[goodvibes]

  umm...I'm reading that in iOS works alac 24/48 too....

Perhaps they do now but it may be resampling. Things may have changed by now but when using the dig out via wadia, the most I could get to register out of a pre 5 touch was 24/44. 

 

[goodvibes]

Interesting. I'm pretty sure all the new Cirrus Logic chipsets are capable of at least 24/96, last I checked. Would there be OS level resampling going on though?

I haven't tried it but my understanding is that you get full res via usb but limited to 88 via the headphone jack. Might have something to do with the dig control.
https://itunes.apple.com/us/app/onkyo-hf-player/id704139896?mt=8

 

[_js_]

So that thread about the sound differences between iOS 6 and 7 had someone who was using FLAC files, so my theory is wrong.  So that must mean that the internal DAC in iDevices has firmware?  And that firmware can change and change the sound, and yet leave the FR ruler flat, still?  That just doesn't make any sense to me.  None of it.  All of it.  How could iOS 7 introduce extra bass and yet leave the FR curve the same?  And if it DOES change the FR then we should be able to graph it and see it.  There are a lot of questions here.

 

[vwinter]

I haven't tried it but my understanding is that you get full res via usb but limited to 88 via the headphone jack. Might have something to do with the dig control.
https://itunes.apple.com/us/app/onkyo-hf-player/id704139896?mt=8

I'll give it a shot. $9.99 to not have to convert FLAC files to ALAC... Hmmm.

The main bummer is not being able to amp straight from a line-out with the 5 and up. You'd need an an Amp/DAC through the digital USB out, but then your phone DAC is useless.

So that thread about the sound differences between iOS 6 and 7 had someone who was using FLAC files, so my theory is wrong.  So that must mean that the internal DAC in iDevices has firmware?  And that firmware can change and change the sound, and yet leave the FR ruler flat, still?  That just doesn't make any sense to me.  None of it.  All of it.  How could iOS 7 introduce extra bass and yet leave the FR curve the same?  And if it DOES change the FR then we should be able to graph it and see it.  There are a lot of questions here.

I think these could be a few factors at play here. I didn't read through the whole thread, but (1) I'm pretty sure most people aren't comparing side by side the same device with iOS 6 and 7, so it's from memory, which is fine but should be noted, and (2) that a change in the presentation of the sound, for example a tighter more focused bass could give the impression of less or more bass, or more separation in the midrange could make it sound more distant. Just possible explanations.

 

[_js_]

I'll give it a shot. $9.99 to not have to convert FLAC files to ALAC... Hmmm.

The main bummer is not being able to amp straight from a line-out with the 5 and up. You'd need an an Amp/DAC through the digital USB out, but then your phone DAC is useless.
I think these could be a few factors at play here. I didn't read through the whole thread, but (1) I'm pretty sure most people aren't comparing side by side the same device with iOS 6 and 7, so it's from memory, which is fine but should be noted, and (2) that a change in the presentation of the sound, for example a tighter more focused bass could give the impression of less or more bass, or more separation in the midrange could make it sound more distant. Just possible explanations.

 
vwinter,
 
I read that the DAC in the lightning to 30 pin adaptor is on a par with the built in DACs (it's a Wolfson, IIRC) so you could always do 30 pin adaptor-->LOD-->amp.  I'm sure you knew this already but I (a) wanted to hear your thoughts/impressions, and (b) wanted to mention it in case others didn't know it.  It's a pretty common question, I think.  I know I was wondering about it before/during my upgrade from iP4 to iP5S.
 
As for players, someone earlier mentioned Tuneshell and I just looked it up on iTunes.  Got tons of 5 star ratings and it is only $5.99 for the full version and there is a free version.  The person who mentioned it earlier said it could even output hi-res audio (higher than the DAC can handle or iOS will allow, anyway) through the lightning digital out of a 5/5C/5S.  So if you DID have a DAC/Amp combo, you'd be set.
 
And, I just found out that JDS Labs is now offering a C5D!  Which is a DAC/Amp combo, with the amp section just being the C5--only slightly better (lower output impedance-- 0.6 ohms or something instead of 2.2) and it will work, native digital stream out via UAC1 protocol, with an iPhone running iOS7 via the Camera Connection Kit.  Only issue for me is the idea of two cables.  iPhone-->CCK-->USB to Micro-USB-->C5D.  Kind of annoying.  No reason you couldn't make (or a company couldn't make) a CCK with a male micro-USB termination instead of a female USB-A socket.  I'm looking forward to reviews of the C5D.  In particular, I'm hoping that the volume bar disappears in iOS7 just like with the CEntrance HiFi M8 and Sony PHA-1.
 
Back to the iPhone 5S audio output, I'm really impressed with the SQ of my Senn HD-595's straight out of the iP5S.  Really quite good, considering.  And the Ety ER-4S is significantly better.  I'm so happy with those Ety's!  And I'm really pleased that they also work quite well straight out of the 5S, no outboard amp needed.  Still doesn't stop me from being very interested in one of the DAC/Amp combos that will work with the straight digital out of the iPhone Lightning connector, though!  LOL!
 
