Lets Talk Tuner ... and then a lot about M4L

[TomViolenz]

I for one don't care about M4L in the least. Even the pre-made solutions are heavy on the CPU and have unnecesary limitations. I guess I could have fallen back to a: "ah well, that's what other people use" position, if they hadn't decided to realease new Live devices like the Convo reverb and even more gratingly the LFO and Envelope Follower as M4L devices instead of native ones.

Now I need to load heavy M4L if I want to modulate anything AND on top of it lose all other control of the LFO controlled parameter.

Useless garbage

[Angstrom]

I think the thing that really doesn't make any sense is how they pretty much abandoned all their UI standards when they decided to add more or less an object oriented programming language to their super simple intuitive user interface™.

I really have no idea how this pairing made any sense in terms of their original selling point VS the big dinosaur DAWs like Logic, Cubase, DP, Sonar etc. ??

So now the selling point is something like, "You can get started right away making loops and writing melodies in Session View, and if you want to delve deeper into the program, take a course in Max/MSP and after a year you can write a basic synth as powerful as the original VSTs from 1996!"

Seriously there's an uncanny valley between Live's original intuitive interface and Max 4 Live. Compare this to some relatively simple problems you might run into in DP, Logic, Cubase etc. To get up and running learning 90% of what Cubase, DP, Logic do would take you about a year, in Live, you can get the basic program in a couple months, but to use Max 4 Live at 90% will take you several years. I see how it appeals to certain people, but not all of us are programmers.

I agree 100% and I spend all day programming for a dayjob, so it's not like I can't extend a class or figure out variable scope. It offends me because its so wrong.

The problem comes when people who love Max can't understand the problem of putting an IDE in a music app. They think the naysayers are too daft to program, and so just want "simple" so say "well here's an LFO we made", and "here's a Midi echo device". They can't see it. Cannot grasp it. It's invisible to them. "But look at BEAP, or maxforcats, they made a modular inside the app inside the app! Isn't that good" . No. It is not good. It is ludicrous.

Live is a modular synth, it has audio generators, modulation sources, filters, resonators, delays, waveshapers. Live has all the components of a massively modular synth programmed by experts to have the best audio quality and CPU usage. In Live you can stack and route audio sources and processors, and route inputs to outputs.
However you cant do sub-device routing of those modulators, and audio. You cant use Operator's envelope on the ringmod, you can't run the resonators through Operator. As a massively intuitive polyphonic modular environment it nearly, nearly works. All it needs is a mod-out button. An "audio from". All it needs is a parameter routing facility. Mod-out to destination. It was nearly nearly there! They brought in macros and the routing panel. But then ....

So what do Ableton do? they say "throw all that away, forget everything you have learned about our interface and our instruments and easy intuitive creation, if you want to send operator LFO to the ringmod now you must make your own operator, and your own ringmod, and try to connect the ports, good luck sucker."

Bitwig got it right, Ableton got it wrong.
Sure M4L is powerful, but did that effort pay off for 80% of the users wanting to connect A to B to C to A?
Nope. M4L was a mistake. It was an un-needed side road. A distraction.
That is unless you are one of the 200 people loving making buffershufflers and controller scripts, or selling packs. Just loving their IDE.

[stringtapper]

The problem comes when people who love Max can't understand the problem of putting an IDE in a music app. They think the naysayers are too daft to program, and so just want "simple" so say "well here's an LFO we made", and "here's a Midi echo device". They can't see it. Cannot grasp it. It's invisible to them. "But look at BEAP, or maxforcats, they made a modular inside the app inside the app! Isn't that good" . No. It is not good. It is ludicrous.

…

Sure M4L is powerful, but did that effort pay off for 80% of the users wanting to connect A to B to C to A?
Nope. M4L was a mistake. It was an un-needed side road. A distraction.
That is unless you are one of the 200 people loving making buffershufflers and controller scripts, or selling packs. Just loving their IDE.

Let's not get carried away here with the IDE pejoratives.

Max is and always has been from day one an environment for making music. Period.

Maybe not for making the kind of music you make or the way you like to make it, but made for music-making nonetheless. I'm sorry that it was "your DAW" that happened to be the one to have Max integrated into it.

I know that you don't like Max because you tried it a few years ago and it didn't click with you like SynthMaker did. Fine. Doesn't mean it's not a useful tool for some of us or doesn't offer lots of possibilities for users who don't code in Max.

[Valiumdupeuple]

"But look at BEAP, or maxforcats, they made a modular inside the app inside the app! Isn't that good" . No. It is not good. It is ludicrous.

Live is a modular synth, it has audio generators, modulation sources, filters, resonators, delays, waveshapers. Live has all the components of a massively modular synth programmed by experts to have the best audio quality and CPU usage. In Live you can stack and route audio sources and processors, and route inputs to outputs.
However you cant do sub-device routing of those modulators, and audio. You cant use Operator's envelope on the ringmod, you can't run the resonators through Operator. As a massively intuitive polyphonic modular environment it nearly, nearly works. All it needs is a mod-out button. An "audio from". All it needs is a parameter routing facility. Mod-out to destination. It was nearly nearly there! They brought in macros and the routing panel. But then ....

