What's been going on with Ardour?

There hasn't been much news posted here for a while, so I thought it was appropriate to update subscribers and other supporters of my work on Ardour on what has been going on. Development efforts have ben split (about 60:40) between Ardour 3.0 and continuing work on the 2.X series, both to fix bugs and to support the continuing improvement of Mixbus.

3.0 has been through some incredibly deep changes over the last 2-3 months, some of which I'm reluctant to describe because they involve arcane details of programming and would bore those of you who choose not to track the changes via the commit mailing list. One exciting (and deep) change has been the move towards having the Ardour history list just record changes to things like Playlists, rather than recording their before and after state. This might seem like an obvious way to do things, and indeed, its been in the "todo sometime" list for ages. But as usual Carl H. jumped in and got the ball rolling, and it became so clearly the right approach that I took over and flushed most of the old approach away. The result is that the history list now takes up a lot less memory - with big sessions or sessions that have hundreds or more regions in a track, the impact is very dramatic.

This work went along with a top-to-bottom revisit of the undo/redo mechanism with the goal of making it scale properly to operations involving large numbers of regions. The results? Operations that were taking an absurd amount of time (40 seconds) to undo can now be undone in less than half a second. The overall responsiveness of undo/redo has now greatly improved.

There has also been work to improve the usability and interactivity of the "Strip Silence" dialog (still not finished) and a new Monitor/Main section in the mixer, which provides the kind of control over the final step of delivering audio to your studio monitors that is generally only found on high end consoles (Ardour3 supports Solo-in-place, Solo-in-front, After-Fader Listen (AFL) and Pre-Fader Listen (PFL) as alternate solo models). Features there include per-channel cut/invert/dim along with realtime dim and solo boost controls, global dim and cut controls and more. I'm currently considering whether to make this feature a default in all sessions or continue to make it available as an option for those who really need it. You can of course tearoff this monitor section into its own window.

Work on 2.X has mostly been focused on a few specific bug fixes, and continuing work to improve Ardour's support for AudioUnit plugins. Mixbus has shown us that, unfortunately, the AU specification is nowhere near as "rigid" as it first appears, and there are plugin manufacturers out there doing many different things in their plugins. The general pattern seems to be that they make sure that their plugin works with Logic, and consider the job done. This means that we have to make Ardour copy Logic's behaviour even though we can't directly know what Logic's behaviour is (since we can't read the source code). Over the last 2 or 3 months, a lot of work has gone into this effort, and there are now only a small number of plugins that we know about that do not work perfectly with Ardour (and Mixbus, of course). There are still some general UI issues with keyboard entry in a plugin editor/view, and with those rare plugins that choose to resize their editor/view based on some kind of user action.

If you want to follow Ardour development "as it happens", please do subscribe to the commit list - even if it is all just nonsense to you, it will give you a feeling for the changes that are taking place. As thing stands now, I hope to reach the next stage of Ardour 3.0's release plans within a few weeks, and have another (2.8.8) release of Ardour out within the week or so. Thanks for reading ...

WOW ! Looks like Ardour 3

WOW !

Looks like Ardour 3 will ROCK even more than I hoped for! These performance issues are the last obstacles for me to consider Ardour the millenium DAW :)

Benjamin

Paul, Thank you for posting

Paul,

Thank you for posting this article as it was needed for us who have been patienly waiting on the A3 arrival. The changes to A3 from what you have described seem amazing and not to include the MIDI capability. The best news to us here following this great project is finally the possible release of A3 in couple of weeks.

Thank you for your dedication and hard work that you and the rest of the team put into one of the best (if not the best) open source application. KEEP IT UP.

Rony P.

Good job! I'm checking for

Good job! I'm checking for updates here every day. This application is awesome and I'm looking forward to Ardour 3!

The Ardour 3 release will be

The Ardour 3 release will be a great event. :)

Paul, i don't know if you have considered it already but, what about a switch that allows a track to be connected either to its input or to its "tape". I read that its a feature of the so called "in line" consoles. This is a flexible mecanism when u have a recorder connected to a mix console (doing some live mixing for a start and switching to the tape after recording a track and keeping your EQ settings). In Ardour of course, both devices are bundled together, but we don't have yet this switch that let you set the monitor routing independantly for each track.

For this reason i often add an insert on each tracks on use the bypass button to switch from the direct input or the recorder material. As very low latency is now possible reliably, it's engaging to use the computer as a live mixer (specially with a big multichannel souncard).

Any opinion ?

Cool. So with the new

Cool.

So with the new Undo/Redo approach, could you potentially "remove" certain edits from the history, while keeping edits done since then? For example, you run a normalisation filter over the audio, then cut a bit out of the middle, could you then undo the normalisation without undoing the cut? Maybe a bit excessive, but might have some cool uses.

Can the Ardour have any input into the AudioUnit spec? Even if that's not possible, it might be worth putting out some public comment on it, detailing possible improvements - Apple can then choose whether to ignore it or not, but AU developers might be interested.