• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Measurements of FIIO X1 ii

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
74
#3
This is curious:

View attachment 11870

Why this narrower range than 20-20 kHz?

That is the author's choice. However it makes some sense because not only does audibility go down at the frequency extremes, but so does our sensitivity to variations.

The illustration shows this fairly well known psychoacoustic effect, The broader the frequency response error, and the closer it is to the midrange, the less the ear tolerates it,

The attachment shows how wide and at what frequency a response variation has to be to be in danger of being audible, using a fairly sensitive risk adverse standard.

abx level matching criteria.png


The best recent AES paper relating to this is F. E. Toole, S. R. Olive: The Modification of Timbre by Resonances: Perception and Measurement, JAES, vol. 32, March, 1988, 122-142. Also online for everybody: DETECTION OF AUDIBLE RESONANCES Ivo Mateljan, Heinrich Weber*, Ante Doric** http://www.artalabs.hr/papers/im-aaaa2007.pdf
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,566
Likes
25,243
Location
Seattle Area
#4
That is the author's choice. However it makes some sense because not only does audibility go down at the frequency extremes, but so does our sensitivity to variations.
Well, the idea of measurements is that we get all the data and then apply psychoacoustics to it. If you don't have any response measured above 15 kHz, how do you determine if it is or isn't audible?

Also response below 40 Hz is useful to know if the device is DC coupled or capacitor.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,566
Likes
25,243
Location
Seattle Area
#5
On your enclosure which is from Clark's paper, there is important text that goes with that figure:

upload_2018-4-4_19-56-41.png



So while directionally that may be right, I don't think it provides justification for not measuring what is going on outside of 40-15kHz.

On Dr. Toole/Olive's paper, I have read that many times. Which specific portion do you think applies to this?
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
74
#6
On your enclosure which is from Clark's paper, there is important text that goes with that figure:

View attachment 11872


So while directionally that may be right, I don't think it provides justification for not measuring what is going on outside of 40-15kHz.
I agree, and so I also post a traditional frequency response curve covering a much wider range. I regret that the previous PDF file was missing some graphics, but that should be corrected now.

On Dr. Toole/Olive's paper, I have read that many times. Which specific portion do you think applies to this?
All of it and none of it, because IMO it is not stated in such a way as facilitates comparison with ordinary frequency response measurements.

I've been working on a way to restate its results in a way that is more useful for me.
 

Wombat

Major Contributor
Joined
Nov 5, 2017
Messages
3,869
Likes
2,164
Location
Australia
#7
Bell Research in the past established a link between lower and upper frequency break-points wrt telephony clarity. As the lower frequency break-point decreased the upper break-point was increased (or vice-versa) to give a balanced spectrum for audio balance.

That 40-15kHz spectrum reminds me of that work.
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
74
#8
Bell Research in the past established a link between lower and upper frequency break-points wrt telephony clarity. As the lower frequency break-point decreased the upper break-point was increased (or vice-versa) to give a balanced spectrum for audio balance.

That 40-15kHz spectrum reminds me of that work.
My take is that while the 20-20 KHz range is easy enough to remember, it is more of an icon than a real world standard.

For example, most people who have downloaded and reported on using a CCIF 19+20 KHz file for listening, report that it is silent when they play it.

When CRT based TV sets had 15.75 KHz flyback transformers that spread sound at that frequency, very few but the youngest ever seemed to notice it.

A lot of perceptually coded files are found to be very satisfactory in critical blind tests and many of them are brick walled at 16 KHz.

Very few speaker systems reproduce 20 Hz audibly and cleanly, and life seems to go on for the vast majority of audiophiles without it.

FM stereo was low passed at around 15 KHz using passive analog filters that had wide transition bands and nobody seemed to complain.

Almost all common musical instruments have a note in their 32-45 Hz as their lowest ordinarily played note.

Since about 1978 an increasing percentage of LPs were cut with in-line low pass filters with relatively gently sloped and wide transition bands that intruded well below 18 KHz related to cutting lathe pitch automation, and nobody seemed to complain. In fact many seemed to say that they preferred LPs even though the vast majority cut up until about the mid-90s had this fault.

Conclusion - the 20 Hz - 20 KHz range is generally for almost all listeners overkill and far from being sacrosanct.
 
Last edited:

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,725
Likes
3,212
Location
Central Fl
#9
FIIO X1 ii
04-05-2918

Whoa, Arny must be time warping with Scotty on the Enterprise compiling this data. :eek:
Or is it STNG's Data compiling the data? :confused:
 

Wombat

Major Contributor
Joined
Nov 5, 2017
Messages
3,869
Likes
2,164
Location
Australia
#11
My take is that while the 20-20 KHz range is easy enough to remember, it is more of an icon than a real world standard.

For example, most people who have downloaded and reported on using a CCIF 19+20 KHz file for listening, report that it is silent when they play it.

When CRT based TV sets had 15.75 KHz flyback transformers that spread sound at that frequency, very few but the youngest ever seemed to notice it.

A lot of perceptually coded files are found to be very satisfactory in critical blind tests and many of them are brick walled at 16 KHz.

Very few speaker systems reproduce 20 Hz audibly and cleanly, and life seems to go on for the vast majority of audiophiles without it.

FM stereo was low passed at around 15 KHz using passive analog filters that had wide transition bands and nobody seemed to complain.

Almost all common musical instruments have a note in their 32-45 Hz as their lowest ordinarily played note.

Since about 1978 an increasing percentage of LPs were cut with in-line low pass filters with relatively gently sloped and wide transition bands that intruded well below 18 KHz related to cutting lathe pitch automation, and nobody seemed to complain. In fact many seemed to say that they preferred LPs even though the vast majority cut up until about the mid-90s had this fault.

Conclusion - the 20 Hz - 20 KHz range is generally for almost all listeners overkill and far from being sacrosanct.


Except for electronic and pipe organ music there are few recordings that exceed the 40-15kHz range. Only youngsters are likely to hear above 15kHz.
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
74
#12
Bell Research in the past established a link between lower and upper frequency break-points wrt telephony clarity. As the lower frequency break-point decreased the upper break-point was increased (or vice-versa) to give a balanced spectrum for audio balance.

That 40-15kHz spectrum reminds me of that work.
The best recent papers I've read suggest that wider range within the ordinary audio range yields better intelligibility if there are not strong interfering signals. Reducing bandwidth to attenuate irrlevant signals remains an effective strategy.

If we simply allow ourselves to believe most recent scientific papers, response beyond 20 KHz generally has little or no effect. In order to obtain more positive results than that requires heroics that most people and producers are not going to put forth. For example, most microphones and loudspeakers that are used in recording and playback have roll-offs, even on axis that are either within or just outside the 20-20 KHz range. Many so-called omnidirectional speakers and microphones never were all that flat off-axis.

For example please observe the directionality of a common high quality "omnidirectional" microphone that is used for recording and measurements (attached). Note that there is 20+ dB of attention @ 20 KHz off axis and the laws of physics suggest that it only gets worse as you go higher. Trust me, the cheaper ones are no better unless they are smaller and this one is pretty tiny!


dpa 4006 mic directionality.png
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
74
#13
I believe that this may be the most over-arching of all current audio standards - the standard for transmission of audio via satellites various named J17 and BS 644.1
audio standard BS644.png


I suspect that the vast majority of what you hear via broadcast may have passed through a device that barely meets these standards.
 
Joined
Mar 16, 2018
Messages
6
Likes
6
Location
Montreal, QC
#14
Is there someone who can interpret those measurements for me? I am new to reading the graphs, and don't have a feel yet for what is good or not-so-good...
 
Top Bottom