[LazyListener]

  Most closed-back headphones have a treble peak around 9k-10kHz.
Is that really "severe treble roll-off"?  Treble is all there to the human ear, with some excess in a particular area.

About the treble spike around 9 kHZ, many, but not all, of their headphones, whether open or closed, show that in their curves.  Strange, as I don't recall noticing that on other sites like headroom.
 
Yes, IMO, that FR curve for the 598 Cs, shows severe treble roll-off.  If you exclude the 9 kHz spike, which could be testing or equipment induced, most of the treble is way below the target and even more so relative to the mids.  The only exception really being as you approach 20 kHZ, it creeps back up to target, but most adults can't hear beyond about 17 kHZ anyway.
 
It just reaffirms, except for the boosted mids, the 598 Cs really is a dark sounding headphone.  It lacks clarity and sparkle in the treble.  This based on actually having owned it for a while.

 

[samvafaei]

I've noticed that too. Every headphone graph that I've looked at on that site, whether closed back or open back, shows a spike at 9khz. I came to the same conclusion; that it was thier methodology or thier testing equipment. I still find it a good site for comprehensive graphs and interesting headphone reviews.

We have just updated our compensation curve, and are in the process of updating our measurements with the new test bench (which includes measuring the bass of over/on-ear headphones on 5 human subjects, 5 times each). This update should greatly improve our Bass and Treble scoring, however, we do still experience some inconsistencies above 8KHz in our results. So if you see a +9dB peak at 9KHz, in reality it may sound more like a 5dB peak. This could have a few reasons, (like the amount of smoothing, or different positioning) which we can discuss here if you guys are interested, but for now we are not scoring high-treble as heavily and lower frequencies.

 

[LazyListener]

 
We have just updated our compensation curve, and are in the process of updating our measurements with the new test bench (which includes measuring the bass of over/on-ear headphones on 5 human subjects, 5 times each). This update should greatly improve our Bass and Treble scoring, however, we do still experience some inconsistencies above 8KHz in our results. So if you see a +9dB peak at 9KHz, in reality it may sound more like a 5dB peak. This could have a few reasons, (like the amount of smoothing, or different positioning) which we can discuss here if you guys are interested, but for now we are not scoring high-treble as heavily and lower frequencies.

Love your site, tests, and reviews.
 
How do you measure the bass on 5 human subjects?  Is this strictly a subjective test, or is it an objective measurement and the human subjects are essentially being used as dummy heads?
 
What is "smoothing?"

 

[samvafaei]

 
Love your site, tests, and reviews.
 
How do you measure the bass on 5 human subjects?  Is this strictly a subjective test, or is it an objective measurement and the human subjects are essentially being used as dummy heads?
 
What is "smoothing?"

Thanks!
 
We use the human subjects as dummy heads because getting reliable bass measurements on our dummy head was quite tricky. We're working on a video to explain our procedure which should be out in a month or so.
 
Smoothing is easier to understand if you see it: http://stackoverflow.com/questions/12987905/how-to-make-a-curve-smoothing-in-matlab

 

[gargani]

We have just updated our compensation curve, and are in the process of updating our measurements with the new test bench (which includes measuring the bass of over/on-ear headphones on 5 human subjects, 5 times each). This update should greatly improve our Bass and Treble scoring, however, we do still experience some inconsistencies above 8KHz in our results. So if you see a +9dB peak at 9KHz, in reality it may sound more like a 5dB peak. This could have a few reasons, (like the amount of smoothing, or different positioning) which we can discuss here if you guys are interested, but for now we are not scoring high-treble as heavily and lower frequencies.

Thanks for responding to my post. I didn't know if you were aware of the prevalent spike around 8khz-10khz. Your explanation for this does make sense.

Just an observation, I would like to mention.The frequency response graphs on some of the popular sites use 1khz as the "base line'', then measure the rest of the frequencies relative to that.
I notice on your site, you take your compensation curve, then flatten it and use this as the "base line" (the blue dotted line).The rest of the frequencies are plotted relative to thistheoretically, what humans would perceive as flat)
Am I correct in this Observation?

By the way, as I said, I do enjoy and appreciate your site; and the effort you obviously put into it.
Looking forward to your upcoming video.

 

[Doomski]

Hey guys, I'm trying to find these headphones full potential; Anobody have any tips(equalizer tips maybe?)?