I love M4L but I 1000% agree on this. And OSCiLLOT (which is an impressive piece of Max patching anyway) is the most relevant demonstration of how absurd and inefficient it can go.
I hope that they won't stick with this strategy for too long, it sure won't do any good.

[stringtapper]

I doubt M4L is going anywhere, so don't hold your breath there.

Yes, Live needs better routing options all around. Ableton continuing to develop M4L doesn't preclude development of more modular features in Live. If anything the limitations that people attempting to code innovative M4L devices are running into could be a big motivator for them to make those kinds of changes.

[andrewbrewer]

This discussion really grabbed my attention. It's only recently that M4L has really started to frustrate me. Like many of you I loved the concept originally, but now I find myself facing quality control and version management problems (I don't build M4L devices, I just buy them or use freebies). Stupidly I find myself fondly remembering a time when ableton was walled off from this complexity. I had the prudence to stay within boundaries then.

Now, well, it is very tempting to integrate M4L devices because some are very very creative, and the convo reverb is incredible. So I get to this weird place of either being disciplined and not using what I paid for; or submitting to temptation and enduring the hidden cost of instability. That last sentence reads "lose/lose".

More recently I'm also remembering Gerhard's speech years ago about addressing ableton's stability problems. I have to think that, today, it would be impossible to seriously guarantee that standard of quality with M4L in the mix, unless something very major changes (like "official" M4L devices and packs being placed through the same rigorous bug testing trials that base code most likely endures). That sounds impossible to me. Too many variables.

[stringtapper]

^ Just curious. Windows or Mac?

[andrewbrewer]

Windows

[stringtapper]

Yeah, that seems to be the trend. It's not surprising if you know a bit about the history of Max. It was originally written to run on the early Macintosh and then the NeXT system. There wasn't even a Windows port until 2003 I believe. Those who created it and have continued to develop it over the years are firmly rooted in the academic/electroacoustic music scene, so that Mac preference is not surprising at all.

[login]

I also think they should keep both directions, there are many audio/visual artist that use Max and M4L for custom devices its a very good solution to have it integrated in to a DAW.

But Ableton has to keep developing devices that enhance the modular approach of Live: native LFO's, modulation between devices and so on with out the costs of M4L.

[Machinesworking]

The problem comes when people who love Max can't understand the problem of putting an IDE in a music app. They think the naysayers are too daft to program, and so just want "simple" so say "well here's an LFO we made", and "here's a Midi echo device". They can't see it. Cannot grasp it. It's invisible to them. "But look at BEAP, or maxforcats, they made a modular inside the app inside the app! Isn't that good" . No. It is not good. It is ludicrous.

…

Sure M4L is powerful, but did that effort pay off for 80% of the users wanting to connect A to B to C to A?
Nope. M4L was a mistake. It was an un-needed side road. A distraction.
That is unless you are one of the 200 people loving making buffershufflers and controller scripts, or selling packs. Just loving their IDE.

Let's not get carried away here with the IDE pejoratives.

Max is and always has been from day one an environment for making music. Period.

Maybe not for making the kind of music you make or the way you like to make it, but made for music-making nonetheless. I'm sorry that it was "your DAW" that happened to be the one to have Max integrated into it.

I know that you don't like Max because you tried it a few years ago and it didn't click with you like SynthMaker did. Fine. Doesn't mean it's not a useful tool for some of us or doesn't offer lots of possibilities for users who don't code in Max.

It's not about whether Max is useful or a musical tool, it's that the whole concept of Max is completely counter intuitive to the whole concept behind Live. It's a pairing of complete opposites. Some of the things that make M4L difficult in Live are things Live programmers did to ensure stability in Live. The MIDI 128 limitations, no SysEx, overtly intuitive interface designs, no hassle routing, basically my grandmother could loop a little song together in Live, and I'm lost in Max. The difference is vast, and not at all about whether Max is 'good', it's about whether it made any sense to hoist it on us as a solution. That a guy who gets paid to teach people music is OK with it doesn't make it any easier for a DJ who wants to start writing music to learn it.

The world of difference is vast here, Max 4 Live sits next to dumbed down orchestral libraries, simplified embedded AAS etc. plug ins, and Gigabytes of looped musical phrases for DJ's to piece together and call a song. For you and others into programming, it's cool, but for even people like me who've been using other arguably more advanced sequencers like Logic or DP it's a freaking weird and unnatural pairing. I mean can you really grasp how odd it is to see an advanced object oriented music programming tool paired with a DAW that has about as simple of a MIDI implementation as you can get? and years later none of that has changed much. We can all wish that Live got improved MIDI to advance M4L etc. but we can see with the updates to 8 and now 9 that it's simply not happening.

Don't get me wrong, I'm happy you and others like it, it's just for all the development that went into making M4L work in Live I would have rather had a fucking LFO and modulation implementation that was native and as easy to use as Bitwigs.

