DSP austerity in 2018

Say you want a kick, independent hats and snares, add a 4 note chord and boom , no more channels and no more voices. Buy another synth if you want melody and bass or a stacked sound.

Completely off topic.
We are discussing here digital synths and DSP, and where it is at in 2018.
Nothing more to add other than I do like the Digitone, will probably eventually buy one when OB is released and the system is stable.

2 Likes

Haven’t looked inside a DN yet, but the microprocessor in the DT is significantly slower than the one in your PI.

There’s no dedicated DSP chip in the DT and I’m sure there’s also nothing like that in the DN.

2 Likes

My speculation for the 8-voice poly

  1. Elektron has been spartan with the voice counts ever since the octatrack was released. They do not seem to want to let go of the tracker scene roots, and want to impose tracker-style limitations/features to their products

  2. IIRC Simon said the DN runs internally @96kHz - I bet we could have 16 voice poly if it ran @48k

But I am quite sure there would be more poly if elektron wanted there to be more… Somehow, they want us all to use these like boxes like trackers (?)

Personally I am ok with 8 voice poly / 4 part multitimbral. The DN doesn’t have indiouts anyway, I’d much rather have 4 outs than more poly tbh, this same thing eats me with the A4 mk I.

The sad fact of the matter remains - even if MNM from over a decade ago was technically superior to the DN, I bet the DT and DN will outsell the ole MNM by leaps and bounds. Elektron has grown up now and is also in this to make money, and more of it. It is far more lucrative to sell cheaper, more limited units. The age of Hector is over, its not about being a superuser anymore. Now its all about intuitive use and fast workflows…

PS I love my new DN

3 Likes

Necessary in regards to what ? DT is only 8 voices, yet I always have 1 or 2 voices unused in my patterns…

Computers can have plenty of voices, yet I never found stacking more voices does any good to my music.

I used to have an MPC Live, which can play a lot of voices… yet it failed to keep me inspired.

More voices is good, and I’m glad my DTone has twice the voices of my A4, but it’s not that important for many use cases.

Interesting view… which company is progressive then ?

A modular system will cost you more than 2 Dtones. A lot more if you go by the number of voices. It doesn’t make sense in this context really. And I do own a 12U eurorack.

Well… I do remember the days of 128 poly grooveboxes… that was great !
Oh, wait… no it wasn’t ! These were bad romplers and sounded like sh**.

We now have instruments/toys that sounds very good… and prices went down too. Things do progress.

1 Like

ditto

3 Likes

Ditto @tsutek @acidhouseforall

To get back on the subject, food for thought: http://www.newelectronics.co.uk/electronics-technology/theres-still-growing-demand-for-dsp-say-experts/117710/

If you read closely, you’ll quickly realize that the dedicated DSP market has indeed moved into a direction where it makes less sense to use off-the-shelve DSP components in electronic musical instruments.

Yeah, and with this new NAMM announcement, according to which, new music gizmos will soon be able to use off the shelf IT parts and still have a realtime OS worthy of music making, it would seem that dedicated DSP is getting closer to being extinct, with the exception of some industrial/military systems perhaps…

No, not at all. They’ll just be used mostly for always-on, low-power, high-efficiency tasks in produces with volumes high enough to spend the additional development cost of using a DPS versus a general-purpose MCU.

In terms of audio, you could think about phones or “smart speakers” that are always listening if they can recognize a voice command while their main processor is powered down.

Keep in mind that DPSs are not limited to audio processing in any way. They’re also widely used for motor control, visual processing, cellular systems, etc.

It’s just that what’s available right now, and what is going to remain available in the near future is optimized for uses other than high-power, high-performance audio, and that it makes more sense to use a general purpose MCU that’s more widely available and that supports open development toolchains you can easily run on your workstation without having to spend a lot on up-front license fees, or without having to buy parts in very high volumes.

1 Like

Hence mpc live / juce.

