I've got a little bit of audio code involving a delay line. I've done this before and on every single platform, the code works fine.
Except on iPad4 - ONLY. Iphone5, Iphone4, ipad Air all work fine - but on this one, the delay code (with feedback) is extremely prone to blowing up. It creates an ever increasing feedback loop that again, doesn't exist on any other platform.
As far as differences in processing go, apparently this iPad uses the a6x chip, while the others I've tested are A6 and A7.
Does anyone know any differences here that might explain this signal issue I'm having? The sound is noticably different on this platform specifically, and I've no clue what could be the case.
If you have any inkling of an idea - let me know and I will, as always, report back with my findings.
Wow I wouldn't haven't believed this until I confirmed on multiple devices and at this compatibility list page: http://store.griffintechnology.com/guitarconnect-cable
Notice that ipad4 is the only one not on the compatibility list.
The iPad4 produces a hardware feedback loop if you use a connector similar to the one in the link (that is a Griffin GuitarConnect, but I also confirmed the hardware issue with an IK Multimedia iRig).
This drove me crazy looking in my different effect code trying to find some unset variable or a scalar out of whack...it wasn't until I cleared all input, and I could still here the output feedback through the headphones that I started to look around for external confirmation.
Watch those iPad4's if you are doing something similar.
Crosstalk with headset interfaces (assuming that's the issue here) is a known problem with all iDevices. Some are worse than others - I don't have an iPad 4, but I know that an iPad 2 is much worse for feedback than an iPad Mini Retina. The root cause is electrical crosstalk within the cable and the iPad audio circuitry, where the high level headphone signal leaks into the low level input signal. It gets worse the more gain your app applies to the input signal, so it's a problem for guitar amps in particular...
A delay loop would certainly suffer the same problem, unless you peg the output level with a limiter or restrict the feedback amount.
Phase-reversal of the stereo output (which cancels the crosstalk to some extent) is the only workaround I know of, but sonically speaking it's a crap solution. Better to tell your customers to avoid cheap headset interfaces like the Griffin, and instead invest in a digital interface which uses the dock connector (e.g. an Apogee Jam)
Yea - thanks for the more detailed info. It was with high gain that it became most apparent.
It really isn't a problem on any of the other devices I've tried, but the ipad 4 is just awful...with both the griffin and the irig (same underlying technology).
Having said all that, the griffin is actually a really cool little cable.
