[Karendar]

  I've had the X5iii for about a month on stock rom, and seriously, this modded rom ROCKS  Could easily tell the SQ difference between this and 1.1.4, even with mid-range headphones without external amplifier.  I definitely don't want to go back to 1.1.4. just for the SQ.
 
However I found one thing on my system that might kind of forcing me back to stock 1.1.4.  Fresh clean cache-wiped and factory eased from recovery menu, flashed by the took kit as instructed, No further changes under android setting menu, WLAN left on (Wifi intelligent sleep) and I found my unit won't go to deep-sleep but was being held by kernal wakelock of "wlan_wake" 99% (reportred by BetterBatteryStats), stuck at 312MHz.   Then I turn off the WLAN, sometimes the deep-sleep would work, sometimes doesn't.  I just can replicate the pattern, still kind of random.  No further software installed or disabled.  The stock 1.1.4 would allow WLAN set on and still able to get to deep sleep without drawing much power, but now the constant drop on battery is pretty obvious and fast even not in use and screen off.
 
Am I the only one or Mr.WinX would you be able to shed some light on this.  As said the rom is great on SQ and I need the juice to run it.  Appreciate your work!

 
This is the reason I challenged WindowsX's idea for using an older kernel initially. 1.0.9 might have better sound quality, but usually at the expense of bugs. Final release was 1.1.1 for a reason: They felt stability could be improved further before final release. Not sure if they expected a sound difference in build 1.1.1 vs 1.0.9 though. 
 
Wifi + BT are notorious for for two things: Wakelocks and interference. What might have happened (And this is just assumption at this stage) is that Fiio might have fixed a wakelock by updating the drivers for the wireless chipset, which might also have caused added noise on the components, thus affecting overall SQ. 
 
If there's a way to bring back sound quality without compromising the wakelocks of previous kernels, it'll have to be done at the kernel code level by Fiio (Or a dev if they were to release source, which I doubt they will)
 
**edit** I just installed BBS to see if I get the same symptoms as you on 1.0.9 kernel... Will post findings.

 

[Cabrio]

Yes that's exactly what I am hoping for.  I have switched back and forth with the modded firmware and stock 1.1.4, just to find out if there is really a noticeable SQ difference, and sure there is, and in fact it is quite obvious.  I I wouldn't mind have the wlan off initally in order to keep it in a battery-friendly state, however, once I have it on once then turn it off again, the wlan wakelock won't go away and kept draining, that's the only problem of this kernel to my discovery.  

 

[Karendar]

  Yes that's exactly what I am hoping for.  I have switched back and forth with the modded firmware and stock 1.1.4, just to find out if there is really a noticeable SQ difference, and sure there is, and in fact it is quite obvious.  I I wouldn't mind have the wlan off initally in order to keep it in a battery-friendly state, however, once I have it on once then turn it off again, the wlan wakelock won't go away and kept draining, that's the only problem of this kernel to my discovery.  

Wifi is slower on 1.0.9 for me that's for sure... But that issue was also present on 1.1.1. Could not get over 8 mb/sec until I flashed 1.1.4...

 

[javre76]

   
Don't need screenshots. All you have to do is the following:
 
The URL for the full 1.1.1 ROM and flash tool: Unzip it in a folder somewhere
Unzip WindowsX's 1.1.4 modded ROM inside that folder (Overwrite the file if it asks you to)
Follow the instructions inside the folder's help document (All screenshots are there). It's detailed enough to follow step by step.
 
I recommend removing the SD cards if you have any before doing this operation.

Yes, 2nd that recommendation, remove your SD cards, when upgrading, it deleted all my music from my card (I had a backup and was able to restore it).

 

[Karendar]

  Yes that's exactly what I am hoping for.  I have switched back and forth with the modded firmware and stock 1.1.4, just to find out if there is really a noticeable SQ difference, and sure there is, and in fact it is quite obvious.  I I wouldn't mind have the wlan off initally in order to keep it in a battery-friendly state, however, once I have it on once then turn it off again, the wlan wakelock won't go away and kept draining, that's the only problem of this kernel to my discovery.  

