Quote:
The only mentioning of balance there is in reference to the frequency balance, not the Left-Right channel balance.
In my experience, imbalances have been stable (ie: I have not experienced total failure or balance drift). You will experience this if you go through enough headphones, especially older ones. Many people don't even much notice or care up to 1.5db imbalance with stereo music, thought that sounds like an awful lot. An easy way to subjectively test for channel imbalance is to hook your amp/headphones up to a computer and enable Foobar's "downmix channels to mono" DSP filter - pop in any music CD and give a listen. You should be able to discern differences down to 0.25-0.3db if you listen carefully. The Radio Shack SPL meter can give you more objective readings if the imbalance is around 2db or greater.
Best, would be to send them back to the manurfacturer for installation of new drivers. Barring that, see if you could obtain a spare set of matched drivers and have them installed by yourself or a friend.
When all else fails, it's fairly easy to build a custom attenuator (via a standalone box with RCA jacks or custom interconnect cables) to reduce the louder channel's SPL level. The attenuator should go between your source and your amp. Imbalances of around 3.5db or less can usually be completely fixed without discernable degradation to sound quality - the only exception is if your imbalance is very frequency dependant (that is very bad), which should be rare. I once ran across a pair of K340 that had a busted electret driver - the right and left sides were matched, but starting at around 3 or 4KHz the right side rapidly decreased to 10db below the left side! You can use tone-generator software with an SPL meter to test for a frequency dependant imbalance.
Keep in mind that the db reduction of an attenuator will be dependant on the input impedance of your amp. So, it helps to know this value. You can use a combination of Foobar mono downmixer, soundcard balance control, and SPL meter to determine what db attenuation value you need. Then, you can convert db value to a resistor value (in ohms) by the following formula:
R1 = Za/(10^(x/20)) - Za
where R1 is the resistor value you need (in ohms), Za is the input impedance of your amp (ohms), x is your desired attenuation in decibels (should be a negative number).
So, for a 100K input impedance (like my SDS) and a 3db attenuation you would have:
R1 = 100K/(10^(-3/20)) - 100K = 41.25K
Now, this will cause a slight channel imbalance between the input impedance your source sees (the attenuated channel having the higher input impedance by R1), but the sonic effect of this should be negligible except for very large attenuation values. If you want to be more rigorous you could add shunt resistor (bridges signal and ground) so that the input impedance is the same for both channels. To do so, use this formula:
1/Za = 1/(Za+R1) + 1/R2
where R2 is the shunt resistor value.
Not sure anybody cared that much, but that's all there is to this issue
But, I did notice one thing which I wasn't sure if I was correct about. It seems that the right driver is slightly quieter than the left driver. Not much, but I noticed it in several tracks.
