PRaaS - Plate Reverb as a (web) Service

Shop deals on ebay
shop deals on amazon
Advertise with us
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
D

dharma one

Active member
Joined
Mar 24, 2005
Messages
40
Location
London UK
I've been toying with this idea of an automated online hardware EMT140 service

Imagine the following:

1) You register on the site, post dropbox links to the files you want to process, and pay via paypal/cc

2) The web app downloads the dropbox links, automatically runs them through the effect of your choice (EMT140 in this instance), saves the results and sends links back to your email.

If the service cost a few bucks per track, bandwidth and storage costs would be easily covered and there would still be enough profit left to run it as a viable business. Just needs a beefy pc to work as a server, a good sound interface, Arduino hooked up to EMT140, running 24/7. If it's automated with server side scripts doing the work, there are no labour costs involved and price can be kept low

Reverberation time parameter on the EMT140 could be controlled via javascript with Arduino and breakoutjs.com or similar.

The idea could be extended to work with other hardware gear too.

What do you guys think? I would definitely use the service and I think others would too. Anybody got a spare EMT140/Plattenhall and node.js/arduino/linux chops to make it happen? :)


 
Don't think the ROI would cover building a reverb chamber, would probably be easiest to do this together with a struggling studio who already have the gear/reverb chamber. The jobs could be queued to run overnight/during downtime if users are OK with a 8-10hr latency :)
 
that's exactly what it is, except these guys seem to be using a manual, not automated workflow (with the cost being $100/track vs 5$/track)

 
Maybe.  I have two plates myself, and sometimes in a hurry and/or probability the client will come back with a mix tweak I go with a virtual plate. 

I envision a bunch of people sending you impulses to harvest your plate sound, then it's shared everywhere as a set of impulse responses.  This may be a market that requires a premium charge to suggest value. 
 
There are plenty of EMT140 impulses out there, many of them free. If impulses sounded like the real thing, no one would use the real thing. They don't sound the same - and same for virtual plates, the sound is about 70% there. Stereo imaging and interplay of frequencies doesn't match the real thing.

If it was a manual workflow, a premium charge would be appropriate to compensate for labour. The point of automating it is that it takes zero time and effort from the plate/studio owner beyond initial setup, and the price can be pushed to a point where users don't think twice about spending a few dollars on a real plate reverb sound - and you get larger volumes of use. Scaling  isn't an issue either - whether you have ten users a day or hundred users a day doesn't matter, since it's all automated


 
Kinda a no brainer to spend some extra $$ for the real thing, on some projects I would just use plugins, but once in a while you get an artist where the more organic everything is the better...spaced mics, reamping things in different parts of the studio..stairwells, basement...
This would be a great service!! And I agree, there is nothing like a real plate, especially for those special songs! Great idea.


Like any good fix, great thing about automation is FAST turnaround. That is probably key, if people know they can upload files and get their files back super fast, they will probably be that much more likely to do it.
 
If the plate reverb was reserved for this service 24/7 and not used for normal studio work, turnaround could be pretty fast if there is zero queue - the length of your track basically, plus a few seconds for emailing the download link back.

If it became extremely popular, you'd have to queue the requests, or buy more plate reverbs to deal with the demand, since one plate can only process one track at a time, and every request takes the length of the track to process.

Maybe the pricing should be time based like AWS EC2 instances - $2-4 for each minute of your track.
 
When I bought my ecoplate I thought about doing something like this. Didn't think of the Arduino, great idea.
There might be a great opportunity to create a turnkey system for this - the server, arduino, etc...
Sell the system to the old great studios for their reverb chambers.
Imagine if you could send your vocals through the same reverb chamber the beach boys used? That would be cool.
I also think a quick turnaround would be key to success.
 
