Reflections on "Intuitive"

Maybe we should think of intuitiveness as the ability to get some sort of positive result in spite of the users lack of understanding of the software as a whole.

I’ll use this as a starter.

First, Ardour and Audacity have different purposes; and, while I use Audacity for a few things, the ‘record is a record+play’ behavior is one of my strongest dislikes. Intuitive, in the context of a DAW, should, to me at least, mean ‘familiar to those who have used professional multitrack recorders before’ not ‘familiar to those who have used CoolEdit before.’ Every professional multitrack machine I have personally worked with operates like Ardour in this respect, where ‘record’ is ‘record+pause’ and not ‘record+play’ and where tracks have to be explicitly armed for recording (for track safety, of course).

Ardour has extreme routing and tracking flexibility, but flexibility is a two-edged sword.

The meaning of inuitive is audience-dependent, as inuition is determined by an individuals prior experiences and their ability to observe those experiences.

Your entire post rests on some specific assumptions about what someone wants to do with a tool like Ardour. You are clearly convinced that the default purpose for a new track is to record audio, and experience over the last 10 years has shown that this just isn’t true. It was actually my assumption at first too, and early versions of Ardour completely ignored the possibility that people might have existing audio material that they wanted to use ardour to arrange rather than recording new material. When you throw in the fact that the other tools to which Ardour is most similar do not rec-enable new tracks by default (I’m not even sure its an option), and you can be sure that this is not a change we would make. If someone were to submit a patch to make this a user-controller option, I’d be happy to consider it.

One change that will happen in 3.0 (I actually though I had implemented this already, but it appears not) is an option for 1-click record (which makes pressing the global-rec-enable button start the transport rolling).

I think it is extremely important that you stop comparing Ardour to Audacity. They are no more comparable than Bias Peak and ProTools, which is to say that they each have very important roles to play, and they overlap a little bit in their functionality. However, they are not intended to be replacements for each other - Audacity would have to do an unbelievable amount of fundamental re-engineering to be able to do even half of the editing operations that Ardour makes possible.

@seeblade

  1. By clearly visible I mean must be visible on the screen, a menu item however logical does not meet this criteria. Here such an action must be a prompt as the user may not specifically understand the concept that they must add tracks before they can record. Making this visible acts as a hint or a clue that this is what they need to do.
    My suggestion would be to add a small strip in the track area itself that could sit just below the the bottom track that offers what right clicking in this area currently offers. This makes things slightly more intuitive for new users and I don’t believe it would get in the way of advanced users.

  2. On reflection I would suggest that you are right and instead the default template would come with two tracks come with a two armed mono tracks or a single stereo armed track. My first suggestion was overkill and probably would disrupt the work flow of most users. By having the first two tracks armed it should be easier for a new user to spot that newly added tracks need to be armed to work.

  3. The buttons could be grouped visually but present large enough target areas not to be hit accidentally. If we are very clever the buttons need not be adjacent but still grouped visually (for example a line drawn connecting the elements may be enough to help suggest that they should be used together).

I could perhaps produce some mockups of these ideas if you believe they might help.

Mockups are never a bad idea if you are discussing GUI changes, however I suspect in as far as the first suggestion, that you would be removing Real Estate for little benefit, and screen real estate is already pretty scarse. I am not sure I agree with the need for this myself, for whatever my opinion counts.

The third option a mockup wouldn’t be a bad idea on I don’t believe though. I am as of right now not for it or against it.

   Seablade

@Paul
I was looking at recording as a use case and considering what would happen if I forgot everything I knew about the software and started from scratch. In my case the first thing I would try to do is record something and unless I was very lucky or I read the documentation (which is something that I unlike most people I know actually do) I would have failed at Ardour gone and used Audacity instead. knowing what I do know would have been something of a Tragedy.

I haven’t considered what would happen if my very first task in Ardour was to mix. I suspect it might go a little better than my first thought experiment. It also might change some of my recommendations. (I will try it when I get a chance).

I am not suggesting that Ardour should become like Audacity or any other software. I like the flexibility and power of Ardour and I would not want that compromised. At the same time I can see see some places where new users may easily become unstuck. I think that adding some visual cues will go a long way towards helping this and done in the right way these should not hinder intermediate or advanced users. Also defaults may be changed to reduce the cognitive load of first time users however this may be a little harder to not get in the way.

