Audacity Forum

I've updated the version in the previous post to enable it to accept a comma as the decimal separator.
Unfortunately it is not currently possible to make this for appropriate locales only but I think that it should work OK.

steve wrote:

Gale Andrews wrote:Wouldn't have a clue what the "Label placement value" is without looking at the Manual.

Label Placement method: Number of labels
Label Placement value: 10

Guess how many labels there will be.
For a real user, they can try it and probably will not be surprised when it creates 10 labels. Are you saying that you would be surprised by 10 labels?

Perhaps not as an experienced computer user, but that is really unfriendly to a novice, IMO. You need C++ to make that friendly.

The image I had "no clue" about, in http://forum.audacityteam.org/viewtopic ... 86#p233186 had a method of "Label Interval" and a value of "0:1:30.0". And the time to place first label was "0:0:0.0".

As an experienced user I can probably guess that "time to place" is showing me one of the stated formats (hh:mm:ss but including tenths of seconds)

So by extrapolation is "0:1:30.0" a label every 90 seconds?

I too would "like" to trim down the controls, but if it needs to be as hard as that I'm not so sure.


Gale

I've seen you've now made it more friendly.


Gale

Gale Andrews wrote: "Label Interval" and a value of "0:1:30.0". And the time to place first label was "0:0:0.0".

I agree that is likely to be confusing for many users. I think we'll need to wait for type 4 plug-ins before we can properly implement hh:mm:ss.

steve wrote:

Gale Andrews wrote:As I said, to me that shows 2 labels in the selection with a final (extra) label.

Under the old (weird) scheme, yes you would have to specify 2 labels + one more.
Under the new scheme it is three labels.
Just looking at it, it is OBVIOUSLY 3 labels - they are even labeled 1, 2, 3

Yes but only two of them are inside the selection and will affect the export

At least I think we agree the old scheme is weird / wants changing. Changing to what may be the challenge.


Gale

Gale Andrews wrote:but only two of them are inside the selection and will affect the export

That's making the assumption that the whole track has been selected (which in the example it appears to be) and that regular interval labels is a precursor to Export Multiple (which it may not be) and that no more audio exists or will be added beyond the current selection (which is unknown).

Gale Andrews wrote:At least I think we agree the old scheme is weird / wants changing. Changing to what may be the challenge.

I'll see if I can figure something out that is less weird

steve wrote:Case 1:
You want the "label regions" to fit. [Fit interval]
3 labels.
No label at the end because the next label region would start and extend from the end of the selection.
Label Track "C1 Point" shows this with "point labels" (zero duration labels).
Label Track "C1 Region" shows this with "region labels" (duration greater than zero).

Case 2:
You want to fit 3 labels into the selection. [Fit Labels]
There is a label at the end because we are "fitting the labels" inside the selection.
Label Track "C2 Point" shows this with "point labels" (zero duration labels).
Label Track "C2 Region" shows this with "region labels" (duration greater than zero).

So are the two options:

• Case 1 "Fit interval" AKA "fit the label regions"
• Case 2 "Fit labels"

a straight replacement for "Adjust label interval to fit length"?

Are case 1 and 2 both when choosing the "Number of labels" method?

Case 1 "Fit interval" conveys very little to me. Is it the same as what happens now as long as you don't ask for a final label?

Case 2 looks quite fringe to me. I would never want it, but what would be the use cases for Case 2 Region? That there could be a high chance something required to be deleted or kept would be at the end?

Which is going to be default - Case 1 or Case 2?

Will there be any cases where the last label in Case 2 lies after the end of the selection?

Thanks,


Gale

This is what I mean by "Regions" and "Intervals".

All three of these label tracks have the same "label intervals". The time from the start of one label to the start of the next is 5 seconds in each case. The "label interval" is 5 seconds. The current selection indicates the extent of the "interval" between the second and third labels (same for each track),

The first Label track has point labels. The "label region" has zero duration,
The second Label track has "region labels" the length of each "region" is 1 second.
The third Label track has "region labels" the length of each "region" is 2 seconds.

An example using the "new scheme".

Here's an 8 bar phrase from a dance track.

I selected from the start of the phrase to the end of the phrase.
I wanted the bar line for the end of the phrase so that I can see where the next phrase starts. This is the start of the 9th bar. I manually renamed it "bar 1" because it is also the first beat of the first bar of the next phrase.

I used "point labels" because I only want to mark the "bar lines".
Because I knew that the selection was exactly 8 bars, regular tempo, I did not need to work out the bpm or anything like that. I just needed to know that I wanted 9 bar lines (point labels), spaced regularly and exactly fitting the selection. Here I am fitting the bar lines (labels), so there is one at the start of the selection and one at the end. The final label being the down beat beginning the 9th bar.

The settings with the new scheme are:

• Label Placement: Number of labels.
• Number of labels: 9
• Length of labeled region: 0 (or empty)
• First label text: bar 1
• Label adjustment: Fit Labels

Thinking out loud....

For creating labels at regular intervals, the decision whether to include a label or not is simple - either that position is within the selection or it isn't. If the label position falls at 3 minutes and the selection ends at >= 3 minutes then the label is included. If the selection ends at < 3 minutes the label is not included.

The problem comes when fitting a specified number of labels. The user may want n label intervals (for example for Export Multiple) or they may want n labels with the final label at the end of the selection. Either way I think they should get the number of labels specified.

The first label track fits 3 label "intervals" into the selection.
The second label track fits 3 label positions to the selection.