Sony MDR-7506 Review (Headphone)

[posvibes]

When I am in the right mood I consider them to be really very good headphones (using Amir's EQ) but even without and maybe it's my deteriorating hearing I don't consider them overly bright, but very present. My idea of bass must be somewhat lacking because I consider them to have quite some slam and that atmospheric bass drone stuff you get in movie soundtracks is well presented to my taste.

My AKG371's Harman tuned are much more subtle and relaxed, atmospheric and layered and offer relief after a few sessions with the Sony's, but I like all my headphones and it doesn't take long for me to adjust and enjoy most of them, I am promiscuous in that way I guess.

Except for the AKG240HD's that should be mounted on a tumbril paraded through the streets, cursed and heaped upon with excrement and eventually burnt at the stake. What were they thinking?

 

[GXAlan]

Its frequency response received the highest average preference rating of 94% (scaled, behind the Harman target at 100%) from listeners in Harman's blind tests.

What we don’t know is if these were the samarium cobalt or neodymium magnet models. Presumably they will be slightly different.

 

[Robin L]

My question still is not answered.
Since you read all the research into the finest details I thought you could give some answers to the question surrounding the 'listeners'.
1: Did they actually have 7506 on their head (bye bye blind test) or did they listen to another headphone EQ'ed to mimic a 7506 ?
2: The preference rating of 94% was that determined by measurements or was it scored by 'listeners in a blind test' ?
I assume you knew the conditions.
Oratory measured it and it scored 81.
I owned one (before my flawed measurements) and did not like it. Way too sharp sounding and lacking 'body' and sub bass.
Excellent monitoring headphone for around € 100.-
Great for instruments and voices.

Really great (and seemingly designed for) ferreting out buzzes and hum. Definitely a tool with a specific purpose.

 

[markanini]

What we don’t know is if these were the samarium cobalt or neodymium magnet models. Presumably they will be slightly different.

Depends on how the change was made. It's well known changing magnet types isn't a drop in operation. Assuming it was done in a skilled way there could be virtually no differences, still some concession to efficiency could have been made affecting FR.

 

[GXAlan]

Depends on how the change was made. It's well known changing magnet types isn't a drop in operation. Assuming it was done in a skilled way there could be virtually no differences, still some concession to efficiency could have been made affecting FR.

Yeah. It’s one of the mysteries I would love to have objective data on. We know that the magnet doesn’t matter as long as geometry and design are considered but I have to imagine that there may be variations in production consistency.

Hobbyists have always claimed different sound between the V6 and 7506 and the ground may not be the only difference.

 

[markanini]

Yeah. It’s one of the mysteries I would love to have objective data on. We know that the magnet doesn’t matter as long as geometry and design are considered but I have to imagine that there may be variations in production consistency.

Hobbyists have always claimed different sound between the V6 and 7506 and the ground may not be the only difference.

Another big issue is that Chinese manufacturing isn't always reliable, parts may be swapped without notice. Neodymium grades, wire gauges and coating types have different stock and pricing from week to week, the manufacturing facility might consider them equivalent and use whatever is cheap an available at the time of production, even if it changes the FR a little.

At least for a 20 year old headphone this is more true than newer headphones, which were designed around the current landscape of available parts.

 

[GaryH]

I thought I made this clear from the emphasis in my previous comment, but I'll obviously have to spell it out:

Its frequency response [meaning exactly that, recreated on a replicator headphone via precision equalization to eliminate sighted bias] received the highest average preference rating of 94% (scaled, behind the Harman target at 100%) from listeners in Harman's blind tests [meaning exactly that, not the predicted preference rating from measurements].

If anyone wants any more details on Harman's research I suggest they read this overview (and for more depth the other excellent articles on Sean Olive's blog), which shouldn't be too much effort, even for those who are not much of a 'scientific paper reader'.

 

[solderdude]

I thought I made this clear from the emphasis in my previous comment, but I'll obviously have to spell it out:

If anyone wants any more details on Harman's research I suggest they read this overview (and for more depth the other excellent articles on Sean Olive's blog), which shouldn't be too much effort, even for those who are not much of a 'scientific paper reader'.

