Around 10 years ago, I decided to rip my entire CD collection. A lot of research on the matter made me realize that I needed to rip the CDs at high bitrates and that I had to make sure the metadata was in good shape. So I ripped my CDs at 256kbs using EAC and used EAC's built in CDDB feature to tag them.
Soon afterwards I stumbled upon MC8 and, lo and behold, all my metadata imported properly. The main reason why I started using MC8, was the configurable view schemes, and the fact that most other players had performance issues with my 508 CDs.
A while ago, discussions with Rick.ca and Darichman on the Personal Video Database forum, lead me to develop the PvdImport plugin for MC. So now, my TV, Videos, etc. collection is well tagged. And MC still performs well, even with viewschemes of over 60 000 people (actors, directors, etc.) . I'd prefer this functionality to be "built in" to MC, but the plugin solution works for me (for now). At least, it seems as if PvdImport and the other solutions out there have provided MC users with the means to getting this metadata.
So, my audio and video journey has lead me to two major conlusions regarding information: First, Information without metainformation is worthless. And second, the metainformation must be accessible and not proprietary.
As you may have gathered from the subject of this thread, I'd like to comment on how MC and other applications stand on images relative to these conclusions. It turns out that the situation is significantly more chaotic. There are, in my opinion, no applications that do image metadata handling well. Some applications handle certain aspects well, but there are *always* critical features missing. Could MC emerge as the ultimate image handling application? I think yes, but, as I'll show you below, there is still work to do.
So which meta data standards are there? Basically Exif, variations on IPTC, and XMP. The purpose of Exif is to save information on the hardware tech (shutter speed, camera type, etc.). IPTC shall provide information on image context/content (keywords, authors, etc). Xmp is a variation of XML where the purpose of fields and structure is defined by the vendor (in Schemas). I.e. it's application defined and may contain anything. All these standards are contained in the photo.
Geoff Coupe has a very good survey of various applications for image handling. I'm kind of bummed that I just found his blog because I too have tested a lot of applications. Jeff has a basic rule for image handling which he has learned the hard way: "I insist that any software used in the digital workflow (transfer from camera to computer, image selection, digital processing, cataloguing, publishing and asset management) will respect any EXIF, IPTC and XMP metadata that may be stored in the image file itself. "
I completely agree with Jeff. That's the reason why I cannot use MC for image tagging. Even sidecar files are sub optimal here because at some point, they're gone. Ok, I can agree that MC isn't as proprietary as Adobe or many other applications since there are methods for exporting the data, but you'll have to be quite tech savvy to do that. Also, say I used MC for image tagging, when I backup my images or give them away, should I also give away the database, sidecar files, or playlists? It's a no go.
And the many web services out there? Stay away! You cannot retrieve your tag data from Picasa, Facebook and the like. Think about that!
From a programming standpoint, reading embedded metadata is a lot easier than writing metadata (especially XMP). And for XMP, as I pointed out above, since no real standard exists except from "It looks exactly like XML", we cannot expect the developers at MC to take action. Yet. The good thing about XMP is that sooner or later, some XMP *Schema* will win over enough users to count as a standard. I suspect that one of those schema winners will be Microsoft's "ns.microsoft.com/photo/1.2/" or something similar because it's being supported by Vista, Win7 and Flickr.
By the way, I was experimenting with an IPTC tagger yesterday (Geosetter) and realized that even if MC imports IPTC, there is no 1-1 relationship. The one example I noticed here, was that [Places] consists of IPTC's City, Country, and State/Province. Why? An alternative would be to create a new set of 1-1 MC IPTC tags and let the user set up a calculated tag that merged these three IPTC tags. Then JRiver could say to users "We read Exif and IPTC in a one to one relationship". Later, if JRiver chooses to, they could add "we also write EXIF and IPTC, don't worry, the info is in the image".
And when that is in place, you could say to any user, professional and rookie alike: "By the way, we're also best in the market at displaying and organizing your images, we even have a nice 10foot display. 60 000 images? No problemo."
I'll start tagging my images this fall. I'll use GeoSetter for IPTC and GeoTagging (EXIF). For face tagging, I'll use Windows Live Photo Gallery (WLPG) since it uses XMP to store face tagging data in the image. Facetagging is important because you, and others, would like to know exactly who is who in a photo. WPLG also has a "rudimentary face detection algorithm which I'm sure will improve over time. I also like the graphical user interface of WLPG. Furthermore, Images uploaded to Flicr supports the XMP Schema of WPLG ("ns.microsoft.com/photo/1.2/"). This means that Flickr can display WPLG's Facetags. Flickr can also display GeoTags (EXIF) on a map, supposedly.
Sometime during this fall, I'll write up a WPLG XMP decoder plugin for MC which will translate the Face tags of WPLG into the people tags of MC. Then I can import all my images into MC and "rest assured that all my metadata is stored in the image..."
So then, back to my original question. I'll repeat it if you cannot remember
Where does MC stand with respect to "Information without metainformation is worthless" and "the metainformation must be accessible and not proprietary?" MC is good at displaying and organizing your images. But only if they are properly tagged. MC uses a "slightly proprietary database" where the info is not written to the file. Thus, other applications should be used to tag the data.