Easiest - least effort way to load impulse response in a hardware effect?

[thawkins]

So we all know that the newfangled thing is to cheat and not buy real oldschool tube gear or effects VSTs or whatever, but just pass some sine waves through it to capture an impulse response and use that instead.

From what I understand, the IR is just some audio stems that kind of encode an EQ response so that you can capture the "sound" of some hardware in a way.

Coincidentally that's pretty much what I need for doing room correction EQ. Usually this means that I should run the correction VST on some laptop or whatever runs the software correction. But if I can just capture the IR, then it should be possible to load it into any piece of junk that can do ADC, DAC with some magic in between.

So to my question - what's the smallest, cheapest (while not being terrible) way to load an IR onto a piece of hardware, hook it up to the monitoring chain and forget about it until you need to update the IR or bypass it somehow?

I am sure all the fancy audio interfaces already do this, but I don't really care about all the other fancy stuff this gear does, all I want is to put a signal through the IR-based EQ correction.

Can the Axoloti Core do something like this? Something more insane like a custom built FPGA board?

[TubularCorporation]

The cheapest room correction hardware I know of isn't cheap, it's the MiniDSP and it's still something like 350 euros.

 

Probably the cheapest way is a Raspberry Pi with a Hifiberry DAC and audio input over LAN.  But there might be some latency so I bet it would be no fun to monitor on.

 

Find an old SM Pro Audio V-Machine and run ReaVerb (the free convolution plugin that comes with Reaper and is available standalone for Windows) on it?

[TubularCorporation]

For more effort but less money, an Axoloti can do this, and probably with decent latency.

 

EDIT: I definitely read about a newer, more powerful Axoloti-like DSP platform last year but I can't remember what it's called.

 

EDIT 2: https://www.zrna.org/akso

 

Also sold out, though.  Might have to grab one when I can at that price, though.  Use it for audio and dedicate my old Axoloti to MIDI event processing.

Edited April 7, 2021 by TubularCorporation

[TubularCorporation]

Easiest but not cheapest would be something like this probably:

 

https://reverb.com/item/36008983-yamaha-srev1-2000-black

 

Not even trying to be helpful at this point.

Edited April 7, 2021 by TubularCorporation

[TubularCorporation]

Here's a really long thread about an open-source DSP based processing platform aimed at pedal builders:

https://www.diystompboxes.com/smfforum/index.php?topic=114354.0

 

If you don't mind having PCBs manufactured, doing a bunch of surface mount soldering, and possibly coding your own convolution algorithm (I haven't seen how far the project actually went) then it's an option...

Basically it's the whole "easy, powerful or cheap - you can only choose two" conundrum. I've looked at hardware convolution off and on over the years and I still haven't found anything that ticks all three.

 

Maybe there's an inexpensive guitar pedal out there that has stereo i/o, accepts line level, has a cabinet simulator (i.e. a simple convolution filter) that allows user-created impulses, and isn't expensive.  There should be, but is there?

[TubularCorporation]

I guess you could use two of these:

https://www.amazon.com/SONICAKE-Speaker-Cabinet-Simulator-Response/dp/B08DNBHGWX

Or two of these:

https://www.aliexpress.com/item/4000707337251.html

 

I have a feeling using a different digital pedal on each channel wouldn't treat your stereo image well, though.

 

EDIT: this would do the job but it's as expensive as a MiniDSP:

https://www.yerasov.com/catalog/other-products/iron-cab-impulse-response-cabinet-simulator

Edited April 7, 2021 by TubularCorporation

[thawkins]

Actually those guitar pedals look like something I could use. It's not necessarily something I would instantly put in my monitoring chain, but for broad strokes room correction for a living room listening situation it might be good enough.

I wish there was a stereo pedal with some ADC/DAC that looks like hi quality. Not sure why I don't think a 50 bucks pedal can't have a high quality signal chain.  

[TubularCorporation]

That Iron Cab one is probably fine but it's also too expensive to make sense for room correction because in that price range there are better options.