Oh... I read that (and lots more) and know the answer to my 'questions'.

The real question is .... how well was the emulation of the 7506 done... it kind of hinges on that.
Seeing that the 7506 plots appear to differ from the one made by others (incl. Oratory) it could make one wonder why the 7506 scored this high yet on oratory's measurements it did not.

 

[preload]

I thought I made this clear from the emphasis in my previous comment, but I'll obviously have to spell it out:

If anyone wants any more details on Harman's research I suggest they read this overview (and for more depth the other excellent articles on Sean Olive's blog), which shouldn't be too much effort, even for those who are not much of a 'scientific paper reader'.

Scientific paper reader here.
The fact that the 7506 sounds so mediocre to so many people (including myself), yet supposedly exhibits a curve that was one of the most highly preferred when reproduced on tester headphones in Harman research, gives me pause.

 

[preload]

The real question is .... how well was the emulation of the 7506 done... it kind of hinges on that.

Apparently not that well. If I remember correctly, they didn't attempt to emulate the curve beyond a certain frequency, for instance.

Seeing that the 7506 plots appear to differ from the one made by others (incl. Oratory) it could make one wonder why the 7506 scored this high yet on oratory's measurements it did not.

The other rigs used different external ears and test equipment, right?

 

[GaryH]

gives me pause

It shouldn't, because (in order of likely influence): 1) high unit variance, as I showed just a few posts ago; 2) sighted bias, which will likely be significant and negative for a headphone that looks and feels cheap; and 3) potentially audible distortion (although Oratory has said the 7506 didn't have as high distortion as measured on here so this is less likely).

Apparently not that well.

But in reality well enough to get a high correlation of 0.85 between preference ratings given to real and virtualized headphones, and this despite nuisance variables such as leakage and tactile cues present in the real tests but not not virtual tests, as well as being based on measurements using the old artificial pinnae before Harman started using Todd Welti's more anthropomorphic custom pinnae used in the later preference studies we're talking about, which better simulate leakage on real human heads. Considering this substantial improvement, the virtualization correlation would probably have increased significantly higher than the 0.85 found with the original pinnae.

If I remember correctly, they didn't attempt to emulate the curve beyond a certain frequency, for instance.

They didn't equalize aggressively above ~10 kHz, due to positional variation up there. That still resulted in very accurate matches overall (right red and blue left channels offset by 10 dB for clarity):

The other rigs used different external ears and test equipment, right?

Again point 1 above is likely rather the culprit.

 

[solderdude]

People that have actually owned a 7506 (aside from the happy few that prefer the substantial treble peak) will understand that the real 7506 will not have one of the highest preference ratings but realize it sounds realistic and 'laser-like' (disective) with an almost unnatural emphasis on details.

Sharpness in recordings is in the 8-12kHz range and guess what, the industry standard test fixtures cannot show this (correctly) resulting in incorrect EQ in that area.
Thus also incorrect emulation of that peak which is very audible to most (our ears differ from a 'standard' above a few kHz).

The research was about finding a curve that was preferred and as their '7506' (see the plots from Harman) was EQ'ed to be pretty flat with less of a treble plateau it stands to reason that such a sound signature might be preferred. That does not mean the actual 7506 and the emulated sound are close in reality.
So save to say that the numbers quoted are not really related to the actual 7506.

That headphone is 'laser-like' in the treble. In a different way than HD800 and DT880 I might add. This is great for monitoring instruments and voices as details are high-lighted.
Not a hifi-phone even though one can listen to recordings and even like it.

The 94% preference is based on an incorrect emulation so no reason to buy it for hifi enjoyment unless you prefer laser-like highs (for most classical music it may well be fine).
Build quality is fine except for the pads. Too shallow and start to 'flake' soon when used often.
It doesn't get as high 'confidence' when measured by 'a professional acoustic engineer whose actual job it is to measure and help design headphones'

This is a studio tool not a hifi headphone that adheres closely to Harman (high preference rating).