@seeblade
Consider in the first instance that no extra screen real estate would be used. If there are no tracks then the space where tracks go is empty. As the track list grows the add new tracks item would remain at the end of the list so it would be the first item to disappear off the screen. It would then also be right next to where the new tracks would appear in fact the first newly added track would replace it position wise which would often be desirable.

The third option may take a little extra screen real estate. But I shall see if I can find a way to minimise that.

Consider in the first instance that no extra screen real estate would be used. If there are no tracks then the space where tracks go is empty. As the track list grows the add new tracks item would remain at the end of the list so it would be the first item to disappear off the screen. It would then also be right next to where the new tracks would appear in fact the first newly added track would replace it position wise which would often be desirable.

Ok my first impression based off this description is a good one I will admit.

  Seablade

@danni: as i’ve explain in many different contexts, i’m not really interested in promoting the idea that you can do audio engineering, with a computer or without a computer, without understanding what you’re trying to do. The cognitive load for first time users is important - this task is NOT simple, and software should not be trying to convince them that it is. That doesn’t mean it should be presenting unnecessary roadblocks either, of course. If there are places where Ardour does that, I want to know about that.

@Paul

your position is quite understandable and in large part I happen to agree with it. I am rather interested in the roadblocks myself and it is those that I am trying to describe.I think i particular that it is important that a first time users not be completely unsuccessful at achieving anything at all in the program. If that is the case such a user may assume that either Ardour is broken or it won’t work with their computer or that it is just too unintuitive. I think once somebody gets past that point they are more likely to put effort into learning how to use the software properly.

I could be wrong

@Danni: While I can see where you are coming from with your suggestions, I have to agree with Paul and Seablade. If you want to work an a large multi track project, recorded or otherwise, you would do one of the following:

  1. Use Ardour as you would any professional DAW, in which category it falls currently.
  2. Read the manual, or get a video like the one for Harrison Mixbus.
  3. Get an audio engineer to do it right

Driving a car is quite intuitive, but even so, it takes some lessons, and a licence before you are let loose on the roads. It boggles my mind why people think that driving a mixing desk or DAW that has hundreds of potential controls should be running at full speed in the first few minutes.

I am all for making things easier, but changing the fundamentals of a system control can make it less intuitive.
Think about it. If I were to change a basic control, like my computer keyboard to suit my language better, others could have problems with using it.

From what I’ve read here, we all want Ardour to be relatively easy for the starter without taking what Ardour is. And I think Ardour could be easier for newbies with the tools it already has. First time users need an easy starting point from where they can go further.

To lower the barrier, Ardour could provide templates, such as a “Beginner Session”, where there are tracks already connected to the inputs on the soundcard, maybe also channel templates for vocals, guitar etc. with basic plugins like a compressor, reverb etc. with basic presets already loaded. The timelines could be reduced to only the essential lines. This would lower the entry barrier considerably. If one provides a popup window with a link to the FLOSSmanual when loading a beginner session, it would be a good entry for starters, I think.

This could be a good compromise without altering Ardour and without hiding the necessary complexity, but at the same time making newbies familiar with it and give them an early success to think “I can do this”. And from my experience an early success is just the boost most users need to learn more, even if something is very complex. Ardour has already every possibility to be simple enough to provide an easy starting point, it’s only a matter of providing it. And maybe my suggestion could be an easy way to do that.

Oh, and while we are at the topic, I have one little thing I’d like to mention: I wanted to save a channel template in the mixer window and name it, but the little window for naming the template had only a “Cancel” button. I pressed “Enter” and everything was saved fine, but I think not having a “Save” button there could be a little confusing to some.

I built recently Ardour 3, and i was greatly impressed by the new MIDI pianoroll. You see directly the notes on the editor window and can set the range interactivly. Wow ! :slight_smile: You cannot make something more intuitive :wink:

There are others improvements that need to be noticed like the horizontal sliders for the FX Send, now you see them all at once and can fine-tune the amounts with ease.

The dialog box to route tracks/insert/bus/outs has been revamped (i have yet to dig into this).

And another little detail that says everything, i wandered myself to try the timestrech tool on a MIDI part and, at my big surprise and pleasure, it worked ! This is exactly the definition of something intuitive. Things react as you would expect.

All that to say that the perception of Ardour is going to change positivly anyway, and lot of frustration may vanish.

