[LinkyG]

I mainly listen to the music from my DX50 via the headphone out socket using Shure 215 in ear headphones. I have upgraded the firmware since 1.2.2 and never had an issue with any of them until 1.2.7
It installed fine, and the fixes were good, but the sound was many steps back. The previously posted 'Muddy' and 'Narrow' would adequately sum up my impression of it.
I tried it for 48 hours, but just couldn't get into it, so a quick step backwards and my musical goodness has returned.
So sorry iBasso, another vote on here against 1.2.7 - combine its features with an earlier soundstage and you may have a winner!

 

[spurxiii]

No the DX50 is fine as it is for me, don't change. Its already a winner.

 

[Varoudis]

dx50 is great but the EQ is for sure coded really wrong!!!
 
I canNOT match ON OFF with any "flat" combination of sliders (all 6 or all 7)... volume seems right but the low end is lost with (sub bands) the EQ ON.
Even the "Classic" which has a heavy bass boost (on paper/EQ curve) has less sub bass than the EQ OFF :S
 
I have coded a simple EQ in C++ many years ago and I believe that the corner bands have some anchors low down (-12???) outside the UIs frequencies. They are not coded as 'shelf' EQs.
 
Am I wrong?

 

[xnuthecaveman]

dx50 is great but the EQ is for sure coded really wrong!!!

I canNOT match ON OFF with any "flat" combination of sliders (all 6 or all 7)... volume seems right but the low end is lost with (sub bands) the EQ ON.
Even the "Classic" which has a heavy bass boost (on paper/EQ curve) has less sub bass than the EQ OFF :S

I have coded a simple EQ in C++ many years ago and I believe that the corner bands have some anchors low down (-12???) outside the UIs frequencies. They are not coded as 'shelf' EQs.

Am I wrong?

This is highly probable, since the EQ of 1.2.7 was recoded. Perhaps ibasso has not yet perfected the programming on the EQ. Even with the DX100, the EQ is pretty much useless.

Maybe you could shoot them an email with your findings

 

[Varoudis]

  dx50 is great but the EQ is for sure coded really wrong!!!
 
I canNOT match ON OFF with any "flat" combination of sliders (all 6 or all 7)... volume seems right but the low end is lost with (sub bands) the EQ ON.
Even the "Classic" which has a heavy bass boost (on paper/EQ curve) has less sub bass than the EQ OFF :S
 
I have coded a simple EQ in C++ many years ago and I believe that the corner bands have some anchors low down (-12???) outside the UIs frequencies. They are not coded as 'shelf' EQs.
 
Am I wrong?

A bit of visual analysis (the images are just for description, not exact replication of the curves on DX50)
 
I believe the EQ is implemented like this: (anchors near the limits of the bands)
 

 
or with a low cut band:

 
but it should be something like:
 

 
'shelf eq'.
 
And the unison gain is also a thing!!

 

[Michael Jennings]

I have quite a bit of new gear on the way: a DX50, a JDS Labs O2, and a pair of Noble 4 universals. I hope they play well together! I'll report when it's all in place and I've had some time to settle in.

 

[DrRuss]

  I have quite a bit of new gear on the way: a DX50, a JDS Labs O2, and a pair of Noble 4 universals. I hope they play well together! I'll report when it's all in place and I've had some time to settle in.

great gear choice man! enjoy!

 

[DrRuss]

  A bit of visual analysis (the images are just for description, not exact replication of the curves on DX50)
 
I believe the EQ is implemented like this: (anchors near the limits of the bands)
 

 
or with a low cut band:

 
but it should be something like:
 

 
'shelf eq'.
 
And the unison gain is also a thing!!

i believe so too man! great visual analysis!

 

[MCImes]

+1 for 1.2.2 sound with 1.2.7 bug fixes.

Or Ideally, let us choose which decoding library we want to use. Then everyone is happy!

 

[alphabetagaga]

   
This is normal.  The mp3 format does not support gapless natively.  An mp3 must be encoded with a specific codec that supports gapless, and it must be played back by a player with a corresponding plugin that supports gapless. 
 