Design choice nothing else, the idea is work within the limitations of a limited feature set, I get that some people click with that and others don’t, luckily there are plenty of other options, and in a way that is the hard part, finding a set up that works for you, deciding where and if and how many limitations you want, I don’t imagine guitarists or piano players have such dilemmas, maybe they do?

Not an apologist, but I tend to try to look at “what is” rather than “what should be” when choosing gear, it has served me well but even so I still sometimes struggle with my setup, although it is often about removing rather than adding.

I rarely ever use more than 3 note chords, or have complex polyphony as I quite like minimalism so for me it isn’t an issue, however I also don’t think for example that 16 voices would have prevented me form enjoying a synth that only has 8 voice poly, then of course there is the DAW which effectively allows unlimited polyphony and unlimited options, but for me that doesn’t work.

I would definitely encourage anyone of an open mind to compose a whole piece using a set of very fixed and limited options, like for example only using 3 sounds, or make a whole song using just a monosynth, it can be a great way to get out of a creative rut and its fun and allows a certain freedom that unlimited options do not.

TL;DR Because that is how it is :wink:

6 Likes

I would love a new Mono Machine, with sampling tracks, individual physical outputs, and 32 sequencer tracks. Essentially a new octatrack + including monomachine hybrid system. I would be fine without overbridge, if it would allow to sample to an SSD or USB stick.

Elektron please do it. Make it with an expansion slot, so you could sell us expansion modules, and make money with that continuously. (Expansion modules for additional MIDI controllers, expansion Modules, for additional Interfaces, expansion for additional synthesis etc.)

3 Likes

Sorry that I didn’t read every post so far, so I may repeat something, which has already been said.

IMO it’s a design concept defining, what an Elektron unit is supposed to be. Having my dark trinity I would describe the commonalities as follows:

  • specific sound creation machine(s) per unit, (sampling, digital/analog synths, melodic, rhythmic)
  • step sequencer supporting parameters in the hundreds with many bells and whistles
  • mono-phonic or reduced polyphony per track (also AR does have 4 voices only),
  • a groove-box in the first place, one task, easy enough to operate, switch it on, let’s rock …
  • optionally control of external gear (MIDI, CV)

The intention seems never to be a “Music-Production-Center” like the MPCs, or one box to rule them all, or one box delivering multi-timbral sounds in polyphony like expensive workstations, and there seems to be no intention to replace a DAW with a box.

TBH … I appreciate this one-box-one-purpose concept … and yes … sometimes I use my OT like Ableton, and all boxes together like a workstation … but not as a replacement, only as a setup with many complex features, which I can have OTB and use single or in combination.

If I imagine having so many functions inside one single Elektron unit (MPC-like, workstation-like), but having the same user interface, I wouldn’t think, this would be easy and fun to use.

5 Likes

Thank you, my point exactly, just much more patiently typed!
If people don’t like the digi pair don’t get it, let people who do get on with them!
The dichotomy of this forum is hilarious, half the threads are elektron bashing, the other half are fan boy gushing!

I clicked in fear of being rick rolled

2 Likes

Every forum I’ve ever been hanging out at has its pet peeves. After a year or three your autofilter kicks in and you stop registering the most frequently re-occurring noise… it becomes invisible. That’s the BBS blues I guess.

I don’t remember my MC909 sounding like sh**.

2 Likes

I thought they don’t even use DSP chips, just a single CPU?

My understanding is that because Elektron doesn’t get the stuff mass produced in China, but in relatively small quantities in Sweden, they have had to cut a lot of corners (e.g. use very low res screens, low polyphony, limited RAM ) to make some profit from selling the machines. The hardware is obviously high quality made to withstand a lot of use, including live shows. Living standards and wages in Sweden are very high so the people doing the assembly are getting paid much much higher wages than factory workers (practically wage slaves) in China.

Or maybe Elektron CEOs are trying to buy themselves a private island and some nice super cars. :no_mouth: