Thanks for the link. Watched quite a bit of Amirs video's but forgot to watch this one.
I guess my real problem with the methods of testing is the fact that the refference is so high
@amirm? I understand it is convenient, but I honestly think it would be far more valuable to readers if there would be an additional (or two?) lines lower on the graph.
As for the comparison to live recordings... I see your point, but is that really a fair comparison? I personally would never listen to a full Rock concert without ear protection. And I sure won't even get close to that level of volume in a headphone. If we would be comparing PA systems, or maybe home audio systems, I can see where you are going to. But with headphones, I think this isn't all that relevant. This research just shows that these live events also cause quite a bit of hearing damage if you do not wear ear protection. In my experience hearing protection is very strongly encouraged. Some live events I have been to hand out free ear plugs, and I have even seen it being mandatory for people below 18 on some live events 2 years ago!
As for the A weighting and what sound pressure is most prevalent in music, they get partially cancelled out by the Harmon curve. The most sensitive frequency for our hearing (1-5KHz) is also the highest part of the Harmon curve, which is roughly 5dB higher than the bass. So under the line that only leaves roughly 15 dB difference between bass and 3KHz in sound pressure (assuming the song in the video (20 dB difference between 20 Hz and 1Khz) on a headphone following the Harmon curve). So 124 dB of bass gives us still roughly 109 dB(A) at 1 Khz, extremely high, literally off the (WHO) charts.
On the peak vs average dB, that is a very solid point. Assuming the peak/average difference measured by diy-audio-heaven is somewhat representative for most music, we would get another 12 dB difference between peak and average dB.
So to calculate from peak sound pressure (in bass) to A-weighted average sound pressure in the 1kH region (dB(A)), one would have to substract a total of 27dB (+5 from harman curve, -20 difference level in music, -12 peak vs average). That still means a peak level of 114dB in bass results in an average SPL of 87, something you really shouldn't be listening to for longer than 10 hours per week! There are some assumptions in there of coarse, but I think they are fair.
I think it is very fair you try to stress test these headphones. It is, like you said, your job.
So what is my point then? Do I think these graphs are wrong? no, the data is clearly there. Do I think you get deaf from a short burst of 104 dB? no, but hearing damage is not that black and white. However, I do think these graphs show a skewed picture and uneducated people might draw the wrong conclusions.
The graph is heavily skewed towards a loud + EQ listening style, which is not representative for the average user. So honestly, all I am really debating here is the addition of an additional measurement, at 84dB, and maybe take this more into account with drawing conclusions.
124 dB of bass, is an extremely niche measurement, only relevant for people who listen at very high volume with EQ. Even the 114dB is really up there, but probably more represented in a normal distribution of average users (the 1% you talk about maybe?). Taking the above assumptions and the WHO guidelines into account, people who listen to music often shouldn't really get bass levels above 107 dB regularly!
As a practical example to close it off: I haven't heard my focal Clear clip once, and if I wasn't able to read these graphs correctly, I would have never even gave them a fair try in the shop, and I would have missed out on a great headphone for my personal use case.
Just my 2 cents and I hope you take this into account in the future
TLDR: your measurements are not wrong, but you are measuring extremely niche use cases which can lead to people drawing wrong conclusions.
p.s. I absolutely value this webpage as a source of objective information, but like in science, we can only grow from discussion