Finally, I haven't soldered up my test jig for the OI measurements.  My schedule (graveyard shifts) and the nature of the shift work this rotation hasn't really allowed it.  But I'm going to be off shift and back to days starting Tuesday and will get things setup then.
 
I'm also going to measure the max voltage output into various resistances of both the 5S and 4 to see if it can shed some light on the question of which can get "louder".  Sort of a tricky issue, since loudness has to do with power, but the amp is providing voltage, but which max voltage can and does change depending on the load impedance.  Despite the higher OI of the 5S, I'm guessing it's got higher gain and more power output vs. the 4 in many cases, especially with higher impedance headphones.  In fact, it very well may have been a design trade-off for output impedance.
 
Or, it could all just be accidental and related only to lowest bidder for given specs and codec size for each new iPhone generation going forward.  Still, accidental or not, I'll take it!  Now that I've got some synergistic pairings, and have had some time to do extended listening, I can easily say that my new setups have higher SQ than my old one.  To my ears, anyway!

 

[_js_]

OK!  The results are in!
 
I soldered up an 1/8" TRS plug with leads terminating in alligator clips.  I checked to make sure all connections were good and that there were open circuits between any combination of tip / ring / sleeve.  Next I got four 1/4 watt metal film resistors.  I had 99.4 ohms, 62.0 ohms, 32.4 ohms, and 18 ohms.  I started with the iPhone 4 since we know it's output impedance and because if I screwed something up and damaged it, my iPhone 5S would still be fine.  I also decided that I didn't need an oscilloscope since I was using a signal generator app to output a pure sinewave at 440 Hz, so a Fluke DMM would work fine to measure AC RMS volts.  Initially I used a Fluke 12B I have at my office.  Here are the raw results:
 
99.4 ohm: .174 V
62.0 ohm: .173 V
32.4 ohm: .171 V
18.0 ohm: .167 V
open circuit (= Vin): .177 V
 
Unfortunately, depending on which data point above I used, I was coming up with anywhere from .96 ohms to .85 ohms.  I also used the multiple data points to solve for R1 without assuming Vin was equal to Voc (two equations, two unknowns, in other words).  But, it quickly dawned on me that I didn't have enough significant digits on the voltage measurement.  I also hadn't turned up the volume all the way!
 
Once home, I used my Fluke 179 True RMS Meter, and for simplicity I measured Voc and only the data point for the 99.4 ohm resistor.  Oh, I should mention that the resistance I measured was through the 1/8" plug, and so included all the resistance that the iPhone would see.
 
Anyway, here are the results I got for the iPhone 4 at maximum volume:
 
Vin: .878 V
Vout: .869
 
If you solve the circuit equations for R1 (output resistance) you get the following:
 
R1 = R2 * ( (Vin - Vout) / Vout )
 
Plugging in the values (R2 is 99.4 ohms, from above), you get:
 
R1 = 1.03 ohms.  This is almost exactly what the figure is from the internet, if I remember correctly.  Maybe it was 1.02 ohms or something.  But quite close to my result.
 
And, by the way, you also get a total output power into the 99.4 ohm load of 7.6 mW.
 
Now for the iPhone 5S (I checked and I hadn't damaged the iPhone 4, so all is a go for the 5S)!
 
Vin: .929 V
Vout: .908 V
 
Calculating, we get . . . drum roll please . . .
 
R1 = 2.30 ohms.
 
And, the total power into the load is 8.3 mW.
 
So, we know a few things.  The iPhone 5S is more or less exactly as loud (.383 dB louder, to be precise) as the iPhone 4 into this load, and the iPhone 5S has a respectable (but not stellar) output impedance!  I mean, the JDS Labs C5 has an output impedance of 2.2 ohms.  Of course, it will put 3.3 volts into 150 ohms . . . so, it's much more powerful.  But still.  This result is lower than I would have guessed (than I and others here did guess, that is to say).
 
Not bad.  Not bad at all.

 

[vwinter]

Awesome! Thanks for the hard work!

I was off by .2 in my guess of 2.5 Ohm. I was also right in just a hair louder than the 4S. I feel good. I'm gonna buy myself some dumplings.

You should get some dumplings too. You earned them. Hell if you were here, my tab.

 

[elfary]

  OK!  The results are in!
 
