Fishbase and the rest of the planet is wrong; planetcatfish is right.
Otocinclus affinis is easily recognizable by 1) golden sheen, 2) a very narrow lateral band (it's unusually pale, barely visible in the dead? fish in the Planetcatfish picture), 3) iris diverticulum. It is also, like its close relative
O. flexilis, a large species, almost twice the size of most other
Otocinclus species.
The fish on the Fishbase image does not have an iris diverticulum or golden sheen, but has a wide lateral pigment band: whatever it is, it's not an
affinis.
I've actually mailed Fishbase about the fact that the image for
O. affinis shows some other species, but nothing's happened - presumably they're waiting for the designated catfish expert to check.
At the root of the problem is:
a) that pretty much all
Otocinclus are sold as "affinis" or "vestitus", neither of which I've ever seen for sale in the trade. My guess is that this is because those are two species early on in Schaeffers key to
Otocinclus, and which you get to if you make a mistake on some fairly difficult characters.
...and...
b) that fishbase accepts photos from anyone, and relies on the public to tell them which photos are wrong (if the photo is disputed, the issue is referred to an expert).
If you do a google image search for "otocinclus affinis", you will notice the massive spread in looks of the hits: at least half a dozen different species are sold under that name.
I did however find one other actual
affinis on the net, this young male, which shows the narrow pigment band better than the large, and possibly dead, female on the planetcatfish picture does:
http://www.otocinclus.de/affinis.htm