Just playing devil's advocate.  I agree there's no comparison, I'll also guarantee there are plenty of people with plates who use impulse responses, even of their own plates, to streamline workflow on projects that require lots of picky client mix tweaks, which seem to be most projects these days.  As well at a certain point laziness is the overriding factor.  I say those who don't care a lot can't be coaxed no matter how inexpensive it is, they won't bother uploading/downloading/etc.  That's going to be most people, in my experience.  It'll only be the few who care to bother, and many of them will already own plates.  Can it really be self supporting at such a low price?  It's an interesting question.  Go for it. 
 
I'm up for making this happen if someone wants to collab.

I can design/write the web front and back ends, package the software/hardware to a product and market it, need someone with a plate and EE/Arduino/node.js experience to collab with and drive it forward

emmr: I agree people are lazy. I don't have a plate or a space for it, but would pay a few bucks for a real plate on the final mix vs. impulses, especially on arrangements where you can really hear the plate. So I figured there might be more people like me. Dunno, maybe there aren't and it's a niche product. Maybe the pricing is too low, but for hobbyist users (where you get most volume) I think the price limit is a few bucks. Enough users, repeat custom - I think it'd be self sustaining (until physical modeling algos can reproduce the plate 100% anyway:)

If it's painless enough and pleasant to use, like other modern SaaS, I reckon it'd be an interesting proposition. Simple web based front end, no ftp, press a button to upload (takes a minute or two with broadband), get a result back in 10-30mins - I'd use it.



 
Maybe there are other use cases that aren't viable for hobbyist/small studios where you'd get more demand - well mic'ed self playing Yamaha Disklavier grand - upload midi file, get 24bit piano audio back :) You could even get a live webcam feed of the piano playing itself for added entertainment value
 
business case guys... before you go spending time and effort.

What's the potential market, whats the unit price, what's the effort required.

Don't bother charging $5 for something unless your getting minimum 200 clients a month.

(sorry to be a realist... it's just that there's way too many people who invest way too much money in stuff without checking if they can actually MAKE any money from).


(Says Rochey who has an idle pick and place machine)

/R
 
back in the old days we used be able to do ISDN connections into the chambers at capitol, I forget if it was real time or not. MY issue with processing this stuff is control functions like wet/dry or in the case of a plate volume level of the return. It would be cool if you could do it via real time which in the fast world of modern I am sure there is a way to make it happen

 
pucho812 said:
back in the old days we used be able to do ISDN connections into the chambers at capitol, I forget if it was real time or not. MY issue with processing this stuff is control functions like wet/dry or in the case of a plate volume level of the return.
Click to expand...

A non issue with a reverb, you get back wet only and blend to taste.

 
The biggest issue is damper setting,  or in the case of some Ecoplates, EQ setting.  When the client decides they want shorter/longer decay, pay for another pass, try again tomorrow? 
 
That would be the idea, Rent a certain amount of time with the plate and do it via a java (or whatever) screen control, hearing the fx in mp3 real time, then record the high rez through and download.
 
Rochey - good points, appreciated. I think the pricing definitely needs to be thought about to at least recoup the invested dev time. Maybe there are only a handful of people who care about a product like this

Implementing a realtime preview with controls and mp3/video feed isn't a problem - but giving users an indefinite session time and charging in the end might not be ideal - someone could hog a session for hours and log off without paying in the end. Plus there would be no way to schedule multiple users because their session times are not known in advance - you would have to wait in a queue until your turn is up. Perhaps this wouldn't be a problem initially with just a handful of users.

Another way could be-

1) Upload wav, specify "tweaking time" length and/or default decay setting, take payment, schedule session (cost of the session=length of wav+user specified length of tweak time)
2) User gets an email reminder before when their realtime tweaking session is up. If the user is not present when their session comes up to tweak the settings, the reverb gets rendered with whatever decay length they specified at booking time.
3) Link to final render gets emailed to user
 

Similar threads

ciscan.81
FIRST DIY BUILDS: JLM BAs & 1290 Complete!
Replies
8
Views
9K
fab672000
F
joe-electro
MIDI Controlled Vari-Mu Compressor
Replies
11
Views
6K
gyraf
gyraf

Latest posts

Back
Top