Clock Test

So I did a test of the internal clocks of both the MOTU UltraLite (the original one) and the M-Audio 2626. Using the internal microphone preamps we're looking at stock versions of each device.
The difference is that I flipped back and forth between the internal clocks on each of those interfaces and the S/Pdif clock from an Apogee Mini-Me.
The recording was made at 96kHz.

The EQ I used on the tracks. Just a little scoop at 277 seemed to do the trick.

The guitar is a Martin D28 going into an AKG 460 with a CK1 capsule. I put a tiny amount of EQ on the tracks just to cut out some of that low-frequency stuff and the 70Hz cut was switched on at the microphone itself. I'm recording in my relatively noisy control room but I used no noise reduction.
I then offer a version which is "mixed" with some nice reverb and compression. This is just so you (I) don't get tired of listening to dry guitar.
Remember, all of this is at 96kHz. I did not A/B anything against 48kHz or the like.
Also, I did not use fancy-pants mic preamps. No Neve, no API. So you're listening for clock differences. But you're also hearing preamp differences.
Note that you won't see the above link if you got this post as an email or in an RSS reader. You must come to the blog.
The numbers in the file are from Freesound contributor Corsica S.
Now ostensibly you should listen to the above track (embedded in Bandcamp) without my prejudicing you regarding which track is which. So I won't. Just listen.
S
P
O
I
L
E
R

W
A
R
N
I
N
G

It goes like this:
1. 2626 internal clock
2. 2626 Apogee clock
3. 2626 internal clock mixed with some reverb and compression more the way I'd probably actually do it
4. 2626 Apogee clock with reverb
1.1 UltraLite Apogee clock
2.2 UltraLite internal clock
3.3 UltraLite Apogee clock reverb
4.4 UltraLite internal clock reverb

My conclusion? Wow, at 96k there certainly is no difference that's particularly shocking between the internal and external clock. Knowing this will save me some money in external clocks, that's for sure.
Now I also did a couple quick tests with playback. At 48kHz the MOTU converter sounds much better with an external clock. I knew that before I did this test. With the external clock the high end is much less cloudy. I did not hear those differences when I was listening at 96kHz.