Metadata options for classical CDs

Having found the cataloging/metadata features associated with the CD ripping functionality of my Uniti Star worse than useless for my classical CD collection (see my post of June 23 for related comments), I’m now looking for alternatives. If you have successfully ripped and assigned metadata to a large classical collection (including both the auto-assignment of metadata and the ability to manually edit all of it while maintaining the ability to display the edits in your NAIM system), I’d love to hear what software/metadata source you used and any tips you have. Ideally I’d like to rip from the Star ad hoc (the prospect of ripping my entire collection makes me want to invest in a 55 gallon barrel of cognac) but I’m prepared to give that up if I can’t get useful results. Thanks in advance.

1 Like

Thanks so much, jmtennapel. This is a VERY helpful response. I use windows but I’ve found a tag editor “Kid2” that seems to provide very similar functionality to the Amvidia product. Some questions, please:

  1. To be sure I understand: you’re not running a UPnP server, right? You’re actually editing the tags in the NAIM Music Store using your own schema?
  2. Where are you keeping your Music Store?
  3. When you say you “keep albums together”, say you have collection of the 9 Beethoven symphonies with a single orchestra and conductor. Do you get them to display with a single instance of the “album” cover or does the cover appear 5 or 6 times or however many disks there are? If the former, do you just put the 9 symphony names in the “song” field? Are these fields limited in size?
  4. How about a “Great Performances” collection of performances of several quartets by different quartets (performers)?

Greatly appreciated! Very best wishes.

Can’t help with the Star, but I got seriously into streaming when I retired and my first retirement project was to rip my CD collection which included nearly 1000 classical CDs. I used dBpoweramp to do the ripping. This collects metadata from various sources, but I always edit it subsequently. For this I use Metadatics (I use a Mac). My upnp server of choice for classical is Minimserver and this affects the way I arrange metadata. For example, Minim treats the “Group” tag in a particular way which is particularly valuable for discs containing several multi-movement works when you might want to listen to just one of them. I arrange works in the same order as the old Gramophone classical catalogue and have added a Type tag which classifies works as eg, string quartet, opera, symphony. But that is just my way. I think the main thing is to be consistent. If you can locate a work with just a few taps, your tagging system is working.


I’m finding similar issues as I embark on archiving my 900 CD’s.

My approach is slightly different - and it may yet turn out to be short sighted, but…

When I bought CD’s, my first priority was the ‘composer then work’, so I structure my music database likewise:
Classical > Initial > Composer > Work
So, in the Naim app, I engage the following key presses
CD’s (or Hi Res)
Classical (or other genre)
B (e.g if I want to listen to Barber)
Adagio for Strings
Movement 1 (or first file)

So in 7 keystrokes, I can be playing whichever work I choose. The downside is that I have no album art. But I make no reference to a particular album as I have no need and focus on the work.

Hope that helps.

My music is kept on a Qnap NAS running MinimServer. CDs are ripped with EAC and if necessary tagging is tweaked with Mp3tag using a similar format to jmtennapel…

Song: Work name - part name
Artist: Composer
Album artist: Blank
Conductor: conductor
Album: album name

I use mainly classical downloads which I play through my Naim Core, so I don’t know whether this will be helpful to you. I find the metadata is regularly incorrect and is a common problem across a number of download sites, such as Qobuz, hiresaudio, presto, eclassical etc. I edit the metadata using Jaikoz or Metadatics, the latter is I believe only for a mac whereas Jaikoz is also in a windows form.

Having spent more hours than I can really justify trying to decide on a schema for Classical music metadata.
There any innumerable ripping programs dbpoweramp, EAC, XLD
There are an equal number of metadata editors idtsgs in PerfectTunes, mp3tag
Similarly uPNP servers Asset, MinimServer, Naim Core

Each have their own set of wonderful attributes and you can tag data to a high degree of detail Album
Genre eg classical
Style symphony, chamber etc
All of which make for easy searching, however

Then there the Control Point - Naim App , Foobar2000 etc

Before you start to rip multi thousands of CD’s decide how you are going to interact with the ripped music and your UnitiStar.
Are you going to use the Naim App, on phone or tablet or some other method eg have a laptop with another control point eg foobar2000 running
What do you want to see when the music is playing Composer, work, movement, album, conductor orchestra etc or are you happy to know what work you have decided to enjoy and only need basic info on control point screen.

The Naim App appears to be fixed on
Track Title

So let’s take a example Mahler Symphony No 2
Album 1 Berlin Philharmonic, Sir Simon Rattle, Kate Royal, Magdalena Kozena RundFunkchor Berlin.
Album 2 Budapest Festival Orchestra, Ivan Fischer, Birgit Remmert, Lisa Milne, Hungarian Radio Choir.

Raw ripped metadata on Naim App shows 3 lines of text
Album 1
Symphony No 2, “Resurrection”: I. Allegro maestoso
Simon Rattle
Symphony No 2, “Resurrection”

Album 2
Symphony No 2 in c minor - “Resurrection”: I. Allegro maestoso
The Hungarian Radio Choir
Mahler Symphony No 2 in c minor - “Resurrection”

This as you can see is just an example of the pretty inconsistent state of metadata attached to classical music.

