It may take some time to ferret out just how iTunes match works (or in some cases, doesn't) but I can assure you that at least in some instances, it DOES match vinyl rips. I know, because I've ripped quite a few myself and while not all songs end up matching, many of them do. Some records won't make it at all, but if they do, generally, 40%-60% of the tracks make it. There are a select few albums that have made it through 100% (Nautilus Superdisks), something that's not true even for some CDs. I'm in the process of comparing my vinyl rips to what is on the iTunes store in order to identify any discrepancies in track name, length, order (often vinyl track orders are different) that might in their totality keep a match from being made, but so far what I have are just a few clues to the mystery.
Given that I haven't yet heard of anyone else having success with the matching process when it comes to vinyl, finding out what it is that I do "right" is, I know, kind of major for anyone else wanting to repeat my success. It's important to me too, and I promise I'll keep working on it. You may have part of the answer: wow and flutter, clicks and pops, silence duration between tracks, tracks in different order, all could be factors. My guess is that there are many things that come into play, that Apple and Gracenote are using weighted criteria to determine what matches, what doesn't. If so, the key is to figure out which factors are the most important and to focus on those.
I won't put every detail of my recording process here, but assuming there's something in how I do it that affects the matching proceess, I will give you the basics. I use a Dual 1219, feeding a Yamaha CX-600U preamp, set in Direct mode. The Yamaha's output is fed directly into a MacBook Pro, so I'm at the mercy of its sound card. The Dual 1219 is a good table, but even using a strobe, I'm not about to say it's so speed accurate that I'm not losing or gaining appreciable time in my rips, compared to Apple's database. Even so, I've got songs matched that were off by a few seconds, so clearly that's not a deal-breaker. Conversely, I've seen songs with identical durations not match, so go figure.
I record using Audacity and only use two filters for posting in it. (1) DC offset remover because of the sound card (2) amplify to get the level of the entire album maxed (if I weren't lazy, I'd max each side separately, which I suspect is how the master cutter did it). I record both sides of an album at once and shoot for a -6 Db so I don't clip. Depending on the condition of the record, I may elect to put fade in/outs, but generally only on the first and last tracks of each side. I've recently started using ClickRepair which gets done after the above in Audacity, but I use it sparingly, sometimes not at all if the record's in good enough shape. I import back into Audacity to name my tracks and to output them. I leave the silence duration after each song intact (I don't replace it with true silence), though I do kill dead air at the beginning and end of each side.
(NOTE: I'm looking into whether first and last songs are less likely to get matched, I don't have that answer now).
Finally, while I record and process at 96000 Hz, 32-bit float, I output individual tracks at 44100 Hz in AAC format. These are the files that get imported into iTunes. Those files have some metadata in them, such as album and track names, genre. For some reason, artist name added in Audacity gets wiped before it hits iTunes, so I end up re-entering that, plus adding total number of tracks, something Audacity doesn't have a field for.
Below are the highlights of what I've discovered about the matching process in general and vinyl in specific:
(1) If you think that having a really good recording from pristine vinyl is going to improve the match, maybe so, but consider this. I have a beat up copy of Billie Holiday's Lady in Satin and some of those tracks matched. That recording was cleaned up using the click removal in Audacity, so it's still pretty noisy and probably more dull than it should be. My copy of Janis Joplin's Pearl is in better shape than Lady in Satin, but it's seen better days too. That one got cleaned up in ClickRepair. Several of those tracks matched too. Some tracks, like Mercedes Benz did not, but given the amount of silence, the noise on that one could have kicked it out, since if I think the music is going to suffer, I kick back the cleanup.
(2) The matching process likely puts weight on a scan of portions of the file that actually consist of music, which only makes sense, really. I point this out only to caution that just fixing track names, metadata etc. probably isn't going to do it. Songs in my library (from CD) with corrupted metadata, actually made it through. According to iTunes, this handful of songs had no metadata other than track number and total tracks on the disk. After the matching process, iTunes still showed them as being "track 04," but using iTunes to query Gracenote resulted in them quickly being matched and metadata restored. I suppose it's possible Gracenote was able to read metadata that was corrupted but that seems a bit of a strech to me.
(3) Getting something to "match" and getting it past "ineligible" are two different animals entirely. While it won't be the answer for everyone, all I had to do to get every "ineligible" track into the system was to make an AAC copy within iTunes. I would sort to show me ineligible tracks, batch process them, immediately delete the originals and have iTunes rescan for matches. They all made it through second time, though as usual, some matched, some uploaded.
(4) Deleting songs from your library and from the cloud (you get that option when you opt to delete a track) does NOT immediately delete it from the cloud. Trashing your local copies doesn't do it either. The tracks will be deleted from the directory you see in iTunes on the computer where you're working, but trust me, they stay resident on Apple's servers for some time. I know this because (1) enabling iCloud on an iPad after deleting albums from the MacBook Pro showed them still being there; (2) re-uploading those "deleted" tracks resulted in their status moving from "waiting" to "uploaded" in a heartbeat and never even showing in the status bar as being uploaded. I waited over 24 hours, tried it again, and same result. This is a real pain, as it makes experimenting with changes pretty much impossible.
Sorry that I don't have a magic bullet that will get your vinyl rips through. After all, my own success has been mixed. I'd say take heart, because it can be done, but until we determine just what are the most important factors to getting tracks to match, there's no way to know what you'll have to do to get your stuff to pass muster with Apple. Meanwhile, enjoy the music.
Message was edited by: JiminMissouri, fixed typo
Am in the same boat. Virtually none of my vinyl rips have matched, and in fact a lot of my CD rips haven't either. Compared to e.g. Shazam the algorithm seems very finickety. I expect this is to avoid false positives that would be acceptable to Shazam... Match has to be absolutely 100% to match a track, Shazam can just be 90%. What I suspect is that the algorithm is matching our songs, just not to the right confidence threshold...
I also wouldn't be surprised if a condition of record company permission to launch match wasn't to make that confidence threshold very high. It's not in their interest to launder terrible quality Napster era 128k rips into pristine 256k AACs. And unfortunately those of us with vinyl fall into that same boat.
What I'd like to see in future is something more like the way TuneUp handles artwork - i.e. the software gives you a list of possible songs and you choose the one that's right. I would be amazed if the record companies ever allow this though as essentially it would let you choose other songs you don't own and get them for free.
I do think that the matching algorithm will get some fine-tuning in future, though, so hopefully we'll see accuracy improve.