ETI style for take numbers

Tags: #<Tag:0x00007f75756a5378>

I’m adding a historic release that includes take numbers on its back cover, formatted as:
“It’s a Lonesome Old Town -2”
“Honeysuckle Rose -1”

My first impulse was to remove this from the track titles; they’re not part of the title, and in the liner notes only the title itself appears. However, there are multiple takes of the same title, so removing them entirely would be confusing.

Per ETI style, these should go in parentheses:

“It’s a Lonesome Old Town (2)”

That’s a bit ambiguous as to what that number means, so I think it’d be more sensible to take it one step further. This is a format used on some actual releases:

“It’s a Lonesome Old Town (take 2)”

Anyone have thoughts on this? Again, it’s a reissue of 1920s material, so artist intent doesn’t apply.


FYI the release is here; I have used the “(take 2)” style for now.


Would this be more like the ETI style for parts? Or like the (alternate mix)?

It’s a Lonesome Old Town - take 2

It’s a Lonesome Old Town, take 2

It’s a Lonesome Old Town (take 2)

The brackets style is clearer. It is more like a remix \ alternate version of something. Not a “part”. And the dash does look a bit vague but is what appears on covers.

I was working on a Charlie Parker release that has similar multiple takes. This was listed on the cover with dashes: and

I’ll watch this conversation and tweak that if needed. Especially the recording level.

I certainly agree that this is vital data. Just like the (mix) details on remixes. Otherwise someone comes along later and merges them in error. I wouldn’t be surprised if a second and third take can generate a very similar AcoustID and length.

When making a style decision like this, I consider what this looks on the artist’s recording page ( ) and the ability to spot the different recordings on that page.

1 Like

This as it’s a version information. :slight_smile:


For the recording itself, I would rely more on the disambiguation comment as you can fit more detail in there than you’d want to see in a title (see here for some examples). With this type of recordings, they have almost always been on multiple releases, with different detail on the cover - take numbers, “alternate take”, or no distinguishing information at all.

So that brings up a followup question. Suppose this release didn’t include the take numbers on the cover at all. From the liner notes I can identify which takes are which. Would it still be okay to label the tracklist with “(take 1)” and “(take 2)”? Now we’ve moved from standardizing format to adding information.

1 Like

I’d say the track list should be as close to as‐written as possible, but name the recording with the " (take N)" (and maybe even also add it to disambiguation). If the track list doesn’t mention this, then it’s likely not that important for that specific release/track list, so IMHO shouldn’t be included.

@Freso in this example the cover is clearly showing -1, -2, -3 so it is showing that the take numbers are important. Seems to make sense to take that to the more compliant level of (take 1) instead as the text looks weird otherwise.

That slight tweak is like when we remove the ALL CAPS that those covers are written in.


is replaced with

It’s a Lonesome Old Town (take 2)

Or that is how I have understood this question.

I really prefer the second example with (take 2). I think it adds clarity.


I’d just like to second that. :slight_smile: If I had seen that back image without having read this thread, I wouldn’t have the slightest idea “-2” means take 2. It seems like a pretty rare notation to me.

1 Like

@IvanDobsky, @Freso was responding to my hypothetical followup question, not the original one.

I think he’s correct that alternate takes are usually indicated in the tracklist somehow, when they are important (in particular, if there are multiple takes of the same title). I’m pretty sure I’ve seen exceptions to that, but none come to mind.

However, there does seem to be general support for the idea of standardizing the format of take numbers when they are present.