I soldered up an 1/8" TRS plug with leads terminating in alligator clips.  I checked to make sure all connections were good and that there were open circuits between any combination of tip / ring / sleeve.  Next I got four 1/4 watt metal film resistors.  I had 99.4 ohms, 62.0 ohms, 32.4 ohms, and 18 ohms.  I started with the iPhone 4 since we know it's output impedance and because if I screwed something up and damaged it, my iPhone 5S would still be fine.  I also decided that I didn't need an oscilloscope since I was using a signal generator app to output a pure sinewave at 440 Hz, so a Fluke DMM would work fine to measure AC RMS volts.  Initially I used a Fluke 12B I have at my office.  Here are the raw results:
 
99.4 ohm: .174 V
62.0 ohm: .173 V
32.4 ohm: .171 V
18.0 ohm: .167 V
open circuit (= Vin): .177 V
 
Unfortunately, depending on which data point above I used, I was coming up with anywhere from .96 ohms to .85 ohms.  I also used the multiple data points to solve for R1 without assuming Vin was equal to Voc (two equations, two unknowns, in other words).  But, it quickly dawned on me that I didn't have enough significant digits on the voltage measurement.  I also hadn't turned up the volume all the way!
 
Once home, I used my Fluke 179 True RMS Meter, and for simplicity I measured Voc and only the data point for the 99.4 ohm resistor.  Oh, I should mention that the resistance I measured was through the 1/8" plug, and so included all the resistance that the iPhone would see.
 
Anyway, here are the results I got for the iPhone 4 at maximum volume:
 
Vin: .878 V
Vout: .869
 
If you solve the circuit equations for R1 (output resistance) you get the following:
 
R1 = R2 * ( (Vin - Vout) / Vout )
 
Plugging in the values (R2 is 99.4 ohms, from above), you get:
 
R1 = 1.03 ohms.  This is almost exactly what the figure is from the internet, if I remember correctly.  Maybe it was 1.02 ohms or something.  But quite close to my result.
 
And, by the way, you also get a total output power into the 99.4 ohm load of 7.6 mW.
 
Now for the iPhone 5S (I checked and I hadn't damaged the iPhone 4, so all is a go for the 5S)!
 
Vin: .929 V
Vout: .908 V
 
Calculating, we get . . . drum roll please . . .
 
R1 = 2.30 ohms.
 
And, the total power into the load is 8.3 mW.
 
So, we know a few things.  The iPhone 5S is more or less exactly as loud (.383 dB louder, to be precise) as the iPhone 4 into this load, and the iPhone 5S has a respectable (but not stellar) output impedance!  I mean, the JDS Labs C5 has an output impedance of 2.2 ohms.  Of course, it will put 3.3 volts into 150 ohms . . . so, it's much more powerful.  But still.  This result is lower than I would have guessed (than I and others here did guess, that is to say).
 
Not bad.  Not bad at all.

Thanks a lot for the testing. Much appreciated.
 
Nice to know that 5S has bested the iPhone 5 in the output impedance and output power areas and is somewhat back on track with the (still) great iPhone 4S.
 
Now my sights are set on an iPhone 5S 64Gb.

 

[ILoveGrado]

Does anyone know what the closest Cirrus Logic CODEC and Class D Amplifier these Apple branded chips might be? My guess will be CS42L73 for the CODEC and CS35L0x for the Class D Amplifier. If it's true then the Class D Amp is only being used for the built in speakers. The Headphone Amp is still integrated within the CODEC chip.

 

[ILoveGrado]

Oh wait, no it's not the built in speakers. The Class D amplifier is to act as a output stage for the line-out.

 

[vwinter]

Line-out?

Besides the headphone out, the lightning port is all digital... no?

 

[ILoveGrado]

Oh my, yes you are right. Well, back to the diagram for now.

 

[ILoveGrado]

In the Diagram for CS42L73, you will see that the line output of this codec connects to external class D amplifier. Apple might use this amplifier not for line-out but for driving the tiny built in spkrs.

 

[vwinter]

In the Diagram for CS42L73, you will see that the line output of this codec connects to external class D amplifier. Apple might use this amplifier not for line-out but for driving the tiny built in spkrs.

The only problem is that app makers seem to be saying that the chip is capable of outputting up to 88khz sample rates through the headphone out. And the codec you suspect is only capable of 48khz in general. Though that's far from official and I'm a bit skeptical.

If you don't mind me asking, what's your interest in the chipset? Solid detective work btw.

 

[_js_]

The Cirrus Logic Audio Codec is p/n 338S1201.  338S1202 is the p/n for the Class D audio amplifier:  http://www.chipworks.com/en/technical-competitive-analysis/resources/blog/inside-the-iphone-5s/
 

New Power Management IC and a new audio codec and class D amplifier
As we mentioned earlier, we’re seeing a lot of familiar components inside the iPhone 5s. The major standouts being the new A7 processor and two new MEMS devices. We are also seeing a new power management IC by Dialog Semiconductor and a new audio codec and class D amplifier by Cirrus Logic. The Power Management IC by Dialog Semiconductor has the part number 338S1216-A2. The new parts by Cirrus are numbered as 338S1202 for the Class D audio amplifier and 338S1201 for the Audio Codec. Note that the audio codec is ~30% smaller than the codec inside the iPhone 5.

 
And I would be very surprised if there were a different amplifier for the built in speakers than for the headphone out.  This is why you can't have both sounding at once.