Paul and team, you are on the good route ! :wink:

Apple’s Logic is for me one of the most unintuitive pieces of software around. I would use it if I could, but after half an hour of using it I’m already exhausted with the over cluttered interface and silly icons, and some basic functions (like nudging) being hidden away. As far as DAWs are concerned protools and Ardour are the best I’ve ever tried (and I’ve tried more or less all the major ones). I’d say Cubase next. Ardour has the advantage of not being hardware dependent (over protools) and more flexible when it comes to multi channel stuff (you are not stuck with the standard 5.1 7.1 formats).

Hi,
Reading the posts at the beginning, i think the definition of “Intuitivity” of a new tool as “beeing able to apply prior in depth knowledge, learned from other tools of the same category, to the new tool” is a little week. For me “Intuitivity” is much more then that, and much more subjective: the “intuitive” interface of a new tool should not only reflect the internal logic of the tool but also that of the user, even if the user has NO PRIOR self experience with the category of tools (e.g DAWs, mixers ). I only demand that the user not only knows what he wants by using the tool, but also has an idea, of how such a tool works. The user should ask: “If I were the tool (the mixer), how would I do the job?”. I think most people looking for an “intuitive” tool unconsiously ask themselves this question.

I’ve worked for a long time with Cubase, prior to using Ardour, and I think both are extremely intuitive, Ardour even more. Prior to Cubase I had absolutely no experience in audio recording, either soft- of hardware. But I had a faint idea of how such a DAW could work, and within hours I recorded my first “song”. Ok then came the time I had to learn how to master my stuff, and I’m still learning. But the essence is that both tools reflected from the start on my internal logic. Other examples of “intuitive” software are, in my point of view, the incredible Reason, Hydrogen, freqtweak, QJackCtl.

Counter-intuitive are N.I Battery, Rosegarden (sorry Chris, but I love librubberband :wink: ) since it took a long time to get a tone out of them (with softsynths), and lots of LADSPA plugins, mainly due to LADSPA its self.

So let me come to the point. The reason why Ardour is soooooooo goooooood is the fact, that due to strong contact of the devs to the userbase, Ardour reflects the internal logic of the majority of users, regardless to the prior knowledge of the users. So even newbies get quickly acquainted with Ardour.
Having said that,
Keep up the good work Paul, Dave, …

We’ve had this discussion many times in irc, and seem to get to the same point each time. I make a workflow suggestion, usually involving adding a action to which a keybinding can be assigned, and one or several voices chime up with something like “I can’t see how that would be useful”, or more commonly, " well I wouldn’t do it that way, so do it my way, and you’ll be ok." That’s not always true, no matter the good intentions behind it. And to kill a myth here. I’ve used music related software since its inception, more or less, and have a cabinet full of apps (win and mac) which i cut my workflow teeth on. Even so, being openminded about workflow efficiency is not just the province of developers, but users too. (Yes, some of us have solid experience, and are still willing to take a fresh appraisal of workflow, even as crusty middle aged farts.)

Ardour has a lot going for it, and yet i still find parts of it unintuitive. A simple exercise is performing a series of actions as a sequence, then examining how FEW actions can be used to complete that sequence. I think most users would be surprised at how many hundreds of actions we carry out during a session, and for those of us who work at this for extended periods each day, in large projects, these numbers are important to deal with issues like hand and arm fatigue, etc…

For the sake of politically correct semantics, you could say “intuitive” is just a word, but i think the intent is a lot more common than one might think. Most users will probably a) use a mouse for just about everything, and b) never examine their own workflow to make sure they’re using the best the app can provide, for the greatest efficiency. That doesn’t mean we all have the same inclination, but i think it does go a long way towards a sometime feeling of reluctance on the part of developers, to take on board more complex code work, filling in the user’s perceived “gaps” in his or her workflow.

Incidentally, this discussion topic rages in the commercial world on a regular basis, and usually ends with a group of users fighting over which workflow tools are the “best”, at which point the devs either quietly disappear from the discussion, or make lots of non-commital but soothing noises, using the difference in viewpoint to do nothing.

I’ve been honest and forthright here, from my perspective, because i don’t think any app development benefits from hearing extremes, ranging from warm and fuzzy to something entirely less savoury, both of which offer little in constructive contribution. Ardour has many strong points, imho, and has some weak ones too, which in my use case is the less than optimal qwerty driven navigation, and the gaps in that navigation.

