Headphone CSD waterfall plots

[purrin]

 
Quote:

Any chance at getting harmonic distortion graphs too?

 
Most definitely yes. They are a pain to do, but I need to select a few standardized points (not to many.) I am open to suggestions though. Here are my thoughts - in terms of priority:
 

• 50Hz single tone
• 6000/7000 two-tone
• 500/3300 two-tone

 
Quote:

Mh, sorry for deing so demanding purrin, but would it be possible to measure the T1 again, this time with an output impedance of 100 or 120 Ohm via a stick-inbetween-adapter or something? Beyer's own A1 has that output impedance, curious to see what effect it has on FR and the rest. This reminds me, what kind of source are you using for the measurements, anyway?

 
I sent the T1s back. I can calculate approximate results (assuming output Z of 120ohms and based on Tyll's existing T1 impedance graphs) in my head for you: +2db at 100Hz.
 

 

[Head Injury]

Quote:

I sent the T1s back. I can calculate approximate results (assuming output Z of 120ohms and based on Tyll's existing T1 impedance graphs) in my head for you: +2db at 100Hz.

Isn't it about 1 dB? Or am I doing something wrong?
 
600/(600+120) = 0.833
1400/(1400+120) = 0.921
 
20 x log(0.921/0.833) = 0.869 dB

 

[purrin]

Quote:

Isn't it about 1 dB? Or am I doing something wrong?
 
600/(600+120) = 0.833
1400/(1400+120) = 0.921
 
20 x log(0.921/0.833) = 0.869 dB

Yup. For some reason I was was thinking 300 as base the instead of 600 for the T1.
 

 

[ZeNmAc]

Quote:

 
 
Most definitely yes. They are a pain to do, but I need to select a few standardized points (not to many.) I am open to suggestions though. Here are my thoughts - in terms of priority:
 

• 50Hz single tone
• 6000/7000 two-tone
• 500/3300 two-tone

 

Why not a full sweep?
 

 

[purrin]

Quote:

Why not a full sweep?
 

Each one of the above measurements will be visualized full spectrum (to see what lights up which shouldn't be there). It's better than saying 0.5 distortion at 100Hz. That gives you a number, but you have no idea of the kind of distortion.
 
Also, when you start doing two-tone tests, you end up with a lot of different combinations to chose from.
 
Best way is to show:
 
http://www.head-fi.org/t/566929/headphone-csd-waterfall-plots/120#post_7743752
 

 

[maverickronin]

Quote:

Why not a full sweep?

Quote:

Each one of the above measurements will be visualized full spectrum (to see what lights up which shouldn't be there). It's better than saying 0.5 distortion at 100Hz. That gives you a number, but you have no idea of the kind of distortion.

You could do full sweep of a single tone and plot it in some kind of 3D like in the burn in tests.

 

[ZeNmAc]

Quote:

Each one of the above measurements will be visualized full spectrum (to see what lights up which shouldn't be there). It's better than saying 0.5 distortion at 100Hz. That gives you a number, but you have no idea of the kind of distortion.
 
Also, when you start doing two-tone tests, you end up with a lot of different combinations to chose from.
 
Best way is to show:
 
http://www.head-fi.org/t/566929/headphone-csd-waterfall-plots/120#post_7743752
 

Makes sense.  Still though, wouldn't you want a full sweep of just the harmonic distortions for an easier comparison between headphones over the whole spectrum?  I don't mean THD, a full sweep showing fundamental and at least second and third harmonics separately.
 
IMO its easier to read and gives you the 'big picture' of linear distortion.  Then also couple for nonlinear.
 
Again, you have more experience, so if I'm missing something I'm all ears

.

 

[stv014]

Quote:

Originally Posted by ZeNmAc /img/forum/go_quote.gif
 
IMO its easier to read and gives you the 'big picture' of linear distortion.  Then also couple for nonlinear.

 
All distortion is non-linear (unless you consider a non-flat frequency/phase response a kind of "distortion"). But testing sweeps is a good idea, because sometimes there is high distortion only in a relatively narrow range of frequencies, for example due to some resonance problem. Also, especially for bass frequencies, distortion vs. level is worth testing. Dynamic headphones often have problems reproducing loud bass with low distortion, so a comparison in this aspect is definitely useful.
 

 

[xnor]

Quote:

 
All distortion is non-linear (unless you consider a non-flat frequency/phase response a kind of "distortion").

 
That unless part you wrote is in fact called "linear distortion".
 
 

 

[stv014]

Quote:

That unless part you wrote is in fact called "linear distortion".

ZeNmAc was apparently referring to the swept THD test as linear distortion, though, if I read it correctly.
 

 

[plin]

Quote:

I am amazed by how clever some of you guys are! Somehow, people who have no background in measurements / acoustics have such ability to think outside the box! This verification by reciprocity is very neat!
 
......
If you can send me a text file with the time data you got (need the sampling rate or time & response vectors), I'll be happy to generate a frequency response and possibly a csd curve.
 
Cheers, arnaud

 

Well we do get inspired by all of you the experts 

. People like you, Tyll, Purrin, and others are really great contributors to our hobby! 
 
 As a side note, I am an engineer but in a different engineering field.
 
Thank you for your kind offer. I will take more measurements (with and without the headband) and I will send a text file to you, just for the fun of it, because I doubt that we will get really meaningfull results from this non standard setup.
 
Quote:

EDIT: I am wondering if it's the metal gimbal assembly. It is sort of shaped like a tuning fork. If I have time today, I'll run it on the RTA and tap on the gimbal.

I suspect also the metal grill that covers the outside. If you do find some time to take the extra RTA measurements, it would be very interesting, thank you. 
 
Cheers, Panos

 

[ZeNmAc]

Quote:

 
All distortion is non-linear (unless you consider a non-flat frequency/phase response a kind of "distortion"). But testing sweeps is a good idea, because sometimes there is high distortion only in a relatively narrow range of frequencies, for example due to some resonance problem. Also, especially for bass frequencies, distortion vs. level is worth testing. Dynamic headphones often have problems reproducing loud bass with low distortion, so a comparison in this aspect is definitely useful.
 

 
Ah thank you for the correction.  I meant harmonic distortion.

 

[purrin]

Non-linear distortion thread started: http://www.head-fi.org/t/597670/headphone-non-linear-distortion-measurements#post_8172836

 

[purrin]

Ultrasone Edition 8.

 
As always, subjective impressions first:
 
Wow! These sound like slightly more refined Beats Studios. Recessed mids. Less muddy than Beats. The treble is kind of sharp! Turning up the volume to see how they behave a louder listening levels. Ouch! Ouch! Taking them off NOW. Holy Cow! I think I'm deaf or at least I've lost some high-frequency hearing. To my buddy Questhate: "WUT? These cost $1500? You've got to be kidding!" Since the EU limits volumes on music players, I think the EU should also place limits on treble nastiness. This headphone will kill any high-frequency hearing you have any left - after that - you will be all set.
 
I usually set the floor lower to ~36db these days, but any lower than the -30db used below just shows too much crap. The usual stuff (note frequency and time ranges, etc.) applies.
 
 
 
 
 

 

[plin]

Quote:

Ultrasone Edition 8.

 
 
I usually set the floor lower to ~36db these days, but any lower than the -30db used below just shows too much crap.

OMG, it is very discouraging to see this bad behaviour to such an expensive headphone. Your valuable work Purrin is even more important in cases such as this!