I confirm: Huge wlan wakelock on 1.0.9... Disabling Wifi doesn't stop it, but disabling wifi and then rebooting does. Probably why this firmware wasn't production-ready. 

 

[WindowsX]

I seconded removing micro sd card. I also find someone failing to flash this properly with unreliable sd card causing freezing on boot. About flashing tool, you can also use DX200 Factory Tool too. from here.
 
http://www.ibasso.com/uploadfiles/download/FactoryTool_v1.39_DX200.zip
 
The instructions in DX200 batch tool instruction.doc file are detailed and easier to follow for new users.
 
Regards,
Keetakawee

 

[javre76]

  I confirm: Huge wlan wakelock on 1.0.9... Disabling Wifi doesn't stop it, but disabling wifi and then rebooting does. Probably why this firmware wasn't production-ready. 

Yeah, back to stock. Sound is good on stock as well, Windows X Custom is faster, but battery was poor. Will check these couple of days to see how it goes.

 

[enragedlemon]

 
Yeah, back to stock. Sound is good on stock as well, Windows X Custom is faster, but battery was poor. Will check these couple of days to see how it goes.

 
I think this underlines the core issue I have with this custom firmware; I have thus far seen no solid evidence that there is a difference between sound signature on 1.0.9 and 1.1.4. Everything that has been said so far points to a confirmation bias, seeing as nobody (except maybe the ROM author, who has another bias) has done an A/B test. The official response from FiiO is that they didn't find any difference themselves.
 
I have no doubt that this ROM can produce better performance by tweaking things like the CPU governor etc, but at the clear cost of battery life. Until sound signature differences have been confirmed, I would recommend limiting claims to the facts.

 

[LCMusicLover]

  ...I have switched back and forth with the modded firmware and stock 1.1.4, just to find out if there is really a noticeable SQ difference, and sure there is, and in fact it is quite obvious...

 

   
...I have thus far seen no solid evidence that there is a difference between sound signature on 1.0.9 and 1.1.4. Everything that has been said so far points to a confirmation bias, seeing as nobody (except maybe the ROM author, who has another bias) has done an A/B test...
 

 
I wonder if there is such a thing as Contradiction Bias?
 
Honestly, when I first installed the modded ROM the difference I heard was subtle enough that I probably could have been convinced that it was just a placebo effect.
 
When I installed the 1.0.9-based ROM, I immediately perceived a very un-subtle difference.
 
Could FiiO's official response be a case of my proposed Contradiction Bias?  Or a case of confirmation bias, confirming the proposition that there are no meaningful SQ differences between the kernels used in 1.0.9 & 1.1.4?
 
I'm willing to wait for more opinions to be expressed.
 
This kind of arguscussion is rampant on Head-Fi.  Burn-in and cable upgrades are just two examples.  Or compare Pinky's opinion of Opus vs X5iii to HBB's"
 
It really comes down to 'your ears, my ears', since nearly all of this stuff is subjective.  So, to my ears, WindowsX's mod delivered a significant sound improvement, and I thank him for it.
 
As far as battery life is concerned, bummer, but there are ways around the problem. I generally carry a 15,000 mAh Quick Charge battery with me anyway, so I'm unlikely to be affected

 

 

[enragedlemon]

   
 
I wonder if there is such a thing as Contradiction Bias?
 
Honestly, when I first installed the modded ROM the difference I heard was subtle enough that I probably could have been convinced that it was just a placebo effect.
 
When I installed the 1.0.9-based ROM, I immediately perceived a very un-subtle difference.
 
Could FiiO's official response be a case of my proposed Contradiction Bias?  Or a case of confirmation bias, confirming the proposition that there are no meaningful SQ differences between the kernels used in 1.0.9 & 1.1.4?
 
I'm willing to wait for more opinions to be expressed.
 
This kind of arguscussion is rampant on Head-Fi.  Burn-in and cable upgrades are just two examples.  Or compare Pinky's opinion of Opus vs X5iii to HBB's"
 
It really comes down to 'your ears, my ears', since nearly all of this stuff is subjective.  So, to my ears, WindowsX's mod delivered a significant sound improvement, and I thank him for it.
 
