Why don't plugin developers venture into hardware offerings more?

DSP, Plugin and Host development discussion.
RELATED
PRODUCTS

Post

I agree that money is the main thing for making hardware to sell to the masses.

Before software for the masses was possible I built many custom singles and some very small runs of small devices. Never had the nerve to look for financing to mass produce. At that time the lowest quantity price point for affordability was 1000 units and prices got more reasonable in quantities of 10000 or more.

Maybe nowadays there are different "minimum quantities" than in the old days.

The trad way to seek venture capital (or even a biz bank loan)-- Formulate a detailed biz plan with "hopefully realistic" marketing data. Then shop it around to deep pockets. Which is easier if seeking money for a pizza joint or auto repair shop, because the market and risks are better known.

Even building a thousand analog stompboxes, with nice paintjob, in a nice shrinkwrap cardboard box. Frightfully expensive for a small fry. Maybe nowadays you could sell em on the internet without fooling with dealers and distributors, dunno.

But back then, OK you spent a bunch of money and you have 1000 shiny boxes sitting in the living room. What, maybe roll a wheelbarrow of em downtown and peddle em on the street corner? Take an extended road trip visiting every music store in the nation trying to talk each one into taking on a couple of the boxes? Run expensive advertising? Concentrate making big sales to the big music stores?

Software has some biz costs of course, depending on the marketing plan. But just about any simple box you make, especially if it can attach to a computer or contains a clock, needs FCC testing in USA and presumably about any country requires the equivalent. When working with a company that made midi interfaces, it costed something like $5000 or $10000 lab fees to certify every new design. So far as I recall even minor circuit changes or chassis redesign was sufficient reason to have to pay another certification fee.

If it still costs that much then maybe it would be more profitable to run a testing lab than to build stompboxes!

Unless you can internet direct sell in sufficient quantity nowadays, the sales people are the most important people in the company because if the salesman can't sell then everybody in the company starves regardless how good a job they do.

Unless you can swing a deal with a distributor that will push yer products along with many others, you need to exhibit at NAMM every time which is expensive. A big distributor will always be at NAMM so it would save some of the expense but its a good idea to fly out a couple of yer own folks to help man the booth and talk up the products. Which is thousands of bucks expense every show.

Just lots of money to sell the boxes and lots of money to build the boxes. Back then I was friends with some rich folks and probably could have conned somebody into bankrolling hardware, but the odds of success so low. It is nice being friends with wealthy people. It is not so nice if you manage to get wealthy/powerful people pissed off about a bad biz deal! :)

Post

stratum wrote:
Hardware is very different, especially if it's using mains power. It needs more expertise and much more money.
I have seen hobbyists working with fatal voltages building pretty safe products for their own use. It's a skill that can be learned without a PhD. (not that I would recommend).

In practice ability to build hardware is mostly about money, and a lot less about skill.
I agree with you about the ability to field a prototype hardware - most people can learn how to use an oscilloscope, printed circuit board fabrication tools, and metal fabrication tools more easily than they can raise $150k for a small production run. However, the ability to create a production run of an electrical device with the required listings to be sold in the U.S. or Europe, the knowledge to create a supply chain, distribution channels, repair facilities, the ability to do quality control of the electronics in Taiwan and the physical device fab in China, and more is not the same as developing a prototype in your basement. It is also a very different animal than developing software. My lab has created a few spin-off companies that have hardware and software components...they're both difficult in non-obvious ways (e.g. the iPhone app portion must be continually re-compiled against new iOS releases to prevent API's from breaking), but of the two I find the hardware component much less forgiving of error.

You know how often developers release bug releases to fix small errors? It's much more painful to apply that fix by hand to your remaining inventory of 2,750 items from the first production run (because, say, an endplate wasn't attached with the proper adhesive).

So I'd say that while it's relatively for most EE's to create a one-off hardware device, it's far more difficult from a skill and capital perspective to create a company structure that designs, produces, and manages the repair of that prototype after it is re-engineered for manufacturability, price, and reliability.

Post

---
Last edited by Chapelle on Sat Oct 07, 2023 12:18 am, edited 1 time in total.

Post

Minimum number of units today is 1.

For example there's the Owl kit: guitar pedal case with input, output, four knobs. You load it with your own dsp algo, job done.

But the target is guitar players. Most synth guys rather keep everything in the daw....
Last edited by BertKoor on Wed Aug 16, 2017 11:21 am, edited 1 time in total.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

quikquak wrote:I see hardware these days as big empty plastic boxes with a chip stuck in the corner somewhere, running software anyway.
That was really the crux of my initial post. I didn't really refine it. I guess I'm thinking along the lines of the Roland Boutiques...

They are essentially just the plugins running inside of a cute box outside of the PC environment. And they made a killing on them. The JU-06 sold out and the JX03 is on its way to inflated prices on the resale market soon enough.

I just thought it'd be cool to have a box like that for a DSP based multi FX.

EDIT: but I understand now a company like Roland has near zero-risk doing this kind of thing lol

Post

Chapelle wrote:Aren't most effect plugins emulations of hardware anyway? :borg:
No. And even if they were, making software is probably the most risky and expensive thing you can do today. Very hard and risky work (for normal people with a wife and kids anyway).

It attracts naivety ("just needs a gui and..."), it's just too easy to underestimate the effort... :)
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post

FabienTDR wrote: making software is probably the most risky and expensive thing you can do today
O RLY?

https://www.youtube.com/watch?v=vvbN-cWe0A0
my other modular synth is a bugbrand

Post

lol

business wise I mean :)
Fabien from Tokyo Dawn Records

Check out my audio processors over at the Tokyo Dawn Labs!

