iPad MP3 problem

[JerryK]

This is not really a Bandhelper problem but I can't think of anywhere better to ask, that would have more experience with iPad audio into a mixer, so here goes.

MP3 from iPad 9th Gen is distorting.  Any clues please?
Regular MP3 files but with some loudness compression previously applied.  Playing from BandHelper via Headphone socket into X-Air XR16 ch7+8 (combo sockets), TRS inputs.
Running the iPad on 4 button-clicks (of the 16) of 'hard' volume and 50% 'soft' (bandhelper on-screen) volume, I hear significant distortion on some recordings.  If I cut the 'hard' volume to 2 clicks, it's better, not perfect, but the level is inconveniently low.  In both cases, the input level to the mixer is way below clipping.  The exact same file plays fine on the PC and there are no run-time changes applied (pitch, level etc) in Bandhelper.
Since the iPad is pretty much designed around MP3 files, I want to think it can't be happening there.
An older iPad behaves very similarly.
I tested a mono DI box on one half but no difference.
I tried using the XLR inputs but no real difference.
I re-loaded the MP3 (which works fine on the pc) to BH but no change.
If it's an impedance problem, it's weird that both XLR and TRS inputs behave the same but I COULD send the audio over USB to my Helix guitar board and out of its XLRs to the XR, if that would even fix it, but I'd prefer to keep the Helix out of that chain.  The Helix guitar output is a single jack to the mixer, so there would be no conflict as such.
What am I missing?  Long time user.

So I added the Helix LT to the chain, now taking USB audio from the iPad and out of the Helix XLRs. No change.
Can it really be that the iPad is not coping with the MP3s?? That's madness. That's almost what it's for.

I could re-create all my backing tracks at some lower level but there are hundreds, so I want to understand what's happening before I do very much.

I have used loudness compression in Media Human for a long time, when making or in fact when copying to my target pc folder but I've found lately that it doesn't treat the whole song the same, so the apparent level of some parts (stems, if you like) changes, in places.  I'm thinking I'll try MP3Gain.  It's a very long job though, to re-load them all from Karaoke-Version and start again.

Anybody had the same troubles?

[arlo]

What happens if you connect a pair of headphones to the headphone jack instead of connecting the mixer, and listen to the recordings through the headphones?

With the mixer reconnected, what happens if you open the Music app instead of BandHelper, and play a recording from the Music app?

[JerryK]

Good questions.
The problem is not with all recordings in BH but I've been testing with one that is an obvious problem.  Not sure if that's skewing my view but I suppose it is.  An obvious problem, though, at only iPad 'hard' volume 4 is therefore not very far away on all of my recordings.

Anyway, replying to your questions: It seems much behaved on headphones.  I barely have a music app on the iPad but 3 demo songs in Anytune, at some 'decent' level are not sounding good on the mixer.  Much better in the cans.

Since the input meters are nowhere near clipping, this suggests an impedance problem but it behaves about the same whether I use TRS or XLR.

Looking closer at the mixer spec, XLR un/bal is 5/10k, TRS un/bal is 20/40k.  So both are way higher than headphone impedance, which is what the iPad 3.5mm socket is intended for.
Shame that the USB via my Helix wasn't a better match.

So what are my available choices, assuming that this continues to be a problem on more than a very few tracks?   I'd prefer it to be a bit more foolproof / less picky.

1. An impedance matching device, maybe just a transformer (x2). Passive preferred if I can get away with it.
2. I'm using Yamaha BT01 for mixer control.  Can I run another one for this?  Latency would not be a problem but I'm not very keen to use more BT or wifi, though it may be worth considering.
3. I tried a DI box but it made no difference.
4. Anything else?

Thanks

[arlo]

So it sounds like the problem is in going from the iPad headphone output jack to the mic/instrument input jack on a mixer. I would suggest:

1) Try going into channels 9-10 on your mixer, which might work differently than 7-8 (1/4" vs combo jacks).

2) Ask Behringer what's the recommended way to feed a headphone signal into the mixer. (I don't see a Line In jack.)

3) Buy a USB audio interface and put that between the iPad and the mixer, instead of using the headphone jack.

[JerryK]

Yes, that how I see it, impedance mismatch.

1. Combo TRS are supposedly the same as 9-14 but I'm not entirely conviced of that, so I tried channels on the second row, with no appreciable difference, as expected.  The impedance is allegedly the same as combo TRS, so the same problem.  Ch 15-16 (TRS) are the high impedance inputs, which would presumably be as bad if not worse.

2. TRS are all considered to be line-in, with the subtle difference in 15-16.  There's no line-in/out (2-track send & return) like you would see on an old analogue mixer, often in phono.

3. I expected the Helix USB in and XLR out to behave as a reasonable USB audio interface. So I'm a bit baffled by that, if it's digital.  In fact, I'm wondering whether it's audio, the same as the headphone socket, being sent to pins on the Lightning connector.  It is a proprietary Apple connector, of course.  I've not looked up what to expect there but it behaved just the same.  Maybe I'll try that again, just in case, then, if it fails, look up a 'proper' adaptor.

(Out of time now - must get some songs into BH and learned for Thursday eve but have 2 days of day-job in the way)

[davelson]

I had a similar problem with MP3 backings, with some distorting... I discovered that the MP3's were recorded at quite different levels. My solution was to Normalise all my files to the same dB level and resave...

Using Audacity on a PC I created a macro to bulk normalise the MP3s and resave

[JerryK]

"I had a similar problem with MP3 backings, with some distorting... I discovered that the MP3's were recorded at quite different levels. My solution was to Normalise all my files to the same dB level and resave...
Using Audacity on a PC I created a macro to bulk normalise the MP3s and resave"

I've been doing that, though not with Audacity, with Media Human, during import from Downloads to desired target folder.  But I've recently discovered that it had used different levels as some songs progress, which is not good.  So I'll be using MP3Gain or perhaps Audacity in future and to fix a few that need it.

[JerryK]

However, my problem and the cause of this post was not different levels, exactly, it's about some songs distorting, despite the levels at every point being well below clipping.
This was surely an impedance mismatch.
So I set about making a passive matching device.
It presents about 30 ohms signal-to-ground to the iPad and, tapped halfway up that resistance, takes a 10k to the output, which then has 47k to ground.  This should satisfy the needs of the iPad and the mixer.
No.  It doesn't.
Somehow, I'm getting the same distortion, even at the same settings.  It's maddening.
I do know that the iPad alters its output, depending on the detected impedance but that's only a switch, yes or no, not continuously variable.

As I said before, the exact same file plays perfectly on the PC.

So what's next?  What's left?  Could it be some falling-out over my reduced bitrate?  The original tracks were 320k but I've been converting to 192k, to save space on one rather ancient tablet.  So perhaps I'll try that - but it's not an effect I've heard of.

[arlo]

Does the distortion affect the same recordings consistently? If so, you could compare the bit rates of the good and bad recordings to see if that's related.

Regarding the USB audio interface, apparently even Apple's $10 Lightning headphone adapter has a digital-to-analog converter in it, rather than taking analog audio from the Lightning port. So that could be worth a try. But I don't know why you had the same problem using your Helix as an audio interface.