Plus Cycling 74 readily admitted to sucking at implementing plug in architecture into Max, that was their main excuse for discontinuing Pluggo, that it was too much work to keep up with hosting Max in plug in format. It's not any surprise that M4L adds instability to peoples set ups with them needing to work with other third party embedded plug ins, Ableton code, and two OS's. The whole idea of multiple layers of embedded intellectual property just screams drawn out beta testing phases, and slow implementation of new features. I mean when you have two large code bases hosted by two different companies that share most but not all code, you're going to have more issues than when you have one developer and one code base, it's just basic math really.

It's all just tears in the rain, but I definitely think it would have been a wiser choice to not spend as much effort on embedding M4L and to go further on their own path, but it's not my ship, I'm just a passenger.

[re:dream]

How much extra burden does M4L put on the system?

For example. I often want to slap a whole bunch of LFOs on different tracks. But my impression is that this eats into my CPU burden. (I am not sure whether this is so, haven't tested it). Has someone been able to quantify how much extra CPU burden an M4L device adds, as opposed to a native device?

[TomViolenz]

How much extra burden does M4L put on the system?

For example. I often want to slap a whole bunch of LFOs on different tracks. But my impression is that this eats into my CPU burden. (I am not sure whether this is so, haven't tested it). Has someone been able to quantify how much extra CPU burden an M4L device adds, as opposed to a native device?

That's easy to test no?!

Just put 10 or 20 of these LFOs in your set and compare just the non running LFOs numbers to before.

Now do the same with Utilities or SimpleDelays etc.

The differences are staggering even on a i7 MBP.


I'd use non running devices as the comparison because it would obviously be hard to quantify the difference between what a Utility has to do and what an LFO does to the audio processing in a running system

But that could be tested too: Just use in one experiment Operator/Sampler etc with their in-built LFOs and then assign the same thing to the M4L LFO with the same settings.

I haven't done that test though.


I avoid the M4L LFO regardles of CPU use though, because of the limitation that I will lose control over the LFO controlled parameter via other means.

In my opinion releasing the LFO as a M4L device was either extreme lazieness, because they just repackaged existing M4L devices (if that), or an attempt to push M4L on users, who would otherwise not use it.

With the added "benefit" that you now need Suite to have an LFO in Live.

[re:dream]

Yes the LFO pretty much sucks canal water. It should be small, light on resources and flexible.

[stringtapper]

It's not about whether Max is useful or a musical tool, it's that the whole concept of Max is completely counter intuitive to the whole concept behind Live. It's a pairing of complete opposites. Some of the things that make M4L difficult in Live are things Live programmers did to ensure stability in Live. The MIDI 128 limitations, no SysEx, overtly intuitive interface designs, no hassle routing, basically my grandmother could loop a little song together in Live, and I'm lost in Max. The difference is vast, and not at all about whether Max is 'good', it's about whether it made any sense to hoist it on us as a solution. That a guy who gets paid to teach people music is OK with it doesn't make it any easier for a DJ who wants to start writing music to learn it.

The world of difference is vast here, Max 4 Live sits next to dumbed down orchestral libraries, simplified embedded AAS etc. plug ins, and Gigabytes of looped musical phrases for DJ's to piece together and call a song. For you and others into programming, it's cool, but for even people like me who've been using other arguably more advanced sequencers like Logic or DP it's a freaking weird and unnatural pairing. I mean can you really grasp how odd it is to see an advanced object oriented music programming tool paired with a DAW that has about as simple of a MIDI implementation as you can get? and years later none of that has changed much. We can all wish that Live got improved MIDI to advance M4L etc. but we can see with the updates to 8 and now 9 that it's simply not happening.

Don't get me wrong, I'm happy you and others like it, it's just for all the development that went into making M4L work in Live I would have rather had a fucking LFO and modulation implementation that was native and as easy to use as Bitwigs.

Plus Cycling 74 readily admitted to sucking at implementing plug in architecture into Max, that was their main excuse for discontinuing Pluggo, that it was too much work to keep up with hosting Max in plug in format. It's not any surprise that M4L adds instability to peoples set ups with them needing to work with other third party embedded plug ins, Ableton code, and two OS's. The whole idea of multiple layers of embedded intellectual property just screams drawn out beta testing phases, and slow implementation of new features. I mean when you have two large code bases hosted by two different companies that share most but not all code, you're going to have more issues than when you have one developer and one code base, it's just basic math really.

It's all just tears in the rain, but I definitely think it would have been a wiser choice to not spend as much effort on embedding M4L and to go further on their own path, but it's not my ship, I'm just a passenger.

Those are fair points. It you look at it only from the standpoint of production then it would have made a lot more sense to make Max for Logic.

But even from its inception Max was meant to be a live performance tool. So in that sense its pedigree is not so incompatible with Live as it was originally envisioned: a "sequencing instrument."

I have always been in the camp that wishes Ableton would focus more on the live performance aspect of Live than the production side. But I don't see M4L as being incompatible with that vision even though they haven't done the work to make sure it's not crashing Live all the time.

For the record I hardly use any pre-rolled M4L devices (the convolution reverse is nice though). I make my own devices that solve little problems or allow me to have more control over how I make noises than what I can do (or at least do easily) with Live's native devices, and I rarely have any crashes.

Horses for courses. I think it's a badass sandbox inside of what was already a badass sandbox.