As far as battery life is concerned, bummer, but there are ways around the problem. I generally carry a 15,000 mAh Quick Charge battery with me anyway, so I'm unlikely to be affected

 

 
'Switched back and forth' is not an A/B test. There is no issue of contradiction bias, as you put it, because the burden of proof remains purely with this ROM. Consider the following points:

1. No hardware changes have been introduced, therefore the only changes could be software either a) changing the EQ or b) adjusting the output power. In the case of the former, just doing your own EQ would produce the same result. In theory, in the case of the latter, using something like the iFi IEMatch would produce the same result and this could be tested.
2. This modification requires flashing an unofficial ROM. In many jurisdictions, this permanently voids the warranty and is not recommended.
3. FiiO themselves have weighed in and said there is no appreciable difference. They designed the hardware and software and have been working with it a lot longer than any of us. You could argue that they are protecting their interests, but that doesn't change this fact.
4. There is a paid option for this ROM. I don't take issue with that directly, but we should be holding the developer to account for all claims made.

 
I am open to the idea that this ROM could produce a sound signature change. In fact, this would be a good thing because it provides users with more choice and that's a big win for a free market. However, without more than just anecdotal evidence, especially seeing as logic would suggest there shouldn't be any difference, the risks outweigh the rewards. All I am doing is providing a sensible counterpoint to unsubstantiated claims; but consider me a convert if we have some conclusive proof that there is a difference (disregarding whether people believe it sounds 'better').
 
Edit: with regards to things like burn-in, whether or not it produces a change to the sound, you're not materially altering the equipment outside of it's designed usage. Although I personally believe that voiding warranties due to software changes is ridiculous, it is nevertheless a reality. Burn in due to normal (or even abnormal so long as it isn't outside of spec) listening won't void a warranty ever. There's a big difference between that and flashing an unofficial ROM.

 

[LCMusicLover]

   
'Switched back and forth' is not an A/B test. There is no issue of contradiction bias, as you put it, because the burden of proof remains purely with this ROM. Consider the following points:

1. No hardware changes have been introduced, therefore the only changes could be software either a) changing the EQ or b) adjusting the output power. In the case of the former, just doing your own EQ would produce the same result. In theory, in the case of the latter, using something like the iFi IEMatch would produce the same result and this could be tested.
2. This modification requires flashing an unofficial ROM. In many jurisdictions, this permanently voids the warranty and is not recommended.
3. FiiO themselves have weighed in and said there is no appreciable difference. They designed the hardware and software and have been working with it a lot longer than any of us. You could argue that they are protecting their interests, but that doesn't change this fact.
4. There is a paid option for this ROM. I don't take issue with that directly, but we should be holding the developer to account for all claims made.

 
I am open to the idea that this ROM could produce a sound signature change. In fact, this would be a good thing because it provides users with more choice and that's a big win for a free market. However, without more than just anecdotal evidence, especially seeing as logic would suggest there shouldn't be any difference, the risks outweigh the rewards. All I am doing is providing a sensible counterpoint to unsubstantiated claims; but consider me a convert if we have some conclusive proof that there is a difference (disregarding whether people believe it sounds 'better').
 
Edit: with regards to things like burn-in, whether or not it produces a change to the sound, you're not materially altering the equipment outside of it's designed usage. Although I personally believe that voiding warranties due to software changes is ridiculous, it is nevertheless a reality. Burn in due to normal (or even abnormal so long as it isn't outside of spec) listening won't void a warranty ever. There's a big difference between that and flashing an unofficial ROM.

 
Just one point, and then I'll shut up 

 
Whether flashing the ROM voids the warranty is completely irrelevant when debating the possibility that the SQ is changed/improved.  (Although, I get that it is a concern when considering whether or not to install the mod.)
 
Oops, 2 points.  
 
In your arguments you state that since the hardware isn't changed, logically there can be no difference in the sound except EQ and output power adjustment.  I don't know how you can know this (that is, that these are the only elements of audio output that can be affected by software). 
 
