Hey guys, lots to do and I can’t get back to every thread immediately, particularly when it’s really a feature request (and really one for @mike!). I’m sorry that this one went unnoticed/unanswered for so long.
That said, back cover art (and more) is absolutely something we want to make happen. However, it’s more complicated than simply doing what allmusic does. Allmusic takes Rovi/TiVo metadata and, although it supplements it in some areas, fundamentally, it’s just Rovi. This makes it easy for them to simply serve up whatever Rovi gives them.
For us, Rovi is not enough; we need to fill in very real album coverage gaps, so we use multiple metadata providers. And then there’s TIDAL… This creates a pretty unique problem for us, equivalencing different editions of albums for which we have metadata, and identifying your content as accurately as possible. Now bear in mind that for rear cover art to be unobjectionable, we can’t serve up incorrect album release / image matches and then have to deal with user complaints that we’ve got it wrong; this is a terrible waste of our time and a major distraction from getting work done that really improves the user experience. So we have to get it right.
Finally, having the metadata is only part of the story; we’ll need to do some user interface design work to ensure an awesome user experience when we deliver this.
How are we going to get there? In general, we think that we do a pretty darn good job with equivalence; however we recognize that there are some problems and it’s clear to us that our current metadata infrastructure and pipeline could do with some work. Certainly, they were never designed to be able to ingest and equivalence an arbitrary number of metadata sources, as well as finely control the pecking order of these sources for every possible type of metadata (album art, classical composition metadata, reviews, etc.) . So we’re in the process of designing a “next gen” metadata processing pipeline and service. We’re taking this opportunity to fix a number of systemic problems, allow the system to grow in various directions, and facilitate new product features (of which rear cover art will actually be quite a trivial one from the metadata perspective). This isn’t a small project that can be delivered in a couple of months, but it’s an important one; and it will be worth it.