 

The MiniDSP really is the best be as far as I've ever been able to see, but cheap by room correction hardware standards still isn't cheap.

 

https://www.minidsp.com/

 

EDIT: actually, this isn't too bad at all. Probably not the most amazing converters at that price but still might be passable:

https://www.minidsp.com/products/minidsp-in-a-box/minidsp-2x4

 

I'd only really looked at the HD version, and that's $200USD for just the board with no enclosure.

 

I have to refresh my memory on these things, actually.  I can't remember if you can configure their outputs to work as a, A/B speaker selector, because if you can then I might have to try to get hold of one of the HDs in the next year or so.  Being able to use it as a speaker selector, room correction, monitor source selector AND free up the pair of outputs I'm using for monitoring right now actually makes the price seem pretty fair.

 

EDIT AGAIN:

 

Never mind, I remember now why I never went for one of these - only the DIRAC models (which cost twice as much) allow impulse response loading. $200 for a firmware update = no thank you.

Edited April 7, 2021 by TubularCorporation

[TubularCorporation]

The other thing is that doing it in software is almost always going to sound better becuase of one less generation of conversion, so for hardware to even be worthwhile you'd have to get something that had some kind of digital input (or worked as a USB audio interface) AND had converters at least as good as the interface you use now.  And those unfortunately seem to start in the $400 US range and up.

[thawkins]

Well, I wanted a small correction box so I could stick it between speakers and the vinyl turntable / aux cable / airplay device in the living room so I would not have to run a laptop or a computer in between and would not have to set up a correction EQ thing on whatever is actually playing the audio, be it a phone or something else.

So for this scenario, the MiniDSP looks ideal, except I am not sure how easy it is to load a "standard" IR into it - from the description it works with some RoomEQWizard format. This is purely a convenience thing as I think I can use my IK ARC 2 MEMS microphone with RoomEQWizard as well, it's just one more calibration thing to learn.

[TubularCorporation]

Oh yeah, it would be perfect for that but I don't know what that model can do in terms of correction.  I think it has a 5 band parametric EQ or something, and you generate the curve using a test mic and Room EQ Wizard and then load it onto the MiniDSP.

 

I thought you were lookign to offload the room EQ for your monitors onto hardware.

Room EQ Wizard is great once you figure out the interface (which can take a while, it's not exactly a masterpiece of UI design).  I forget, but it MAY be possible to convert a standard IR to an EQ curve for the MiniDSP.

 

[TubularCorporation]

As far as using Room EQ Wizard for making standard IRs, which is what I've used it for, the video I always point people to is this one:

 

The basic idea is that instead of measuring from a theoretically perfect listening position, you're measuring continuously while you move the microphone around inside the listening sweet spot.  Room EQ builds a response curve cumulatively as you measure, so you just watch its display and go until the curve stops changing, repeat for the other speaker, and then use it to make a stereo IR from that. You can load it up into any convolution plugin or hardware you want.  You can also do the measurement as many times as you want and then average them all and get the IR from THAT.

I did it the more conventional way first, then I saw this and tried it, and after 5 or 6 times through the full procedure for each speaker, the final impulses I got were almost identical so I'm confident that it works well enough, and waving the microphone around in your listening area is a whole lot easier than measuring and getting it set up exactly right.

 

Took me about 2 hours but I'd say at least 90 minutes of that was figuring out Room EQ Wizard's interface and then convincing myself that the measurements were actually OK.

Edited April 7, 2021 by TubularCorporation

[thawkins]

Yeah I think this is a nice procedure for setting things up, and I hope they do not require a specific kind of measurement microphone so I could use the MEMS mic I got with the IK ARC kit.

Honestly I would consider setting up a dedicated hardware DSP box for this on my main studio monitors, but I am worried about having more latency in my system. If I get one box for the separate stereo system then I can test if there is a big latency gain or not.

All this is armchair posturing as of now, because I don't even have a separate living room setup, just sketching out how to best build one if needed.