Headphone Measurements Using Brüel & Kjær 5128 HATS

[bobbooo]

Do the two have the same (approximately) frequency response? Because if you look for cheap very low THD earphones, you can go with the HifiMan RE-400 which you can find for less than 30$.

No they don't have the same frequency response - the Sony MH755 fairly closely matches the Harman target, whereas the ER4SR follows more of a diffuse field target. The RE-400 seems to have somewhat uneven treble response from the measurements I've seen. The MH755 used to be able to be had for just $5, until people realised how good they were and supply started to thin out, bumping up the price a bit. It might be hard to find a pair now, but its cousin, the MH750, still seems to be floating around, and has similar frequency response and low distortion to the MH755 as measured by Speakerphone/ClarityFidelity:

 

[Feargal]

That is insane. $10k for a moulded plastic torso? It would be cheaper to go and find a real one.

Maybe an ASR reader is willing to make the ultimate sacrifice? Perhaps we, or should I say I, draw straws.

 

[Cahudson42]

Maybe an ASR reader is willing to make the ultimate sacrifice? Perhaps we, or should I say I, draw straws.

Ah yes - plastination. Fascinating process. A wonder of modern science! If you didn't know:

https://en.wikipedia.org/wiki/Plastination

 

[Thomas_A]

Been wearing the 240 for half hour now and my ears hurt. The cups are too small for my ears. You all know I have large ears so I can hear so well and outdo all of you all. The cost of that is that I need man sized headphone cups!!!

My ears are large too and the DT150 fits better and it sounds good IMO, when the pads have been changed for the DT100 pads. The pads are however not the softest around though.

 

[infinitesymphony]

Perhaps eventually it will be possible for more people to make their own measuring appliances. I understand there are liquid silicone rubber (LSR) 3D printers, though they are probably prohibitively expensive to own right now. Maybe something a community makerspace will have in the future.

 

[Robbo99999]

Perhaps eventually it will be possible for more people to make their own measuring appliances. I understand there are liquid silicone rubber (LSR) 3D printers, though they are probably prohibitively expensive to own right now. Maybe something a community makerspace will have in the future.

I like that idea, but that pic is showing lots of crinkles where the ear has been built up layer by layer.....I don't know exactly how sound waves of different wavelengths bounce off surfaces but I can imagine waves bouncing off at strange angles when hitting the grooves & steps that can be seen in that pic.....bouncing off at strange angles that I imagine a smooth real human ear would not exhibit. So I'm thinking if it's gonna be 3D printed it can't have all those visible contours/steps, if we're wanting it to behave properly as a real ear.....that's just my intuition but others who know specifically how sound waves of different frequencies would react to that would know better. (I think it would affect the high frequencies the most).

 

[bluefuzz]

There are a host of sellers on Etsy making bespoke silicon sex toys. Surely one of them would be willing to make some silicon ears for a suitable remuneration. Just say it's for aural sex ... ;-)

 

[Mad_Economist]

Y'all're overthinking the 3D printing angle - Todd Welti had this right, just DLP/SLA print a mold at high res and cast/injection mold the silicone like so

 

[Robbo99999]

Y'all're overthinking the 3D printing angle - Todd Welti had this right, just DLP/SLA print a mold at high res and cast/injection mold the silicone like soView attachment 79166

That way you don't get the visible ridges in the artificial ear? Both because it's Hi-Res & because it's injection moulded?

 

[Mad_Economist]

That way you don't get the visible ridges in the artificial ear? Both because it's Hi-Res & because it's injection moulded?

Resin printers can manage really high resolution - 10 micron isn't impossible, and 50 micron is standard for a lot of consumer stuff. That's better than you'll manage with pretty much any other form of additive manufacturing I'm aware of - and that goes double with a substance like silicone that's really annoying to extrude consistently. It's sort of a horses for courses thing: if you want a smooth piece of silicone, you're better off starting from a rigid mold that's well made and then putting liquid silicone into it.

Edit: Note that notionally an FDM or similar can have a vertical layer resolution that's comparably low...but good luck getting a comparably smooth print out of one without a massive amount of post processing.

 

[Francis Vaughan]

Wavelength at 20kHz is roughly half an inch. Ridges in a 3D printed pinna will be inconsequential. What may matter is how well the structure matches the flexing properties of the cartilage in the ear. I do wonder how much this affects lower frequencies.

 

[amirm]

Wavelength at 20kHz is roughly half an inch. Ridges in a 3D printed pinna will be inconsequential. What may matter is how well the structure matches the flexing properties of the cartilage in the ear. I do wonder how much this affects lower frequencies.

At very high playback levels, I can get my ear lobes to resonate with music so there may be some impact there.

 

[BYRTT]

At very high playback levels, I can get my ear lobes to resonate with music.....

Impressive ..

.....so there may be some impact there.

Sounds be a same effect as using a dynamic EQ process ..