Either your mp3s were not encoded with a gapless codec, or the DX50 does not support gapless playback of that particular codec. 

 
thank you. so whilst that may be the definitive explanation, it does not help us 'average joes' who simply want to rip the gapless albums they own the using appropriate software and then look for a dap that *supports gapless. i rip all mine using dbpoweramp set to mp3 (lame) 320 cbr. the gapless albums play without any issue whatsoever on a sansa clip+ with rockbox (gapless supported) and also an ipod nano. rockbox says it supports gapless. so do apple. and on the dx50 you can turn gapless 'on' or 'off'. but only on one of the 3 products, does gapless work correctly.
 
gapless is still not fixed fully on the dx50.

 

[Varoudis]

  i believe so too man! great visual analysis!

I received a very good and fast reply from iBasso ( I have to admit Im very happy with they helpful attitude! A+++ )
 
==============
HI,
Thank you for your email.
I dont think there is any bugs on the EQ setting.  Please see the attachment.
I turned on the EQ, and set the 30Hz to +12 position, then I got the attached screenshot.  That prove the EQ works correctly.
After you turn on the EQ, the volume will drop down.  You need to raise up the volume a little.
 
Sincerely
iBasso Audio
 
 

=============
 
So the reply seems correct. -6.5db is the volume lowered with eq ON.
+12 clicks seems to also be +6db! Nice.
 
I don't know why my initial tests seems to give different results but I think its was my brains perception of volume diff and 'eq technology overanalyses'
 
PS: i bet this is the first time in this forum that someone accepts that the problem he was hearing was actually neural network malfunction

 

[GoldfishX]

  Ugh, 1.2.2 sounds the best...I want to shoot myself in the head instead of dealing with this buggy firmware, but that might damage my hearing. I'm pretty sure I skipped this one initially. The price one must pay for audiophilia...
 
Best vocals by far, a tad warm, but very large soundstage and extremely detailed. Lacking the midbass hump of 1.2.5, but has a lot more...hmm, soul(?) than 1.2.6 and despite the warmness, is not muddy like I found 1.2.7 to be. I can really crank 1.2.2 because it's so relaxed. Euphonic is what I would use to describe it, even though it isn't quite perfect. I will try it tomorrow on my Magni at my office setup, maybe the warmness comes from the Lyr. Magni has issues with harsh vocals/treble, I am curious to see how the firmware change affects that issue. I suspect a more neutral amp will help things out here. The sound is warm, not muddy, if that makes sense. Songs with backup singers are really well-done on this one.
 
So that means, excluding 1.2.3beta, my favorite firmwares have gone in reverse order. 1.2.2 being the best (for reasons above), 1.2.5 adding a midbass hump (that doesn't bleed into the mids), but also worse vocals and some sibilance (minor quibbles), 1.2.6 being clear and analytical, but lacking the soul of the previous two and I still haven't bothered with 1.2.7 since my initial sour impressions and because I have been enjoying 1.2.2 too much.
 
Again, settings are as followed: Coax out into Peachtree DacIt into Lyr (Siemens tubes) with Mad Dogs. EQ has 30hz band minimized, everything else in the middle.
 
Ibasso, if you are reading this...PLEASE COMBINE 1.2.2 SOUND WITH 1.2.7 FUNCTIONALITY FOR THE NEXT FIRMWARE! THX!

 
My theory is correct...The vocals are more forward on the Magni, but not harsh at all. Very clear, very airy. The warmth is there, but the last of the mud is gone in 1.2.2. The bass has more definition overall and setup seems to change from recording to recording more. Very, very impressive. I need to see what happens when I add an Alpha Dog to the equation.
 
Here is what this is like. Imagine being in a room with music playing onstage. If you listen carefully, you can make out everything that is happening. Vocals, notes, etc. But if you want to focus on something else (like a TV or magazine or something), the music kind of plays casually in the background.

 

[Wapiti]

  PS: i bet this is the first time in this forum that someone accepts that the problem he was hearing was actually neural network malfunction

Indeed.  
 
Thanks for posting the response and graph.  Very interesting.  The wide Q is appropriate.

 

[highfell]

I've seen a lot of positive reports about the D12 connected to the DX50 via Coax, so be sure to try that as well   

Indeed and with 2 WM8740 24Bit DAC Chips rather than the Dx50's one, shouldn't it "automatically sound better", than the Dx50?

It is certainly one of the things I would like to review :bigsmile_face:

 

[Varoudis]

  Indeed.  
 
Thanks for posting the response and graph.  Very interesting.  The wide Q is appropriate.

yes its very good! As it should be. great iBasso.
 
the problem for me was the volume difference which is also the case for many 'quality-persetpion' problems, with for example mastering tracks. (the louder the better - syndrome)