[anearfull]

I’m a relative Newbie here trying to figure out the significance of the “driver matching” specification given by different manufacturers. I understand why close matching would improve stereo imaging but how meaningful are the numbers I’m seeing?

Grado specify that their lower priced models are matched to +/- 0.1 db and their higher priced models are matched to +/- 0.05 db. Looking at Sennheiser: their top of the line HD600 is matched to +/- 1.0 db and the slightly lower-end HD580 only has a driver matching spec of +/- 3.0 db.

This board sports a “Team Grado” and “Team Sennheiser” that delight in throwing darts at each other however after all the insults have been traded I think it’s fair to say that both Grado and Sennheiser provide an audiophile sound that is “different” but "competitive" depending on your personal tastes. If driver matching were a very important parameter then, you’d expect the Sennheisers to sound like earmuffs compared to the Grados.

Are the two companies simply using a different statistical reference for their number? By that I mean that Grado might be using 1 sigma or an averaged difference while Sennheiser might be using a maximum difference at a particular frequency or a 2 sigma overall difference in the smoothed difference. Can anyone clarify what’s happening?

My first question leads to another one: what do they mean by the word “matching”? I see two scenarios:

Scenario A)
"Matching" is a verb signifying that someone is taking each driver and actually “tuning” it to match the spectrum of its twin.

Scenario B)
"Matching" is an adjective that passively describes how well the driver’s measured spectrum matches an ideal or statistically derived spectrum.

If “matching” is a verb then a technician is first measuring each driver’s spectrum and then actively modifying its structure until it “meets spec”. The physical changes he makes are not disclosed but I assume they are somehow “tweaking the diaphragm”, “scrunching the voice coil” or some other trade secret. Any thoughts on what magical activities are involved in this sort of matching?

If however “matching” is an adjective then all the drivers are manufactured using the exact same materials and technique. The technician is merely sorting the resultant drivers into various bins. In the case of Grado, the same “meat grinder” would produce all the drivers used in the SR125 all the way up to the RS1 (below the SR125 there is a stated change in their voice coil and diaphragm construction). All the technician would do is measure and compare all the driver spectrums and sort them into various bins inside of which all the drivers have similar performance. The drivers used in the RS1 would come from the bins such that their spectrums differed by no more than 0.05 db. The drivers used in the SR125 would come from the bins such that their spectrums differed by twice that much.

My question in a nutshell: is “matching” intended as a verb or an adjective?

Moving over to the Sennheiser camp, here’s another question that might be worth kicking around. I see that I can buy a replacement HD600 driver for $34 (part # 49294). If I buy two of them, can I glue them into any open-backed headphone shell that I have kicking around and expect to hear a sound that’s similar to a genuine HD600? If not then why not?

Peter

PS:
This is a really great forum

 

[skippy]

i believe the drivers are manufactered and then measured and then matched according to those measurements (usually only at one frequency or a small range of frequencies).

regarding the hd600 drivers, the earcups and earpads play a large role as to how the headphone sounds. so i would imagine that buying a pair of drivers and sticking them into an earcup they were not designed for will not result in the ideal sound.

 

[jude]

I haven't visited Grado yet, but I did visit Sennheiser's factory, and I can say that they do all sorts of testing before the headphones are boxed up and shipped, with the higher end ones receiving more human attention and testing than the very inexpensive headphones (which go through more automated, and apparently less rigorous, testing procedures).

I didn't stick around any of the testing stations long enough to give you a detailed description of the testing procedures, nor did I get the impression from Sennheiser that they wanted me to do this, but I was very surprised by just how much testing is done to the various headphones there.

One of their tests (I don't remember exactly which products were being tested with it, and I don't think I'm supposed to discuss anything specific anyway) involved a small room that I believe was heavily shielded from RFI/EMF. I'm not sure if they used it for final testing of some products, R&D, or both.

Anyway, my point is, thorough testing does happen.

 

[Beagle]

Driver matching is critical if you want any semblance of soundstaging and imaging. You need matching in output and timing. I suppose it is as critical as tube matching. where there are sonic benefits to be had.

Grado tests by listening and it is fairly easy (to the trained ear) to spot driver mismatch. Without proper matching, the benefits of mahogany chambers and UHPLC copper cannot really be appreciated to the fullest.

 

[Hirsch]

One day I will figure out why it is important that headphone drivers be more closely matched than my ears are.

 

[gdahl]

I believe our brains compensate for the mechanical nonlinearities in our ears. Our perception of the direction of sound sources is created through the interaction of several different mechanisms, such as differences in arrival times between left and right ear, and timbral changes from different arrival paths around the head and through the pinnae. Everyone's head and ears are shaped differently, but the brain calibrates itself to normalize its response to real-life stimuli, so that "reality" will sound "real".

I suppose this may be part of the reason why it is so hard to fool the ear into thinking reproduced sound isn't an imitation.

(...although I believe Hirsch's comment was tongue-in-cheek)

 

[pedxing]

I think it would be a lot easier to build a batch of headphone drivers and then match them accordingly instead of tweaking and modifying them to match another driver.

The enclosure makes a big difference in the sound. For example, if you cup your hands over a pair of open headphones, you would probably notice a difference in the acoustic characteristics. Shift the headphone around your head backward or forward and you would notice that the bass and treble characteristics would change too. The position of the driver and the acoustics of the enclosure play an important role in how a heapdhone sounds.

From a previous thread like two years ago, some people have reported that some HD580 have tighter matching drivers than some HD600. I guess its all luck of the draw because some HD580's are just going to happen to have closer matching drivers than some HD600's. I doubt Sennhesier would go about and sabotage a good pair of HD580 with closely matched drivers to ensure that no HD580 is better than a HD600.

 

[Hirsch]

Actually, my comment wasn't particularly tongue-in-cheek. I know that one of my ears has a slightly different frequency response than the other (audiology testing). In a speaker setup, I can understand driver matching because the signal from both sides is reaching both ears. Driver matching is key to creating the illusion of a spatial image.

However, in a headphone setup, the right ear doesn't hear what the left ear does, and vice versa, not even with crossfeed. The stimuli are independent. The key matchup would seem to be between the driver and the ear. Matched drivers in theory would only be useful to matched ears, which I suspect are less common than we'd like to think. OTOH, I could well have accomodated by now, so that since I'm expecting the distortion of the differences between my ears, unmatched drivers would accentuate it. Just rambling, and wondering what drivers matched to 0.05 dB means, when my ears aren't that matched that closely.