Dude you’re giving too much false info about the AR
Yup. Already have. 
3 Likes
I think Arturia are releasing something today as part of their 25yr celebrations
Hopefully it’ll be a disappointment mk2
4 Likes
Florian will be delighted 
Introducing the Disappointakt
Features:
Can only be turned on once. Ever.
Comes with pre selected samples. Samples may be added but must be in the inaudible human ear range.
Play button means turn off. Press stop 5 times for normal play behaviour.
Samples can only be manipulated via ai prompts
Comes with edible instruction manual.
Compatible with overbridge on windows xp
Premium users get the ad free version
20 Likes
Why would anyone feel disappointed by a machine that yet does not exist, that no one is obligated to buy, in an completely full of other options market?
1 Like
There’s a significant difference between what humans perceive as complex, and what is complex for computers (especially limited computers like embedded microcontrollers or DSP chips) to do.
Some of the things you may be perceiving as “complex” new features—e.g. the new Euclidean sequencer mode—may actually consume very little resources compared to an LFO. An LFO has to be computed in real time for every tick of the sampling rate of the machine, as well as its effects on whatever it’s modulating. Computing a Euclidean sequence requires a much lower time granularity, and moreover you can handle it even more efficiently by pre-generating a truth table of all possible permutations of the available Euclidean parameters, and then just doing a simple data lookup from the table when you need a pattern. Then all you need is enough memory to store the table, which can be quite tiny, and frees up CPU cycles for other things.
That brings us to another reality: embedded systems like this have multiple kinds of resources—CPU, memory, the bandwidth of the memory circuits, the processing power and memory of various subsystems such as DACs and USB, etc. When a hardware system like this gets initially designed, there’s typically some advance planning for future growth in firmware updates, but the designers can’t always anticipate what updates will be needed and what resources they will require. You may think you’ve given yourself enough room to do everything the users might ask for after 1.0, but sometimes the users surprise you by wanting different things, and you have to adapt within the limitations you have already set yourself. It only takes exceeding the limit of any one of these resources for a feature to become technically infeasible—e.g. you may have tons of free memory but be all out of CPU cycles, or bandwidth to communicate with the memory.
TL;DR—when the developers of embedded systems tell me a feature is technically infeasible, I am inclined to believe them. It may seem like a simple task to me, but they are the experts in the limitations of the specific system they have designed.
16 Likes
I’d go with emotional system associated expectations, crash of reward system feedback loop and such. The whole thing why PR and advertising exist in the first place
didn’t wanted to sound like the thread police, cause I bite my lip quite often when I come here. As I expect some interesting new thoughts or maybe even some serious info
and ill slap myself later for repeating myself now with this:
but I really think synth manufacturers should have a close look at the concept of gaming devices. since its very hard to come up with new hardware every 2/3 years in order to get people interested and money in, they should create devices that are used as platforms for additional software, that has to be bought (like video games)
I don’t think I would like that myself, but I find this a good survival strategy for bigger companies with more employees. Ableton is kinda doin it already with push. But the market could be more diverse and interesting
-im out
6 Likes
Yes, I understand and know all that. The difference being personal adjustments or settings of the automatic biological judgement systems between different persons. I cannot change the fact how it works and what the result is. I know it maybe wrong, after all, it’s a black box (unlike the digitakt which is a black box).
Elektron follows this advice, so is Akai. Don’t they? They establish a platform and improve software and software titles over time while leaving everything else untouched. Like good old C64 ans Amiga 
Edit: ah - I missed the part with pay per function. Nah, Elektron is trustworthy because they don’t do that. People pay money because they can rely on several years product maintenance and software development including gradually introduction of new features. The TR-8 mki buy 606/707 was not good and Roland dropped it with TR-8S.
but they don’t get direct money from updates. it maybe motivates people to buy the device
but just imagine machines that cost 99 euros
its very subjective, but I never buy sample packs. but I would think about buying a machine that offers new effects or methods
2 Likes
3 Likes
Noise Engineering has gone that direction with their newest lines of Eurorack modules. Each module is actually one of three standardized underlying platforms—Versio, Legio, or Alia—and you can swap in any of the firmwares for a given platform to change it to whatever module you need it to be.
Really cool, and I hope more manufacturers follow suit!
2 Likes
yeah its cool, but again the software is free
EDIT:
I actually have a kxmx blumechen that sports a daisy seed
so, I can just put the software on that device
I paid 150 bucks for it
but its a lot to fiddle with and not super user friendly
You don’t have to run an LFO the same speed as the sampling rate as the machine. And i bet it doesn’t (does not go into audio space too far). It’s writing a register. And this must be read out and added, multiplied, whatever with the inputs of the parameter which yields a value maybe per sampling slice. That could indeed add load. But since the thing accepts midi CC IIRC, the whole thing could also be achieved by routing an LFO with slow sample rate to CC internally. Or define a max of LFO. And if you don’t use it in one track, you can use it in another track. Etc. Doesn’t matter. The new Sid machine is much more valuable to me than a second LFO could be. Whatever the reasons, I like their paths and results.
Yeah, I think that’s a difference in the Eurorack market vs standalone devices. In Eurorack, it makes more sense to charge for additional faceplates instead of additional firmwares, because if you need two oscillators or VCAs or LFOs instead of one, you’re gonna need that extra faceplate.
For standalone devices, charging for the software would make more sense. For example, the little 1010 nanoboxes appear to be a standardized hardware platform, and it would be awesome if they would decouple the firmware costs from the hardware. So you can pay $300 to get a hardware box, and then $100 for whatever firmware you want to run on it.