So back to my original questions
if you are happy with knowing randomly selected bits of the whole picture then just rip to a storage device with just a check on accuracy of data at the ripping stage ie is the attached data correct.
Or if you are a bit more OCD about what you want to know you will probably have some work to do tagging data to the relevant fields before (or later) in a consistent manner in your chosen schema.
jmtennapel schema above looks a good way of showing on the Naim App and fitting most search criteria.

So in our example the App will show
MAHLER: Symphony No 2 in c minor “Resurrection” - I Allegro maestoso
Sir Simon Rattle - Berlin Philharmonic Orchestra, Kate Royal, Magdalena Kozena, RundFunkchor
Symphony No 2, “Resurrection”

However one could argue there is duplicated data in having Symphony No 2 displayed in 2 fields and maybe the SONG should just be
I: Allegro maestoso

I still haven’t come up with a satisfactory solution, however I think having a uPNP server is a good idea as that will help searching for something to listen to.
There is a wealth of information out on the inter web on tagging classical music

You chose a schema that works for you, be consistent, TRY IT OUT ON A SMALL NUMBER OF CD’s to see if it works for you.

Alternatively go to you CD storage area, select a CD put it in the Star sit down and enjoy the music :grinning:

Thanks to each of you for detailing your various approaches for digitizing your collections and creating the associated “metadata”. I appreciate each of your responses and am particularly appreciative of PeakMan, jmtennapel, Harpo, and VinoVeritas who went beyond the call to explain your solutions.

I have yet another solution which was waiting in front of my face unrecognized. For many years I’ve been cataloging (i.e., creating metadata for) my CD collection using the Windows Readerware Music app which is, essentially, a classic database. It supports all fields one might need…composer, multiple artists, conductor, orchestra, soloist, title, record or cd label, etc etc…as well as a cover image. I’ve learned that it has a mobile app which syncs to the db. I’ve synced and, while it has some issues, it will do the job. I’ve selected a field in the db to add a control number and I’m going to add that control number to the .wav files that my Star creates. So I will have a 2 step process: find what I want to play in the synced Readerware db and then retrieve the related tracks from a simple file manager view that includes a few identifiers depending on the nature of the CD (typically composer title artist) and the control number.

And I’m going to rip gradually and opportunistically, not en masse. So who knows whether I’ll ever complete the job but I’m thinking why, if I never play the CD again, would I bother to rip it?

Anyway, that’s my solution for now. If this approach is of interest to anybody, let me know and I’d be happy to discuss it further.

1 Like

Before I started using Roon I tackled this issue and had some success.

I wrote about my method here.

It assumes you are running Asset. It’s the UPNP server that will define how things look, not the Naim App.

What I mean is that the ‘now playing’ screen of the Naim App only displays what is stored in the Title, Artist and Album tag.
Yours is probably the best schema I have seen and will probably adopt it for my use Thank you for sharing.

Running Asset 6.2 on Windows 10 currently unregistered as may change to Raspberry Pi
Thanks I had a quick go at following your instructions. Must have done something wrong as despite the changes being saved to MediaDatabaseFieldsv5.txt and the browse tree and restarting Asset and the App there was no change to the Search bit of Naim App. Will have another go.

However my original point still stands all these changes affect the SEARCH bit of the App, complex searches can be set up using detailed meta tags.

BUT it is what is put in the Track, Artist and Album fields that is displayed whilst listening to the piece of music selected. Therefore, a decision as to what one wants to SEE when the movement is playing had to be made before spending the time setting up the metadata to the chosen schema.

So from my examples above in a bit more detail…

Mahler Symphony No 2

Album 1
Berlin Philharmonic, Sir Simon Rattle, Kate Royal, Magdalena Kozena RundFunkchor Berlin.

Album 2 Budapest Festival Orchestra, Ivan Fischer, Birgit Remmert, Lisa Milne, Hungarian Radio Choir.

The RAW ripped metadata from dbpoweramp without tidying up when played on Naim App shows 3 lines of text

Album 1
Symphony No 2, “Resurrection”: I. Allegro maestoso data stored in TRACK
Simon Rattle data stored in ARTIST
Symphony No 2, “Resurrection” data stored in ALBUM


Album 2
Symphony No 2 in c minor - “Resurrection”: I. Allegro maestoso
The Hungarian Radio Choir
Mahler Symphony No 2 in c minor - “Resurrection”

My arguments for want of a better word are…

I) it’s inconsistent - but as we know most classical metadata is.
II) what does the user want to see when the music is playing: a lot of info or having found the piece from a detailed search just the basics and therefore tag 3 displayed fields appropriately.
III) there’s a duplication of information on the display in the standard format.
III) there is a lot of effort involved in retagging data to individual fields which may or may not be use.
IV) go with what is given and just enjoy the music once its been found :grin:

Thanks, I am wrestling with the same issues and got some interesting ideas from reading respo9ses here.

I use a tool called SongKong Pro that allow me to make changes and make the collection ripped so far conform. The combination of Minimserver 2 and SongKong Pro is promising for those interested enough to invest in the learning time.

The next problem is finding a nice controller app to make use of it all.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.