For the past couple months I’ve been wondering about what bit rates Inter-App Audio supports.
Obviously nobody answered my question prior to the launch of iOS7. Since its launch, I’ve emailed several iOS music app developers asking this very question. Not one has replied with an answer. Strange? Maybe.
Its a question I think needs to be answered. Today I’ve started asking that question more aggressively by posting it on all of my social media pages and now here. Additionally I’ve sent more emails asking more developers, and some of the same ones I asked already.
We all know that Audiobus, and AudioCopy are always 16bits no matter what your output (iDAW) app is capable of or set to.
It’s not a problem, especially for those like me who create electronic music. However, it’s still a legitimate question.
It may really be a big deal to other people, and or even a deal breaker for some new folks who insist upon the highest bit rates and depths available.
Now, I am no technical wizard. I do not proclaim to fully understand every implication of what bit rates mean to everybody.
I do want to at least know what I am getting with IAA.
I use Auria for the majority of my music productions. It defaults at 32bits audio processing. It can be set to 64bits audio processing. I can’t really imagine any need for 64bits myself. Perhaps you can? It really seems overkill at 64, and holy cow, it must put a huge burden on CPUs not too mention larger project files. All of that aside, I just want to know. What bit depth does IAA support? 16, 24, 32, or what? For some reason nobody has chosen to answer this question. I know I am not the only person interested in the answer. So if you know the answer, please comment on this post. If you also want to know the answer, please leave a comment as well.
It should be a fairly simple thing to answer. Right? So who knows?
WELL! I just got the answer. Inter-App Audio works at 32 Bit floating point. Thank you Rim Buntinas of WaveMachine Labs!
So there you have it. Unless I completely misunderstand the answer, it appears then if you use IAA to stream audio up-to 32bits, the host (if capable) app maintains the very same bit depth. No more being pegged at 16bits.
If that even matters?
Now keep in mind its all really just aesthetics. No human ear on earth can hear bit depth. In order for one to even try they’d have to find the (inaudible) noise floor from an absolutely silent part of the track, crank it up to levels that if music were playing, it would shatter your ear drums and destroy your speakers, just to try and hear what would sound like tape hiss. It’s there, but at frequencies beyond human capabilities. Its a myth that anybody can hear it. It’s not a debate, there’s no golden ear, the science is proven, and the facts are the facts. The human ear and brain simply cannot process such extreme sonic realms, no matter how trained, or how good the equipment is.
It’s not even the point.
That said, 32 bit floating point is still good, or even technically better, but as a person using it, its doesn’t really matter if you know all the technical details. Its nice to know though. What it does mean to you is that there is more room for additional processing of audio without clipping. In other words you can be more relaxed with leveling and whatnot. You get a greater dynamic range to work with. That’s a good thing. It has nothing to do with what you playback and hear.
The most noticeable difference is 32bit will create about %50 larger project files.
More than anything its about what your computer is doing with the audio code, and not so much what you are hearing.
Want to know more about 32 bit floating point audio? Check out this Wikipedia entry or google “32 bit floating point audio explained”. You’ll find all sorts of information on the subject. Most of it is stuff you don’t really need to know. It won’t change how you hear anything. I’m kinda nerdy like that though so its something I’ve been researching.
For the record, I still have a huge amount to learn with the technology aspect to combine with my comfortable understanding of human hearing.
I hope you found my questions, information and answers of some use. I’d still love to hear and learn from your answers what this all means to you and iOS music in your world.
So, you can find out anything on the Internet, we were promised. Just type in your question and am answer will appear from a variety of more knowledgable types than me. My issue is just that, from my search engine I type in obvious key words about ‘How do I….’ Or ‘which apps support…’ or whatever else I want to know and I find nothing, this is where the Forum seems to excel. My point is the Internet itself often reveals no answers, so am grateful to these odd little Oasis that has answered the seemingly unanswerable 🙂
Well, that’s answered the bit rate question then, then 🙂 useful to know, as you said, won’t change my life, won’t get me laid (well that’s what my wife said when I enlightened her over breakfast of my new found knowledge), or mean that I have to move to an island away from marauding Zombie hordes who have been infected by bit rate. Like you though am glad I know, from the Internet. Where you can find out anything, or can you, I guess this proves you can, but how frustrating that very journey can be. Thanks Smitey, from over the pond.
I love your comment. I’m holding my coffee mug trying not to spill from laughing.
To be clear we are talking about bit depth not bit rate, as I initially made the error of posting in the title. Bit rate is actually something else entirely. Like 44.1 is rate 32 bit floating point is depth. It’s a difference (in very brief) of speed vs width. Sort of. Anyway, I have a lot to learn. Cheers
That looks like I was being sarcastic. Sorry, that’s not at all what I was trying to convey. I thought the zombie and getting laid parts were funny, and that what made me laugh.
