Pardon me for asking, but what source were you using to listen to the AM6?
Your description sounds to me as though the output stage of your source may have been struggling to control them sufficiently (my UM Miracles sounded awful from a Cowon J3 but fantastic from a DX100, and I mean massively different).
Hmmm... further to my earlier remark, by sheer coincidence I just stumbled across this discussion, in the DX50 thread (http://www.head-fi.org/t/661167/dx50-a-new-smaller-dap-coming-from-ibasso-specifications-page-1/1320#post_9645407):
Nope I had an RWA AK100 and sold it because I didn't like the sound. I preferred the Studio to it. I won't risk modding my AK120. I'm very happy with it as is and don't want to take the risk. I was not happy with the RWA AK100 so I don't know if I'd honestly like the RWA AK120 or RWA AK120S either. Then factoring in I void my warranty and have to use an amp instead of having a choice just doesn't appeal to me. I'm glad if it works for others though. I also have a Tera Player here that I'm borrowing. It's a nice unit. Still not better than the AK120.
These days though it seems the sound of a DAP isn't solely based on the h/w (DAC, OpAmp, mods, etc.) but also by the firmware. Have you had a chance to listen to the RWAK100 with its newer firmware (may it be iRiver's 2.01 or Listrid's)?
It does concern me somewhat how much firmware plays a part in sound signatures - kinda makes impressions even more complicated.
Edit: & I didn't even touch the SDCard size vs sound impact topic :-P!!
Nope I updated it one time when I owned it. Didn't notice any changes in sound at that time. Still sounded rather harsh in the highs for me and the bass was bloated. With usage it did calm down a little but still was muddy and harsh for me. I was just glad to sell it off in the end. Was not very happy coming from the DX100 at that point :(.
Regarding the SD Cards well that's just what I hear. I'm not gonna purposely inconvenience myself if I didn't hear an appreciable difference. That really makes me wonder though what that one guys posted earlier on (the read speeds of the 32 GB cards being faster than the 64 GB cards?). Whatever it is I didn't like what I heard when comparing them. Now if I just wanted convenience and didn't mind sacrificing a bit of sound quality then I'd still have the 64 GB cards.
I agree it's strange how firmwares impact sound and how much it can impact sound. But I've been accustomed to this since the Studio V then the DX100 :P. Wasn't really expecting the AK120 to change in sound though as I didn't notice the RWA unit do this. But I'm guessing on the newer firmwares it does indeed change the sound?
(N.B.: Please note that emphasis and enlargement of text was done by me, not the original authors)
@ Iron Lung:
Lee's AK100 was a Red Wine modded one. I don't know what variant yours is. However, your description is extraordinarily similar to the one offered by Lee. Therefore, I speculate that the AM6 you heard may not have had good synergy with your source (even the iPhone 5).
The more complex multi-BA driver CIEMs can potentially present an unusually reactive load to an output stage. It seems congruent with this fact that it was the most complex of the AM line-up that you felt sounded the worst.
Let me reassure you that I have no vested interest in defending the AM line-up. I'm speaking about this from a relatively neutral standpoint. From personal experience, the UM Miracle taught me an important lesson about how poor a multi-BA CIEM (6 drivers in the case of the Miracle) can sound if not driven by a very capable output stage that has good synergy with the CIEM. When driving my Miracles with my old Cowon J3, the bass was inadequate, vague and muddy, the mids lacked definition, and whilst I wouldn't describe the treble as 'harsh', it was definitely splashy, erratic, and poorly-defined. The bass was so poor that I was ready to downgrade to the Merlin, and then... then, I bought a DX100 and wow! What a spectacular difference in the sound of the Miracles. It's not really worth bothering, but if you care to search Head-fi, you'll see I did discuss exactly this, at the time, so I'm not making any of this up to suit our discussion here
Incidentally, Lee's remarks about his (ex) RWAK100 are particularly interesting because the RWA mod lowers the output impedance of the AK100 (22 ohms as standard, or thereabouts, if memory serves me correctly), so one would expect the RWAK100 not to perform so poorly, in the manner he describes (as you may be aware, one of the biggest hurdles in driving multi-BA CIEMs is having too high an output impedance in the output stage).
Edited by Mython - 7/25/13 at 12:12pm