treble, but if you buy a high-quality consumer speaker with a well-tuned frequency response well it should perform just as well as a high-quality monitor.
Frequency response is by far the most important quality of any loudspeaker, and most other characteristics (transient response for example) stem directly from having accurate frequency response.
The only drawback of buying a consumer speaker is that it might require more research to check if it is well-tuned. You would expect monitors to be neutral by default (although this is not always the case, Yamaha NS10s are far from being neutral). In any case you should check the FR graph of any speaker you are planning to buy.
Nearfield monitors are designed to be placed close to the listening position, in order to minimise the effect of the room. However in most rooms this is impossible since the bass frequencies will always be affected by the room, and the first reflections from the side walls are also impossible to avoid without room treatment.
In summary: there is nothing inherently special about monitors, they are just speakers. A good speaker is a good speaker.