 

[SAMplify101]

MATLAB! man... I haven't used matlab since my Georgia Tech days! But it did open up the doors for R so naturally, I totally approve of your site. Not like I didn't before, you're stuff was great as a reference when testing the m20-m70 range

 

[samvafaei]

Thanks for responding to my post. I didn't know if you were aware of the prevalent spike around 8khz-10khz. Your explanation for this does make sense.

Just an observation, I would like to mention.The frequency response graphs on some of the popular sites use 1khz as the "base line'', then measure the rest of the frequencies relative to that.
I notice on your site, you take your compensation curve, then flatten it and use this as the "base line" (the blue dotted line).The rest of the frequencies are plotted relative to thistheoretically, what humans would perceive as flat)
Am I correct in this Observation?

By the way, as I said, I do enjoy and appreciate your site; and the effort you obviously put into it.
Looking forward to your upcoming video.

You are correct. Since we calculate a score for our measurements, using 500Hz or 1KHz as the base line wouldn't be fair to the headphones that deviate from the norm in the 1KHz region. Our method probably isn't perfect, but better than normalizing the response at a single frequency.

 

[gargani]

You are correct. Since we calculate a score for our measurements, using 500Hz or 1KHz as the base line wouldn't be fair to the headphones that deviate from the norm in the 1KHz region. Our method probably isn't perfect, but better than normalizing the response at a single frequency.

Yeah, that's fair enough.
I'd like to ask about an anomaly I noticed the other day. I was looking at your review of the skullcandy crusher and the frequency response graph. Then I went to the page where you can compare the graphs of different headphones (overlayed on each other). The frequency response of the crusher was completely different from the response shown on the graph in the review. It looks like the response from a completely different headphone, inadvertently, shown where the crusher's response should be. Is this the case? Just curious.

 

[samvafaei]

Yeah, that's fair enough.
I'd like to ask about an anomaly I noticed the other day. I was looking at your review of the skullcandy crusher and the frequency response graph. Then I went to the page where you can compare the graphs of different headphones (overlayed on each other). The frequency response of the crusher was completely different from the response shown on the graph in the review. It looks like the response from a completely different headphone, inadvertently, shown where the crusher's response should be. Is this the case? Just curious.

I can't replicate what you reported. Is the problem still there? Could be old cache, or that you checked the page just after an update.

 

[SAMplify101]

people on reddit are telling me that the m50x coiled cable will also double for the 598cs. Im sitting here like a moron trying to get this in and the fat part of the plug wont let me extend deeper into the hole. Luckily this is only a problem with headphones, whereas I can see this being a huge problems with "other" issues

 

[LazyListener]

  people on reddit are telling me that the m50x coiled cable will also double for the 598cs. Im sitting here like a moron trying to get this in and the fat part of the plug wont let me extend deeper into the hole. Luckily this is only a problem with headphones, whereas I can see this being a huge problems with "other" issues

Does it really need to extend deeper for you to be able to turn and lock?  Is it possible the 598 plug fits in M50x but M50x plug doesn't fit 598?

 

[SAMplify101]

 
Does it really need to extend deeper for you to be able to turn and lock?  Is it possible the 598 plug fits in M50x but M50x plug doesn't fit 598?

that's exactly correct. the hd598CS cable will go into a m50x like it belongs! Apparently the 598cs port is deeper than the 598se or 598 and those redditors were claiming it worked based on the openback versions. 

 

[gargani]

I can't replicate what you reported. Is the problem still there? Could be old cache, or that you checked the page just after an update.

Thanks for your response. I checked it out again today, what I found was if I go to the crusher's review page and look at the frequency response curve, then go to the "compare the frequency response" page, the curves are the same;but if I go to a different headphone review and select the "compare" page with the crusher's as the headphone being Compared, that's when the completely different curve appears. I tried this with a number of different headphones and this anomaly only happens with the crusher's. It's not a big deal. I just find it puzzling.

By the way I forgot to mention, I was pleased to see your review and graphs for the 598cs.

 

[samvafaei]

Thanks for your response. I checked it out again today, what I found was if I go to the crusher's review page and look at the frequency response curve, then go to the "compare the frequency response" page, the curves are the same;but if I go to a different headphone review and select the "compare" page with the crusher's as the headphone being Compared, that's when the completely different curve appears. I tried this with a number of different headphones and this anomaly only happens with the crusher's. It's not a big deal. I just find it puzzling.

By the way I forgot to mention, I was pleased to see your review and graphs for the 598cs.

You seem to have found a bug! Will report it to our devs, thanks.