Actually, 3 points (but that's all, I swear)
 
I read FiiO's reply which you state '...settles that'. It could easily be considered an 'All the President's Men' non-denial denial. They say that their engineers 'test the parameters' of the firmware, and that 'nothing special are changed in that firmware'. But it doesn't say that nothing is changed in the firmware, and it certainly doesn't say that they A/B tested the mod against stock. Even if someone bought or borrowed second X5 to A/B it, wouldn't that person still be saying 'Yes, I did an A/B comparison with careful volume matching, and the modded ROM sounds better'? So how can we get to your 'conclusive proof' that it IS better. Bottom line is that I don't think we can. So let's agree to disagree. I think my DAP sounds better with the modded ROM, and you think I'm full of equine manure. I can live with that.
 
OK, so I cheated, and there were actually 2 points in that last paragraph.  Sorry, I really will shut up now, on this subject, not in general 

 

[enragedlemon]

   
Just one point, and then I'll shut up 

 
Whether flashing the ROM voids the warranty is completely irrelevant when debating the possibility that the SQ is changed/improved.  (Although, I get that it is a concern when considering whether or not to install the mod.)
 
Oops, 2 points.  
 
In your arguments you state that since the hardware isn't changed, logically there can be no difference in the sound except EQ and output power adjustment.  I don't know how you can know this (that is, that these are the only elements of audio output that can be affected by software). 
 
Actually, 3 points (but that's all, I swear)
 
I read FiiO's reply which you state '...settles that'. It could easily be considered an 'All the President's Men' non-denial denial. They say that their engineers 'test the parameters' of the firmware, and that 'nothing special are changed in that firmware'. But it doesn't say that nothing is changed in the firmware, and it certainly doesn't say that they A/B tested the mod against stock. Even if someone bought or borrowed second X5 to A/B it, wouldn't that person still be saying 'Yes, I did an A/B comparison with careful volume matching, and the modded ROM sounds better'? So how can we get to your 'conclusive proof' that it IS better. Bottom line is that I don't think we can. So let's agree to disagree. I think my DAP sounds better with the modded ROM, and you think I'm full of equine manure. I can live with that.
 
OK, so I cheated, and there were actually 2 points in that last paragraph.  Sorry, I really will shut up now, on this subject, not in general 

 
I don't think you should shut up on this topic. You're not wrong, and from the point of view of preference for sound your view is as valid as any other. Let me just reiterate though that my point isn't about whether or not the sound is better (because that is entirely subjective and open to endless debate), but rather if it is different at all. That can be easily verified by a blind A/B test though it challenging to setup. The ideal would be to actually produce response curves of the two so we can see side-by-side.
 
With regards to the software aspect, I may have overlooked some detail, but essentially 'what' and 'how much' are the two things that can be controlled. The base signature of both the amp and the DAC are fixed and cannot be changed outside of those two instructions (with the exception being the DAC filters which are baked into the chip). Therefore what remains it to change the bitstream before it hits the DAC and/or to adjust the output power of the amp. Firmware may potentially have more control, but seeing as the chips are proprietary, this would be at the whim of their manufacturers.
 
As for what FiiO said, I do admit that I took a liberal approach to interpreting their response in the light of English not being the first language of the respondent. I too would like some more information on what kind of a review they did... Actually, more information and fair comparisons is what I want for this ROM in general!

 

[LCMusicLover]

   
...The ideal would be to actually produce response curves of the two so we can see side-by-side.
 

 
OK, not quite shutting up 

 
For me, the change in tonality was the subtlest change in the sound.  For that reason, I think side-by-side response curves would miss the meaningful changes I perceive -- specifically, more detail retrieval, and more resolved imaging/sound stage. Can differences like that be seen in response curves?

 

[enragedlemon]

   
OK, not quite shutting up 

 
For me, the change in tonality was the subtlest change in the sound.  For that reason, I think side-by-side response curves would miss the meaningful changes I perceive -- specifically, more detail retrieval, and more resolved imaging/sound stage. Can differences like that be seen in response curves?

 
Yes it should, especially if you isolate and test the left and right channels separately.

 

[dukeman]

I hear a difference in sound between hibby and neutron/uapp when listening to the same note4/opus 11 combo, just because the hardware is the same doesn't mean there can't be differences.