Ripping a ‘mood’ CD on Linux

1.0 Introduction

This is the fifth article in my series showing how I would suggest you go about the business of ripping and tagging different CDs of varying configurations. Previously, I’ve discussed:

In this article, I’m going to discuss how I might go about ripping a single CD containing a bazillion compositions by one or many composers. This is one of the most problematic areas of ripping-and-tagging, because value judgments have to be made -and your approach to making them might not be the same as mine. To illustrate the problem, I'm going to use two CDs as an example of the sorts of decisions you might take.

Firstly, there's this one:

That's 21 tracks all by a single composer as it happens -but the point is that it's lots of compositions, none of which is really significant in its own right. Almost every track falls under my '4 minute rule' (which isn't really a rule, but is discussed in Section 4 of my Axioms of Classical Tagging article) -and the ones which don't cut it pretty fine!

The second CD I'll discuss is this one:

That's a CD containing 18 different tracks that really are by 18 different composers -and, again, a lot of the tracks (but by no means all) fall under the '4 minute rule' threshold and so probably shouldn't be considered 'standalone compositions' in their own right.

I won't be describing the specific point-and-click processes required to rip or tag these CDs, as I think the general principles and practices involved have already been discussed at length in the earlier articles in this series. In this article, I'll concentrate instead on the issues arising when considering how to rip these CDs and what the best outcomes from an appropriate ripping-and-tagging process might be in each case.

2.0 Thinking in the Aggregate

The problem with both these CDs, in other words, is that nothing on either of them really stands out as a composition 'worthy' of being regarded as a separate, standalone 'album' in ripping and tagging terms. In the first case, Britten didn't write something called 'Folksongs'; all 21 tracks were genuinely separate compositions, some as early as 1940, some as late as the 1970s. But no one composition is of sufficient significance to stand on its own two feet and every one of the 21 tracks gets its worth largely by being part of a set of 21.

Putting it yet another way: do you ever sit down and think to yourself, 'I'd like to play O Waly Waly'? Or do you instead think 'I feel like a half-hour of Britten folksongs'? If you want to play things as individual pieces, then catalogue them that way. But if you find the main value of a CD is in listening to it in the aggregate, then catalogue it as a single entity containing lots of tracks. Resist the temptation to remember the exceptions, too, when thinking about how you approach a CD. In my case, I adore The plough boy, Tom Bowling and The bonny Earl o' Moray. There are, in fact, times when I do think, 'I fancy listening to The plough boy', all on its own. But I more generally think in terms of the Folksongs as a collection, so the exceptional factors governing three of them shouldn't blind me to the fact that I usually think of them in the aggregate.

This is also something that effects the Love Came Down at Christmas CD: it's a CD all about Christmas carols and similar choral works which evoke the 'mood' of Christmas -and they do that effectively in the aggregate. Whilst there are a couple of tracks whose length certainly makes them candidates for separate ripping (see track 6, for example, at over 8 minutes long), does it make sense to rip them out into standalone albums and leave the sub-4-minute mass behind as a lump of Christmas-themed short pieces? Your view on that might be different from mine, and that's fine.

If you thought that, say, tracks 6, 8, 13 and 16 should be regarded as significant works in their own right, you're in the same territory as I already discussed in the '1 cd, multiple composers, multiple composers' article. So the steps outlined there should work for you in producing 5 separate 'albums' from the 1 CD (i.e., track 6 on its own; track 8 on its own; track 13 on its own; track 16 on its own; and everything else as a 'Christmas music from Wells' virtual album). I won't say such an approach would be wrong -but in this case, I think tracks 6, 8, 13 and 16 only really get their 'power' by being part of the entire Christmas CD experience. The 'mood' the CD is promoting is more significant than the merit of its individual compositions. This is indeed commonly true of 'mood' or 'themed' CDs: break them apart into separate compositions (i.e., separate 'albums' ) and you diminish the intended effect.

Where you 'feel' the contents a CD in the aggregate, in other words, don't attempt to rip it as a set of separate 'virtual albums', but rip it entire and whole, as a single 'album'.

What you call that one 'album' is then really up to you -though taking the lead from the CD manufacturer/music publisher isn't a bad place to start. "English Folksongs" and "Love Came Down at Christmas" seem like good album names to choose for my two specific examples, in other words.

The real point here is that it's not 100% clear-cut. The decision to 'split into virtual albums' or 'treat as a single CD' is a matter of judgment and the way you approach your music collection. What's right for one person may not be the true for all. As I say, if you like the 'multiple virtual albums' approach, I've already described how to rip and tag those. The remainder of this article is for those occasions when you feel that it's imperative to rip the disk as a single, 'mood' or 'theme' CD.

3.0 Who is the Composer?

The trouble with aggregating multiple, diverse tracks as a single 'album' is that we need to find something to tag the resulting 'virtual album' with an ARTIST and COMPOSER tag -and Axiom 2 says that ARTIST and COMPOSER should always be the same thing.

