I would suggest starting by going through the manual and understand the basic concepts of how Ardour handles files and the terminology used.
This is a good first page to read: Ardour terminology page
And then jump to the section on working with regions: Ardour working with regions page
This is the key concept to grasp:
“Although a region can represent an entire audio file, they are never equivalent to an audio file. Most regions represent just parts of an audio file(s) on disk…”
Ardour will never separate an existing file into multiple files, but what you can do is mark sections of a file that you want to be distinct by making new regions (or splitting one large region into multiple smaller regions), and then export them separately when you are finished editing, processing, mixing, etc.
If you really just want to take an existing large file and split into multiple smaller files with no processing, there are better tools for that job than Ardour, which is primarily designed for recording, editing, and mixing audio in a non-destructive way (i.e. the original file is never changed), and generating new output files which are the end result of that editing and mixing.
If by “splitting” you mean the split to happen only while exporting the project to audio files, then you can mark each “movement” as a “range”, using the “Range Markers” ruler. Then, at export time, you can select those ranges on the “Timespan” tab of the export dialog: Ardour will create separate export files (.wav, .mp3, whatever) for each range, using the range name as part of the file name. See this page in the docs.
Hi Chris,
Thank you for your reply. I quite take your point re. separating the actual track itself into movements. I should have been more specific, as my query was always about exporting the marked movements as new tracks.
I am used to working in Wavelab (also non-destructive), which does this very easily with the “Autosplit” facility but of course, Wavelab is a Windows program, & in trying Ardour, I was after a full featured program which would work in Ubuntu Linux.
I have managed to do what I wanted to do using Range Markers to mark each section & then exporting them via Timespan.
Thanks for your help. Actually, after much delving, I managed to work out this method as you have described!
I was a bit surprised at the number of steps necessary in Ardour to perform what, to me, is a common & simple procedure, compared with Wavelab (using Auto-split), which I work with in Windows.
Ardour, at 1st experience seems very strong in recording, mixing & editing, however.
Depending on what steps you took there may be some more efficient methods. For example, there is an option to select either playhead or mouse as edit point. If you select playhead you can listen to the audio and just press s to split at the points you want. If you are looking for silence you may be able to see that in the waveform display, and with mouse as edit point you can just put the mouse cursor at a point that looks like silence and press s to split the area into two new regions.
Hi Paul,
Thanks for your question. I’ll try & explain with the following example:
I have recorded a program from FM radio, say, & this was a Beethoven symphony with the usual 4 movements. I have a WAV file of the complete broadcast imported into Wavelab, so after I have trimmed any unwanted bits & am left just with the music, I want to save the program as 4 movements in my music library.
In Wavelab, I find & mark each section. There are several types of markers in Wavelab, but for this purpose the CD markers are the most convenient. The file as displayed in Wavelab will have CD start & CD finish markers, so I just have to add CD track markers in between each movement.
Having done this Autosplit comes into play. When I start the Autosplit dialogue, it will have a screen for what marker to read in order to split the file (it is non-destructive, so it will be copying). Then it will request file names for the files it will generate after the split operation. This can be pre-prepared text from the clipboard if you want to give exact names, otherwise it will just number them, requiring re-naming after. Your choice. It will also allow you to choose the folder you want the new files to go into, AND the file type (WAV,FLAC, Orbis, mp3 etc.)
After you hit OK it will copy the required data & deposit the files in your chosen folder.
It is a very smooth operation, whereas, in Ardour, it is a bit clunky in comparison - taking into account the rest of Ardour’s functions, which appear very powerful.
Wavelab is, of course, expensive commercial software, so one should expect it to be good, but one needs deep pockets!
BTW, the markers in Wavelab are stored in a .MRK file with the same name as the original .WAV file. There is no interference with the original file by Wavelab.
Hope this ramble explains it in a helpful fashion.
So IF I am understanding your use case, here is how I would approach it:
I would start by stripping silence:
Note that you can specify both a threshold and a minimum length.
Ideally that creates regions according to the ‘movement’. I would then select all regions on tracks, right click, and ‘Add Range Marker per Region’ which will create a range marker per region which the regions are hopefully made to the correct length via the strip silence command above.
Then I would do Session>Export>Export to Audio Files. In that dialog I would just select the ranges that are now made the length of each region, that was made by stripping silence. This should give you the end result you are looking for I hope. If you find settings that work particularly well for strip silence, I would bet it can be scripted via Lua to create a basic script you can make with one button click.
I used to do similar with two track recordings destined for podcast where I would export sections to different files for different purposes. Haven’t done it in a while, I think since Ardour 6 and the workflow has shifted a little, but not to bad.
But just going totally back to basics for a moment: I think you need to be slightly clearer on one very fundamental point:
Wavelab is an audio file editor: it’s purpose is to allow you to carry out edits on a file and save the result to either a new file or the original.
Ardour is DAW: although it does the capability (via Session > Export) to create new audio files that reflect the edits you have performed, the basic workflow is not intended to be “load file, edit, get new files”.
Wavelab is even advertised as “mastering software” which is a whole different thing, conceptually, than what most people would do with a DAW.
The reason the “auto-split” operation is not directly present in Ardour is that in general, you would use “split at ” as merely one small step in producing a new piece of audio, and there’s no obvious reason why that step would create new files (a step that in a DAW is reserved to be a distinct, separate process than editing).
Hopefully a few other users will offer suggestions for how to get close to the Wavelab workflow you are used to – as Seablade noted, scripting can almost certainly replicate most of this.
Ok, I seem to be wanting to do things that Ardour was not designed easily to do.
I am not all that interested in the differences between DAWs & Mastering Software - I just want to do simple audio management at this tme of my life. 20 years ago the recording, editing, and mixing side of things would have been just what I wanted!
I don’t like Audacity particularly, & I thought Ardour would be worth investigating. I can do what I want to do, but it is much easier in Wavelab. A pity really, as I’m tryng to move away from Windows.
Anyway, thanks very much for your help & information, & all the best to you & the other developers of Ardour.
Thanks for the reply. I’ve managed, after a bit of delving, to do what I wanted to do.
Stripping silence is not really necessarily a help, as many classical pieces (particularly opera) require sections within a continuous file to be separately “tracked” without gaps,. (not the right terminology in Ardour, it would seem!)
A tool is a tool. I wouldn’t use a screwdriver for putting in nails. If wavelab does the job, fine. I also use a Windows software for my tax declaration 'caus there’s nothing on Linux.
Oh I am aware (I worked with a noted opera company for years), I was thinking classical orchestra when I wrote that which is why determining threshold and length become important. When I did this before I just did it manually with the play head as edit point, found where I wanted to cut, and hit ‘s’ to create region bounds. At some point especially in more classical genres things are better done manually and more accurately. The point is once region bounds are defined it is fairly easy to export multiple files based off those regions.
By the way if you are monitoring the recording don’t forget you can drop in markers as your are recording as well by a simple keystroke to mark the movements.
There may be a way you can follow a similar approach in Ardour. There is an intermediate step in that Ardour does not produce individual files for each CD Marker range on output. What is does is to produce a CUE file which is used by CD manufacturers to indicate the start of each track.
However, there are tools which can easily use this to create multiple tracks, found from the following post:
If you are comfortable with command line tools, shnsplit (in the shntool package) looks to be a good option. It should be able to set this up as a post-export command in an export profile so that it runs and splits the file automatically.
Not a problem, as you mentioned Opera might be a bit difficult, especially live opera, but a long enough minimum length and it may work, not sure, haven’t tried that one. I suspect that it should work for many classical pieces though.