No problem, and no sarcasm taken at all. Glad you found response funny. I have that very odd English view of life, and I often read and re read my comments and posts as I can make a right mess of words….dyslexic, you see so I often mis read but get the context right. Anyway, great post as usual 😉
Thanks. I inquired on the Audiobus Facebook site about this this morning, and I should have also asked exactly what they mean by the “HD” in Loopy. A well known school/ hardcore studio I studied at in Vancouver insists on 24bit 48k at the lowest for tracking. I think in most instances, that is probably overkill. HOwever, even though it all comes down to our ears, I think there needs to be some major myth-busting about what is really gained by tracking in say 24-44.1 as opposed to 16 bit, other than double the file sizes.
All of the new Apogee interfaces from the Quad down to the Jam support 24 bit or greater, so why not let us users make the choices based on future mastering needs, etc. The Apogee Mic is amazing, but without true support when running through all those crazy 32 bit Kymatica presets in Audiobus, so what?
In one of my discursive rants on your site a while back, I discussed this stuff a bit, and the fact that either Apple iOS specifications or the developers creating apps for Audiobus should be clearer about what can actually be achieved audio wise! AudioMaster, Master Record, Meteor, Auria, MultiTrack Daw, many synths all support 24 bit or greater, so what the hell is the point of it all if it isn’t even applicable? ARRRGGGH!
Thanks again for dealing with one of the most pressing concepts to be considered for iOS “serious status” evolving as it should: AUDIO, AUDIO, AUDIO…It does matter. To me! Even if I don’t hear the difference, lol…
I guarantee, that the more producers start inquiring and posting about this the more answers are going to start crawling out of the ether. It is time.
Obviously my brain is only 8 bit lol, because I am royally confused, and going back to the “what I don’t know (which is clearly a lot)” probably won’t hurt me. Better off trusting my ears than losing sleep over factors whose significance will probably elude me for the rest of my days…
In any case the number of combinations of bit whatevers and sampling rates is probably too many…and I need a bit (or several bits) of rest
No worries. Chris. I think you are right on target, and I am quite confused by the deep technical aspects. Being an end user I don’t really need to know every detail. Just like an Indy race car driver doesn’t need to understand every part of the car.
The only part I truly understand is the human aspect regarding what can be heard and not. In the end it boils down to what does this mean to me or you? Simply put, it means that (assuming the hosts, nodes etc support 32bit floating point) IAA does provide an environment that allows for a greater depth of processing audio and lessens clipping, but increases file sizes by %50. Some people think they can can hear the difference (16vs32) when played back, OK? But that’s like saying they can hear dog whistles or see atoms. They only “believe” they are hearing it, but they physically and mental truly can and do not. Whatever. More importantly it still matters and having the capability of 32bit depth is important. To you, to me, to a lot of people. Whatever it means to anyone, I think it’s certainly worth having.
Great evaluation of the bit factor! I posted on a site about a week ago about favourite, not necessarily the best, audio formats that balance convenience with quality, and the results were more diverse than I expected.
I had a whack of vinyl sitting around for years from back in the days of actually hating clicks and pops on records; I offed them, and now I go seeking the same sound in iPad apps like Elephantcandy Vinyl to retrofy parts, while leisure listening to 128 kbps AAC files like everyone else these days on an iPhone, and still pining over the glory days of dubbing all my favourite songs onto the best TDK 90 minute cassettes I could afford before I heard the first audible pop or click on a record…
I am way, way out of my bit depth…
Oh the memories of those TDK 90 Minute cassettes. I used to have so many. Lots of recordings off the radio of my old favorite electronic music show called ” Alien Air”. I kind of miss that era.
I’d take road trips with bags and bags of tapes taking up at least one passenger worth of space. Those were the days huh?
I find the old tapes much more mysterious than the old records now. I don’t listen to them often, but when I do, I always come across something I had forgotten about. A new perspective on a lost piece of music…They could take quite a beating, or could be mangled by an evil transport in an instant. I notice one of the affiliated sites has a real cassette aficionado, and that a few labels are honing in on the medium. Kind of cool…but will I bother?
I did a bit of experimenting when I was recording a friend and his classical guitar. What surprised me was how obvious the improvement was when I went from 16 bit to 24. REALLY obvious. Another friend told me that it’s actually only the difference between 16 and 20 that we can hear but it was there for sure. 44.1 to 48k made no difference that I could discern.
Interesting. I have an old Lexicon MPX 100 that is 20 bit. On my desktop, I always record at 24 44.1 (never 48, because of speed compatibility issues), and I honestly don’t know why. I may have thought I heard a difference once, long ago, before I clogged all my hard drives with all those massive audio files. Have I been duped.
In photography, there is a guy named Ken Rockwell who always goes on about the “megapixel myth”. More critical factors are lens quality, lighting and technique, and if you do not blow your images up to wall size, there are many instances where the increased file sizes to not reveal any better image quality.
I am not 100% sure, but I think audio systems tend to be the great equalizers, and some super high resolution sounds might be better appreciated with reproduction systems up to the task.
Try the blind tests…I think one of the things Smite is trying to get at is the importance of the resolution in the instruments and effects, where smoother reverb tails, decays, etc. can benefit from all those crazy bits…
When I’m doing drum tracks and ifretless and Nave and whatever 16bit works, but one acoustic guitar 24bit is clearly superior.