Having choice is paramount, imho, and when the user only has one, i.e. the mouse, then he no longer has control over the workflow sequence, and must resort to workarounds, usually with additional actions, presses, and clicks, as a result.

I continue to wish the Ardour team success now and in the future.

Alex.

Very nice post. It’s giving me serious thoughts after reading the post.

Intuitive = instinctive. I think that an intuitive application is one that you can learn without a manual. A large majority of us would eventually figure out how to have sex with a partner on a deserted island as response to certain stimuli, but no matter how you put it, making music using computers is not an innate ability. The skill has to be learned even by most seasoned acoustic/electronic musicians. Therefore I agree with all the people who tend to promote patience and paced learning. It is, however, frustrating to be at ground zero, like me. Life happens to everyone, and ti can be discouraging to not have the ability to just download the knowledge of software into the head within a few seconds. That’s why many things in life can’t be achieved without persistence.

I think you failed to understand a key part of my original post, which was precisely that there is no “intuitive” way to interact with an application like a DAW. There is only learned behaviour, which comes from previous interactions with similar or somewhat similar software. You yourself acknowledge that there isn no innate ability for this stuff - and just as that is true, there is no design that is somehow “instinctive” for the user. There are designs that will appear obvious to someone with some previous experience of suitable other software (and/or hardware), but that’s a very different claim to make.

I think we are all essentially on the same page. My reply is actually in full support of your thoughts. Your claim is very convincing, Paul, and your reasons for working on this wonderful project are very noble. Thank you!

I’m a pro musician who started on an Atari STF. Well, that was when I was a kid.

I’ve used literally ALL the DAWs outthere, bar some few you can’t really get to (MOTU performer?).

Anyway, got my two cents to throw in.

I am a big fan of Logic (as it was when there was a Windows version). I started playing with Linuxes around the same time that 5.5 version came out. Started with RedHat, moved on to MEPIS, then FreeBSD, then Debian, then back to FreeBSD, then Ubuntu, etc.

Anywaywhos, recently I’ve settled on Ableton Live as my sequencer of choice. It was due to pricing.
But there’s a story behind that that I think fits into this conversation.

When I was deciding what to shed money on, it was between Cubase and Ableton Live. (I have deserted Logic since it went Mac only, and if Apple don’t want to be game, it’s their fault. They have meddled with powers that be). The choice for Ableton was driven a little bit around its’ “intuitiveness”. But since then, I’ve changed my notion of “intuitive”. Ableton Live is a peculiar beast: it sacrifices customizability and functionality for their (albeit great) approach to simplicity. But simplicity and intuitiveness are not the same thing.

Also, if something has a setting, it is not related to intuitiveness. (sorry for my spelling). If you can customize it, and there are settings presets, it’s an endless debate about “defaults”. So that is off-topic to me.

I need Ardour. I need a DAW on Linux, even though I dual boot, I simply like to switch quickly between programming and making music. Ardour has been steadily improving over the years, and it’s come to the point where I can now see where it is going.

Now, Ardour is as intuitive as they come. Interface bugs, default settings and such don’t fall under intuitiveness. I liken Ardour to Pro Tools, which I’ve used only briefly, but that is an intuitive program. Intuitiveness is about the learning curve. And yes, no DAW should be simple (exception is Ableton Live, but it only has a linear learning curve, which is to be aplauded of course).

Ardour has become customizable in the right direction. I think at this stage it just needs to be polished, and we need quality plug-ins. The thing that’s missing the most is basically what ALSA and pulseaudio are doing. Personally, I’d switch to OSS, and scrap the whole thing, but ALSA has gone too far already.

I don’t really understand jack, I never cared for interprocess audio, but performance is quite essential.

Anyway, I find Ardour very usable, and it is intuitive in the sense that a lot of things have been simplified, like it working without jack, or autostarting jack - that’s was an ice breaker for me.

I fully support Ardour, and people who find it unintuitive (god, that word is hard to spell), then they should stick with, I don’t know… come on, garageband is not intuitive at all. Logic is. Just as Ardour. It depends. As I’ve said, simplicity and intuitiveness have nothing in common, neither do default options. Ardour is one of top 5 applications for Linux, the list including blender, digiKam, audacity, fontforge.

Good job!