that interesting. does it only affect the hd800's that way? i haven't used my sansui for like a month since it's been sitting in the shop due to tech having backlog of other work.  when i get it back i'll have to test it out. also curious was the loudness button turned on by any chance? that can act like a bass boost. great results though.

Yup only the HD800. The higher the output impedance, the more EQ that you are going to get that looks like the headphone's impedance graph (you can grab those from InnerFidelity or Headroom.) 

The LCD# have a flat impedance curve, so the FR on those are not going to change at all. But a lot of Beyers will also exhibit the same behavior (large bass boost, tiny high treble lift) as the HD800. The AKG701 have similarly flat impedance curves as LCD# - so those will not.

HD800 Z is about 350ish from 1-10kHz with a peak of 650ohms at 100Hz!

I still have the DT990 if you want to run the experiment with them.

headphones are not like phonelines. impedance matching is not a must.

Agreed. Impedance matching is required if the cable run is relatively long and/or the frequencies are relatively high, due to reflections (echo). I reckon that for a 20kHz signal, echo issues will happen if the cable is a few km long. I come from taking impedance matching for granted given my background. It is only recently that I learned that for audio, it is a whole different game... Nice to learn new stuff

EDIT:

Quote from: RexAeterna on January 13, 2012, 08:29:55 AM

that interesting. does it only affect the hd800's that way? i haven't used my sansui for like a month since it's been sitting in the shop due to tech having backlog of other work.  when i get it back i'll have to test it out. also curious was the loudness button turned on by any chance? that can act like a bass boost. great results though.

Yup only the HD800. The higher the output impedance, the more EQ that you are going to get that looks like the headphone's impedance graph (you can grab those from InnerFidelity or Headroom.) 

The LCD# have a flat impedance curve, so the FR on those are not going to change at all. But a lot of Beyers will also exhibit the same behavior (large bass boost, tiny high treble lift) as the HD800. The AKG701 have similarly flat impedance curves as LCD# - so those will not.

Your comment above, and overall thread along with this source (http://www.avguide.com/blog/why-headphone-amps-sound-different-frequency-responseimpedance-issues) helped me understand today why one has to be very careful when putting together a system (headphone + amp). Similar to your comment regarding pairing the Stax 007 with the right amp. Oversimplifying amp characterization to measure really low no longer seems like a golden rule to me.

Given the Beyer 880 impedance characteristics, and the fact that the 990 probably uses a similar driver, it is possible to get more hump out of a higher impedance amp. Howerver, it probably will result in more receded mids and will not tame the tremble. From all this discussion, it is clear the HD800 will gain significant impact with a higher impedance amp.

Thank you very much for making this clear to me! 

Interesting results you got here Purrin. Unfortunately, unless measurements like these are taken for each model, isn't it difficult to assess exactly how much effect would a higher Zout have ?

Besides, AFAIK things are vastly more complicated when it comes to Multi BA drivers CIEMs, with a much lower given impedance. And by extension, all low impedance headphone get in the same boat. IOW, the close to 0 Ohm Zout rule still makes sense for all these, isn't it ?

Hmm. Not sure. That was a long time ago.