 

[Robbo99999]

At very high playback levels, I can get my ear lobes to resonate with music so there may be some impact there.

Ha, I had thought that was a joke when you write that in your headphone amp reviews!

 

[Robbo99999]

Ok folks, I created a couple of different EQ's for the AKG K371 that was measured by Amir in this thread from his results at this post: https://www.audiosciencereview.com/forum/index.php?threads/headphone-measurements-using-brüel-kjær-5128-hats.15352/post-487178. @BYRTT turned the graph in that post into data measurement points which I then plotted in REW and I then manually created the parametric filters in REW to fit the theoretical Harman Curve for the B&K 5128 mannequin (the Target Curve being a result of labour from @Mad_Economist which he used the Chris Struck Method to calculate the Harman Target Curve). I chose AKG K371 because a friend on here has K371 so he can do some listening tests on it, otherwise I would have chosen a headphone I had too (I don't own any of the tested headphones).

First set is based on only one of the measurements, and that was the measurement that had the best bass response and which was closest to that which Oratory1990 measured. Here are the results, I tried to use as few filters as possible and tried to use the lowest Q values as possible, I didn't want to use sharp filters and I didn't do anything above 10kHz, no smoothing used on measurements:

Second set is based on an average of the remaining 4 curves, but I consider these bass measurements to be a little flawed in all of those because it was so far off from what Oratory1990 measured.....but interestingly these were the 4 curves that mirrored each other the most, but I still think they're not as valid as the EQ I showed you above.

Well that was a fun exercise, it'll be interesting to see what you think to the EQ's I've done. I think the first one is gonna be the most valid, but both probably worth trying.

(I created all of the filters manually)

(I'm a bit concerned above 10kHz how far the measurements are below the Target Curve, but above 10kHz is notoriously individual to each person in terms of measured response if you were gonna measure your own ear above 10kHz for example.....and it would need a ridiculously massive High Shelf Filter boost above 10kHz to match the target, so that's not reasonable, and therefore also suspect I think, hence I didn't do that.)

 

[BYRTT]

@Robbo99999 in my browser your post 516 seems miss some picture or whatever data to present your dialed in EQ settings, what i mean you can see in below 50% sized screen capture of that post, so please edit ..

 

[Jimbob54]

Ok folks, I created a couple of different EQ's for the AKG K371 that was measured by Amir in this thread from his results at this post: https://www.audiosciencereview.com/forum/index.php?threads/headphone-measurements-using-brüel-kjær-5128-hats.15352/post-487178. @BYRTT turned the graph in that post into data measurement points which I then plotted in REW and I then manually created the parametric filters in REW to fit the theoretical Harman Curve for the B&K 5128 mannequin (the Target Curve being a result of labour from @Mad_Economist which he used the Chris Struck Method to calculate the Harman Target Curve). I chose AKG K371 because a friend on here has K371 so he can do some listening tests on it, otherwise I would have chosen a headphone I had too (I don't own any of the tested headphones).

First set is based on only one of the measurements, and that was the measurement that had the best bass response and which was closest to that which Oratory1990 measured. Here are the results, I tried to use as few filters as possible and tried to use the lowest Q values as possible, I didn't want to use sharp filters and I didn't do anything above 10kHz, no smoothing used on measurements:

Second set is based on an average of the remaining 4 curves, but I consider these bass measurements to be a little flawed in all of those because it was so far off from what Oratory1990 measured.....but interestingly these were the 4 curves that mirrored each other the most, but I still think they're not as valid as the EQ I showed you above.

Well that was a fun exercise, it'll be interesting to see what you think to the EQ's I've done. I think the first one is gonna be the most valid, but both probably worth trying.

(I created all of the filters manually)

(I'm a bit concerned above 10kHz how far the measurements are below the Target Curve, but above 10kHz is notoriously individual to each person in terms of measured response if you were gonna measure your own ear above 10kHz for example.....and it would need a ridiculously massive High Shelf Filter boost above 10kHz to match the target, so that's not reasonable, and therefore also suspect I think, hence I didn't do that.)

Agree- images are missing.

I'd be intrigued to hear yours (and friends) opinion on stock, the 2 EQ you have done here , but also Oratory own derived EQ from his measurements. Probably a longer term mission , but one worthy of investigation (and I mean the full EQ, including the plus 10k where Oratory gives settings)

 

[Robbo99999]

Ha, will re-edit images......I just cut & pasted from a PM message to the guy I created the EQ's for, so even though it looks OK my end, it looks like cutting & pasting the message messed it up.....will try to rectify now.

EDIT: OK, did that work?

 

[Jimbob54]

Ha, will re-edit images......I just cut & pasted from a PM message to the guy I created the EQ's for, so even though it looks OK my end, it looks like cutting & pasting the message messed it up.....will try to rectify now.

EDIT: OK, did that work?

Yup

 

[Tks]

@amirm

SO boss man, what'd you decide, I must've missed it if you posted it somewhere. You're not going for this thing are you I hope?