 

[oldjackbob]

I tucked a circumferential bead of soft 3/8" poly rope behind the earpads inside the cans and that completely eliminated any ear discomfort (I don't like anything touching my ears).

I use my 7506 for ham radio operation, they're great for ESSB (extended audio bandwidth) and also for pulling very weak audio out of poor band conditions (where even my Kali IN-8 V2's can fall short). I can listen for hours without fatigue.

 

[Doodski]

I tucked a circumferential bead of soft 3/8" poly rope behind the earpads inside the cans and that completely eliminated any ear discomfort (I don't like anything touching my ears).

I use my 7506 for ham radio operation, they're great for ESSB (extended audio bandwidth) and also for pulling very weak audio out of poor band conditions (where even my Kali IN-8 V2's can fall short). I can listen for hours without fatigue.

Interesting. Is this a old school MOD? I never heard of ESSB but it makes sense to have a wider bandwidth headphone for the application.

 

[oldjackbob]

The poly rope is just something I thought of myself, I was trying to expand the inside volume of the earcup so it wouldn't contact the periphery of my ears.

ESSB (Extended Single Side Band) is sometimes referred to as "Voodoo audio" in ham circles. Typically the audio passband width in SSB operation is limited to 3kHz to give everyone enough room to coexist with everyone else in the limited band allocation afforded by the FCC, but that's too tight to squeeze anything resembling high quality audio. But sometimes there is some "elbow room" of band width available so it's acceptable to "open up" your transmit passband to something more than 3k wide, which allows the transmission of higher fidelity audio.

 

[Doodski]

The poly rope is just something I thought of myself, I was trying to expand the inside volume of the earcup so it wouldn't contact the periphery of my ears.

ESSB (Extended Single Side Band) is sometimes referred to as "Voodoo audio" in ham circles. Typically the audio passband width in SSB operation is limited to 3kHz to give everyone enough room to coexist with everyone else in the limited band allocation afforded by the FCC, but that's too tight to squeeze anything resembling high quality audio. But sometimes there is some "elbow room" of band width available so it's acceptable to "open up" your transmit passband to something more than 3k wide, which allows the transmission of higher fidelity audio.

Very cool. SSB is interesting. I studied that a bit when I studied electronics. I had a beefy Motorola RF amp for mobile SSB and gave it away because I saw no future in it for me as a mechatronics technician. I received a Motorola Radio and Trunking Certificate after one RF course that I studied. The RF study was a let down in that they never gave us hardcore fundamentals like in the AC-DC Principles, Semiconductor Devices and Digital Fundamentals courses.

 

[soundtrane]

I wonder why no one mentions the MDR-7509, a contemporary of the 7506...

 

[fpitas]

I wonder why no one mentions the MDR-7509, a contemporary of the 7506...

To my knowledge, the 7506 became a mainstay of on-site sound production. Not sure how popular the 7509 was.

 

[markanini]

To my knowledge, the 7506 became a mainstay of on-site sound production. Not sure how popular the 7509 was.

+1

 

[soundtrane]

Well, the 7506 seems to be mostly quoted here as 'studio' headphones, however, they were widely used for recording on location as sound pros used it while they recorded for fiction, documentaries and maybe also with broadcast - for film and tv (dialogue).

I too used the 7506 initially but switched to the Sennheiser HD25 (which had an issue of intermittent audio on either side) and eventually to the HD26pro (which eliminated that issue without any or much difference in the drivers).

I did use Ultrasone 650 Pro for a brief period - large sitting not on top of the pinna bout around it and very open sound. It did not work out for me for long, i needed better isolation for sure and eventually the HD25 provided this even though they sat on the pinna because they were much smaller than the 7506 or Ultrasone.

I found my ability to focus on the dialogue - more of an exercise in objective assessment of intelligibility - than hearing and finding issues with extraneous noises to work out much better with the Sennheisers than the 7506. Reason being most of the time either those noises are wide band and can be easily removed or they were going to be masked by other sounds in the soundscape of the scene. Nothing to me was/is more important than the intelligibility factor (which i find more and more missing despite better mics, recorders, wireless systems)...