[nmatheis]

Gents,

I tested out the EQ to try and see what levels the sliders would have to be in order to match the volume of the OFF eq.......it's impossible.

Everything between 31 and 8kHz can be matched...but before 31 and after 8kHz is impossible to replicate

Besides the sub-bass and treble roll-off, the EQ-ed FR curve looks pretty smooth. I know some were worried about dips / spikes across the whole range. Doesn't look warranted from these measurements - just the roll-off...

 

[JaZZ]

  Gents,
 
I tested out the EQ to try and see what levels the sliders would have to be in order to match the volume of the OFF eq.......it's impossible.
 
Everything between 31 and 8kHz can be matched...but before 31 and after 8kHz is impossible to replicate
 

Theo, why isn't your reference curve a straight line – since it's presented as X5 frequency response with EQ off? Instead it has a massive bass boost and a slight midrange dip. And how come that nobody cares?
 
BTW: How exactly were the slider positions to reach this quasi-congruence?

 

[uzi2]

Nobody cares, because they listen to music rather than look at it...

 

[JaZZ]

Nobody cares, because they listen to music rather than look at it...

 
You seem to think EQs don't matter. Well, to me they do.

 

[nmatheis]

Theo, why isn't your reference curve a straight line – since it's presented as X5 frequency response with EQ off? Instead it has a massive bass boost and a slight midrange dip. And how come that nobody cares?

BTW: How exactly were the slider positions to reach this quasi-congruence?

Good question.

 

[Brooko]

More importantly - why are the measurements by Theo so much different to Fiio's
 
http://www.fiio.net/en/products/2/parameters
 

  Frequency Response diagram of curves.

Test condition：play- 5db sinusoidal signal， headphone output

Red curve: Left channel , Blue curve: Right channel

Frequency Response diagram of curves.

Test condition：play- 5db sinusoidal signal，Line output

Red curve: Left channel , Blue curve: Right channel

 
FWIW - in Theo's tests, if sub-bass was showing a steady rise of up to 9dB from flat, I'm pretty sure someone would be noticing it by now?
 
Also Jazz - on Theo's graph, the mid-range dip you are talking about is less than 1dB - so chances are you wouldn't hear it.
 
Again though - I'd be questioning the huge discrepancy between Fiio's measurements and Theo's.  Something isn't right .....

 

[uzi2]

   
You seem to think EQs don't matter. Well, to me they do.

I would hope that you adjust your EQ by what you're hearing rather than the graph you're looking at.

 

[514077]

Sorry, Theos
S.  I was just curious.  I studdied recording years ago, and generally, we tried everything from mic choice to placement.  We kept EQing to a minimum so as not to introduce to much out of phase sound.Didn't mean to get you on the hot seat.

 

[JaZZ]

  I would hope that you adjust your EQ by what you're hearing rather than the graph you're looking at.

 
Why this? That's entirely my business. Moreover, I'm always interested in data and graphs to compare them with what I hear.

 

[Dobrescu George]

I think that the reason for theos differences are that he did not play a sin waves, but pink noise through x5. 
 
at the basis of the problem, if the pink noise has not been a flat sine-like line, probably this is why it looks like this.
 
i am wondering if the noise he played is also responsible for the treble rolloff? also, i am wondering why i never noticed this while playing music.
 
Also, @TheoS53 i would have a request, if possible. I do not own the equipment, nor the knowledge to test this, but if it would not take too much time, i am curious if there are differences between low and high gain in distortion graphs, and in freq response graphs.

 

[TheoS53]

Gents,
 
My apologies, I had the wrong calibration file loaded when I did those tests. Busy redoing them now

 

[TheoS53]

ok, after i fixed the calibration issue, the results look spot on
 
 
This is the measurement at a 1dB increment scale (same as the one posted by Fiio):
 

 
 
This is zoomed in to a 0.1dB increment scale:
 

 
 
Red is the OFF eq, and black is the eq I did to try and match the OFF eq as close as possible:

 
 
Same graph as above, but zoomed out to a 1dB increment scale:

 
The settings for this eq was:
 
31 : 2 clicks below +6
62 - 8k : 3 clicks above +4
16k : all the way to the top

 

[Brooko]

Thanks Theo
 
So it's essentially flat with no EQ, and rolls off at 30 Hz (steep) and 8-10kHz (slow) under EQ.
 
Given that there isn't a lot of audible sub-bass in most music below 30Hz, and the roll off is very slow (well under 0.5dB from 10-15kHz), and roll-off is going to be pretty much inaudible anyway (unless you're the type with golden ears who listens mostly to test tones 

 ) - then I see nothing to really worry about

 

[nmatheis]

ok, after i fixed the calibration issue, the results look spot on


This is the measurement at a 1dB increment scale (same as the one posted by Fiio):





This is zoomed in to a 0.1dB increment scale:





Red is the OFF eq, and black is the eq I did to try and match the OFF eq as close as possible:




Same graph as above, but zoomed out to a 1dB increment scale:



The settings for this eq was:

31 : 2 clicks below +6
62 - 8k : 3 clicks above +4
16k : all the way to the top

Thanks Theo! So roll-off but no major dips/spikes in middle of FR.

 

[JaZZ]

Thanks for this reissue – very informative! So the single bands seem to fit perfectly to each other, which means they have a bandwidth of exactly one octave. That's a good thing, as it prevents the creation of unwanted ripples in the frequency response.
 
Of course it would be rather strange to use the above settings just for mimicking the louder no-EQ setting. It would offer no sonic benefit.