The iPod Poor Bass Response

[PHK Brasil]

PART 1

This is not a new, but, for those who doesn't know, iPods got problems on the bass response when using some headphones. The lower the impedance, the worst is the problem.

These are the results of the frequency response test, I did with my iPod 5G.

Although most of the mobile headphones have 16 or 32 Ohm, I used a lot of impedance values to check what happens above 32 Ohm too.

Is not just a sub 32 Ohm problem.






PART 2

I did two simple settings on Rockbox EQ, trying to compensate the bass response attenuation. The graphs shows a better response with the compensating EQ. I didn't spent much time doing this, therefore, is not 100%.






For those who wanna check, compare the sound between these EQ presets and flat, to notice the difference.

Is not a big deal, but exists, and is bigger with 16 Ohm headphones.


Compensating EQ for 16 Ohm:

Advanced EQ Settings:
Low Shelf Filter -> Cutoff Frequency: 50 Hertz
Low Shelf Filter -> Q: 0.6 Q
Low Shelf Filter -> Gain: 3.0 dB
Peak Filter 1 -> Center Frequency: 20 Hertz
Peak Filter 1 -> Q: 0.5 Q
Peak Filter 1 -> Gain: 0.5 dB



Compensating EQ for 32 OHm:

Advanced EQ Settings:
Low Shelf Filter -> Cutoff Frequency: 20 Hertz
Low Shelf Filter -> Q: 0.6 Q
Low Shelf Filter -> Gain: 3.0 dB


PS: Thanks to saratoga for the help.

♪

[lights0ut]

Thanks PHK, glad to see a more tailor made solution rather than just turning the bass up. It's disappointing to know that a company who's market share is ~80% wont make a quality DAP.
(yet another reason why I wish my ipod was a gigabeat )

[Febs]

Thanks for posting the graphs.

PART 1Is not just a sub 32 Ohm problem.

For sub-32 Ohm loads, the graphs clearly show a problem, but for higher loads, it looks much less significant.  The bass attenuation at 50Hz for a 32 Ohm load is less than half a dB at 50Hz and 1dB at 30Hz.  I'm not sure that would be audible.

[DefineByte]

Is the line out any better?

[saratoga]

Is the line out any better?

I don't think you'd want to drive a load with the line out anyway, so it shouldn't matter.

[DefineByte]

Seems I wasn't looking at the graphs properly. I was ignoring the no load part. xD

[soap]

Compensating EQ for 16 Ohm:

Advanced EQ Settings:
Low Shelf Filter -> Cutoff Frequency: 50 Hertz
Low Shelf Filter -> Q: 0.6 Q
Low Shelf Filter -> Gain: 3.0 dB
Peak Filter 1 -> Center Frequency: 20 Hertz
Peak Filter 1 -> Q: 0.5 Q
Peak Filter 1 -> Gain: 0.5 dB

Thank you ever so much, this is something I've been long meaning to do, but never gotten around to.
How much do I need to pay beg you for a RMAA graph of the above EQ setting without the .5dB peak filter?
0.5 dB seems like a very small tweak for the extra CPU load, but I'm curious to see the difference.

[PHK Brasil]

For sub-32 Ohm loads, the graphs clearly show a problem, but for higher loads, it looks much less significant.  The bass attenuation at 50Hz for a 32 Ohm load is less than half a dB at 50Hz and 1dB at 30Hz.  I'm not sure that would be audible.

Yes, I agree. I just mentioned "Is not just a sub 32 Ohm problem", cause once I read that someone is adding resistors to their headphones till they reach 33 Ohm, thinking to solve the problem this way. Or someone thinking that a Shure E500PTH is totally free from this, cause has 36 Ohm impedance.
But, there's not a miraculous cure in 33 Ohm.


Before all of this, I did tests (with music) adding a pot in my Apple 32 Ohm earbuds, and I noticed audible changes in the frequency response, as I increasing the impedance. But the pot I used has a much higher value to a detailed test.

But the impression I had, was that not only the bass changed, but from a certain point in ahead, the mid and highs also were changing. But the graphs don't show changes in mids and highs...

Thank you ever so much, this is something I've been long meaning to do, but never gotten around to.
How much do I need to pay beg you for a RMAA graph of the above EQ setting without the .5dB peak filter?
0.5 dB seems like a very small tweak for the extra CPU load, but I'm curious to see the difference.

soap, the 0.5 dB (in 20Hz!!!!!) is really insignificant. I just focused the graphical change. This was the best graph I got, before I gave up, tired and sleep. I posted this EQ setting, just in respect to the graph.

And if someone wants to use this EQ to compensate, forget about the 0.5 dB peak filter in 20Hz. There's absolutely no audible difference.
♪

[saratoga]

Theres some output resistance built into the Ipod, so adding resistance will change the frequency response everywhere, though probably only very, very slightly.  In general, unless you're seeing 2-3+ dB shifts, its basically ok to ignore them.

[Jamalarm]

I knew it! I was ranting about this about two months ago when i first switched to an ipod...

Thanks so much for those config compensation thingys.

[streetviper]

Hi! Has anyone done the same testing with a Cowon iaudio x5? I would like to compare the results!

[soap]

Hi! Has anyone done the same testing with a Cowon iaudio x5? I would like to compare the results!

Saratoga posted a link a while back to one of the few properly done RMAA tests I've seen on an iAudio.
Link
Note - when I say "proper" I mean one done under the load of headphones.  Most players test better with no load than with a load.
As you can see from the linked graph, the iAudio is not significantly different than the iPod in objective tests.

[LambdaCalculus]

It's disappointing to know that a company who's market share is ~80% wont make a quality DAP.
(yet another reason why I wish my ipod was a gigabeat )

Also makes me a teeny bit jealous of my girlfriend's iriver H320...

I think my next DAP will be either an iAUDIO X5 or a Gigabeat F60... just so long as I can Rockbox it!

[GodEater]

/me is now looking forward to the Gigabeat S port now that it's properly moving.

534MHz processor - oh yeah!

[psycho_maniac]

so since you found there is a problem can  this be fixed?