Post

Very hard and risky work (for normal people with a wife and kids anyway,
Like this? (around min 0:30-1:30)

https://www.youtube.com/watch?v=5MgBikgcWnY
~stratum~

Post

FabienTDR wrote:
Chapelle wrote:Aren't most effect plugins emulations of hardware anyway? :borg:
No. And even if they were, making software is probably the most risky and expensive thing you can do today. Very hard and risky work (for normal people with a wife and kids anyway).

It attracts naivety ("just needs a gui and..."), it's just too easy to underestimate the effort... :)
In my experience, I really think I'd disagree with that.
[dreary story]
A colleague and I were thinking of developing a piece of software which we estimated would take about a year to produce (we'd been toying with it for a few months already). Other costs were actually pretty small. Unfortunately, the hardware became deprecated and in the near term our only real solution was to build the hardware ourselves. This wasn't a insuperable problem (my collegue is an electronic engineer) but we estimated this would add a year to the development. There was, of course, a significant hardware cost hike but almost everything was COTS so still doable. The thing that really killed it was testing which would have cost almost as much as everything else put together.
[/dreary story]

Post

Bump1 wrote:Just a random thought.

How many effect plugins need the processing power of upper tier CPU's??

I ask because I've noticed that multi-FX racks and/or desktop units have lost steam over the years.

Are FPGA's still that far behind in closing the gap or is it just not a viable solution? For example, I think a Waves hardware unit with a select suite would sell well.
https://www.youtube.com/watch?v=XTKYNUK-DJk

4x Quad Core aka 16 core ARM processor on this beast. DSPs are modular, so can be upgraded. I do feel sorry for the developer who has to port all the programming from fixed point to floating point though, it's not like it's a small amount of algorithms!

Post

FabienTDR wrote: No. And even if they were, making software is probably the most risky and expensive thing you can do today. Very hard and risky work (for normal people with a wife and kids anyway).

It attracts naivety ("just needs a gui and..."), it's just too easy to underestimate the effort... :)
Maybe not as steady income as a union plumber but I don't think software-for-the-masses is near as tough a way to make a buck as is hardware-for-the-masses.

Maybe both are about the same difficulty given a business that is already running at a profit? Expenses and complications of a bigger software operation would build up.

OTOH a hardware firm with established track record could more easily leverage capital for new product runs. The rather large routine capital expenses having become of routine nature, just part of the biz as with other capital-intensive businesses.

For the proper person, boutique hardware or software can make a good enough living. Maybe more income than a union plumber or maybe not, but maybe more fun than bending pipes anyway.

It is not real expensive to get into boutique hardware or small-time software and can be started out part-time while keeping a real job. At that stage it is much less expensive a money pit than having a fancy fishing boat or large recreational vehicle. Possibility of some profit and financial hazards too small to even risk sending a prudent person into bankruptcy. The company goes bust and you can eat the loss and still keep the house and car.

Transitioning from part-time software writing into full time decent income, perhaps paying several paychecks, could be capital intensive but not outrageously so. Maybe you could end up having to bankrupt over the sums involved but anything in the world is risktaking.

Transitioning from small-time hardware to mass manufacturing and sales would be lots more capital intensive, a much bigger gamble. Music hardware companies have similar marketing and cost of doing business expenses as higher-profile music software companies. So all those hassles are about the same but the hardware biz is MUCH more capital-intensive than software. Huge routine cash flow issues.

It is not impossible that a talented enough middle class programmer could make a self-financed success with no outside capital required. It seems highly unlikely a middle class hardware guy could do it without outside financing.

High rollers want to make a profit using other folks money. Just routinely run the biz hopelessly in debt. So maybe an ambitious EE would best find some obnoxious high-roller to do the messy stuff, and hope to survive the adventure.

What corporations are for-- So long as you follow the law, you can incompetently run a corp into the ground without putting shareholder assets at risk. You can blow millions of shareholder money, pay yerself enough to buy the house and car, then when the corp goes bankrupt you get to keep the house and car! What could possibly go wrong? :)

Post

JCJR wrote: Transitioning from part-time software writing into full time decent income, perhaps paying several paychecks, could be capital intensive but not outrageously so. Maybe you could end up having to bankrupt over the sums involved but anything in the world is risktaking.
Out of curiosity....what are the main contributors of overhead/cost when deciding to invest in yourself and venture into serious DSP and plugin development?

Post

Bump1 wrote:
JCJR wrote: Transitioning from part-time software writing into full time decent income, perhaps paying several paychecks, could be capital intensive but not outrageously so. Maybe you could end up having to bankrupt over the sums involved but anything in the world is risktaking.
Out of curiosity....what are the main contributors of overhead/cost when deciding to invest in yourself and venture into serious DSP and plugin development?
Opportunity costs and sleepless nights hehe.

Post

ARM Cortex-M4F (and above) based devices are quite nice to code.
Doing fixed point arithmetics (see SMMUL SMMLA) along with floating point calc is really easy (1 cycle int to float conversions).

These days, I'm porting to the Axoloti some "differentiated waveforms algorithms" i experimented with during my thesis back in the 90s with no noticeable issue.

I think that most of the existing DSP C code can be ported with no big deal.

About the hardware business, the ARM processors are game changers.
See you here and there... Youtube, Google Play, SoundCloud...

Post Reply

Return to “DSP and Plugin Development”