In the first of my CD examples, this isn't a problem: all 21 tracks on the 'English Folksongs' album I'm about to create were written by Benjamin Britten, so ARTIST and COMPOSER can both be set to Benjamin Britten with a clear conscience. The tagging in that case is not only functional but entirely accurate!

But in the second CD, this does become a problem, since every one of the 18 tracks on our 'Love Came Down at Christmas' album was composed by a different composer -including the dreaded 'traditional'. Who, then, can be set as the ARTIST and COMPOSER? Well, you obviously cannot set it to be 'William Mathias', since that would be wrong for 17 other tracks. Similarly, you can't say it's 'Sergei Rachmaninov'... since that would be wrong for a different set of 17 tracks. In fact, the only possible 'composer' name that would be valid for all 18 tracks is one that never actually lived: Compilation.

Some people prefer the words 'Various Artists' to be used in this situation. However, I feel this wording implies that we still care about the individual composers or 'artists' when, in fact, we've given up caring about them at this point. I mean only that we are currently thinking about how to tag things at the album level -and at that level, we simply don't have a real, genuine value for ARTIST or COMPOSER that we can supply -so rather than use wording that implies that somehow we care about the 'Artist' at this album level, we would be better off (I think) making a clear declaration that we've got an album which is merely a compilation of works by multiple people, none of whom we can meaningfully identify individually at this level of tagging. I would therefore strongly recommend not to use the 'Various Artists' tag.

4.0 Forget the Composer entirely?

Despite what I've just said, however, I am reluctant to forget the composers entirely. They lived (and usually died!) to give us the benefit of their artistic vision, after all. Just ignoring their existence seems a tad ungrateful!

Accordingly, when you declare the composer/artist to be "Compilation", I would argue that the composer's name should nevertheless be included in the TITLE tag for each track. This is a breach of normal tagging protocol in a sense: we normally use the composer name only in the ARTIST and COMPOSER tag, which applies album-wide; and because we've usually specified it there and we care to not violate Axiom 7, we usually do not include the composer's name in the TITLE tag. In this case, however, since we have not specified the real composer's name at the album level, Axiom 7 would not be violated if we now mentioned the composer's name in the TITLE tag.

For the sake of consistency, I would suggest we use the composers' full names in the TITLE tag of a compilation CD: we have used composer's full names in all previous tagging efforts, so let's do likewise now. For the sake of efficiency, I would suggest tacking it in brackets at the end of the TITLE: when you are trying to read the name of works on a small screen (such as a smartphone) you don't want it to read William Matthias: Si...but Sir Christemas (Wil... In other words, where screen real estate is limited, the name of the song is more important to read than its composer (I would suggest), therefore put the song name up-front and relegate the composer's name to the end. Although this might restrict the ability to display a composer's name, it will in fact still be searchable by most music player/managers search capabilities (so a search for all of William Matthias' work will pick up things tagged as ARTIST=William Matthias as well as things tagged as ARTIST=Compilation, TITLE=anything (William Matthias)).

5.0 Worked Examples

With those considerations out of the way, here's how I would rip and tag each of these CDs in turn. First, my go at the Britten Folksong's CD:

Here you can see another of my last-minute changes of mind: I decided to call this album just 'Folksongs' rather than 'English Folksongs', because I finally noticed tracks 12 and 13 are actually two Britten settings of French folksongs: it pays to be precise in these things! I've mentioned Peter Pears as the 'distinguishing artist', since he's doing the singing -but you could just as easily argue for '(Britten)', as he's doing the piano playing: that's another matter of personal preference.

Otherwise, this is just being treated as a single rip of an entire CD by a single composer: there are no track selections, since I want the whole CD ripped. There are accordingly no track offsets to worry about. The only change from the program defaults is my choice to make the CD be ejected at the end of the rip: since I'm ripping it in one go, I don't need the CD to hang around in the optical drive for second and subsequent runs of CCDR!

For the second CD, the CCDR responses look rather similar:

Again, it's being ripped in whole, with no track selections or offsets. The specific details, like year of recording, album name and so on, have obviously changed... but the most important alteration is that we've used Compilation as the name of the composer.

The completion of the tag data for each of these rips, using the CCDT program, is also quite similar. As usual, the performers, album art and track titles will need to be supplied, but all other tag data should already be correct. Therefore, once we launch CCDT, only options 3, 6 and 8 need to be taken in turn. Here's me tagging up the Britten Folksong's album. First, the performers:

Nothing too tricky there: it's following the normal conductor/orchestra/choir/performers order, with the minor detail that no conductor, orchestra or choir are present! Each performer's role is, however, clearly identified.

Next is the business of applying album art:

Again, nothing particularly out of the ordinary. I'd draw attention only to my use of select-copy-and-paste techniques to 'borrow' part of the full path to the album art from the square-bracketed path that was used the last time I ran CCDT. That allows me to then just type the 'folksongs.jpg' name in the yellow-text part of the screen: this is another example of exploiting CCDT's tendency to remember what values were supplied last time it ran to speed things up the next time.

The other thing I wanted to mention at this point, however, was what that folksongs.jpg file actually looks like:

You'll note that this isn't the album art that I showed you at the start of this article. That was what was actually shipped with my CD, but it's a generic picture of someone ploughing a field and therefore has practically no relevance to the actual content of the CD, except for the fact that the last track on it (and perhaps Britten's most famous folk-song setting) is called The plough boy! When I first got these songs on ye ancient vinyl LP, the cover showed Britten at the piano with Pears looking in from the right: it's more informative, and something I'm also more familiar with, through decades of long usage. So I decided to stick to the album art that made sense to me -yet another matter for personal preference.

Finally, here's me tagging the Folksongs rip up with proper track titles:

The usual 'yes, I want to alter' the generic track titles written for each track by CCDR is then followed by a simple typing of each folksong name, pretty much as supplied on the CD booklet shown at the top of this article. There's no particular subtlety about it, but there's quite a bit of typing to get all 21 tracks named correctly!

Updated to add (16th April, 2020) : Please note that CCDT has just been updated so that if it detects track titles in files which it knows to have been ripped by CCDR, it now will now not ask you if you want to edit or alter them, but will act as if the track title is blank. It will thus just ask you to supply a new track title and will auto-overwrite the generic title supplied by CCDR with whatever you supply. Given that CCDR's track titles are always 'generic' and meaningless, it seemed awkward to have to type 'y' every time you simply wanted to type a proper track title. The screenshot above show the old behaviour of requiring a 'y' response to the 'do you want to update/alter' question.

In the end, as I take the 'q' option to quit CCDT cleanly, it's clear that proper names have been applied sequentially to all the ripped tracks:

Very much the same approach is then taken with the other rip's tracks. The performers are specified in conductor/choir/individuals order:

And the album art is applied in the usual way:

The track titles get slightly more interesting, however:

The thing to point out here is that the individual track composers are being identified at the end of the TITLE tags. Where the 'composer' is indicated on the CD booklet as 'traditional arranged by....', however, I draw the line!

Arrangers are obviously important, but they do not get a mention in my music catalogue, I'm afraid! So, Bob Chilcott is not going to be credited in my track title for track 3. Along similar lines (though you won't be able to tell from the screenshot), I am not keen on joint compositions, so track 18, written (according to the CD cover and Wikipedia!) by Hugh Martin and Ralph Blane simply will get attributed solely to 'Hugh Martin' here. There's no particular rhyme or reason to that, except that my music catalogue is supposed to be a set of usable and meaningful music facts, not a comprehensive encyclopedia of truth. A double-composer in brackets at the end of an already long track name is just less functional to me than a bit of careful abbreviation of the facts!

The complete set of track names (displayed once I take the 'q' option to leave CCDT cleanly) is therefore as follows:

The presence of the composer names at the end of the track titles, whilst breaching normal convention that composer names go in the ARTIST and COMPOSER tags, nevertheless means I should be able to search for any of the composers listed in most music player/managers:

Here in the Strawberry music player, for example, I can type in the word 'Berlioz' in the search box and it will find the 'Compilation' composer and the one track by Berlioz on the 'Love Came Down' CD composed by him. When the entire CD contents are displayed in that same music player:

...I think you can see the album art is nicely displayed; the tracks appear in the correct order; the tagging seems consistent. A success, then!

6.0 Conclusion

The main thing to get out of this article is that whenever you are confronted with a CD containing a lot of tracks which are not intrinsically linked -whether by a single composer or not- you have to make a decision about whether to rip the separate works separately or whether to regard them in the aggregate, and rip the CD as a single entity. If you're going to rip as separate works, see the 'multiple compositions' articles in this series to work out how to proceed, depending on whether a single or multiple composers are involved.

If, however, you feel that the separate works only get their 'oomph' by being part of an aggregated collection, such that ripping them all as a single album is the more appropriate approach to take, then you have to handle the COMPOSER/ARTIST issue: if the entire CD can be attributed to a single, 'real' composer, then do that and assign an ALBUM tag that neatly and concisely describes the entire collection of works (which may or may not be what the music publisher stuck on the front of the CD booklet!)

Where multiple 'real' composers wrote different tracks within the aggregate, you then are dealing with a 'compilation CD' and should therefore attribute the COMPOSER/ARTIST to Compilation. At that point, you won't have a 'real' composer at the album level, so it's appropriate instead to mention the real composers in the TITLE tags for each separate track, in a way that doesn't compromise concision or search functionality.

Back to the Master Index of the Guides to Ripping and Tagging