ASR Acourate users

[Roland68]

So bass management is the function category for all that.

With "phase / delay issues" mostly a part of that

as opposed to "Custom designing speakers from separately driven drivers" ? aka "bi- or tri-amping" which I expressed not a current goal for me.

So, add speaker / room compensation EQ.

Pretty much the list then, excluding convolver-only issues?

...

>> That assumes only one source is available per input?

> I'm not sure what you are getting at

Setting aside the DSP implementation choices for now. Say the Wiim Ultra is my preamp. If my turntable is not suitable to plug in directly, it needs to go through an external box via the regular analog input. Given other analog sources, a passive switcher goes in between.

Likewise if I have multiple TOSlink devices, I may put an auto-switcher box in front of that input, and I may want a different "input compensation EQ profile" for each.

Which I may want even if I am using only one source device per input on the Wiim.

Of course since Wiim offers that, maybe just use it for that function, and the "ad-hoc PEQ" and loudness compensation if desired

And use the "real DSP" to handle the bass management + room compensation, "DSP'ing the DSP".

At least it's all digital transport from the Wiim right up to the output interface DACs.

...

I am not at all averse to DIYing audio PCs or even rPi boards, I've spent thousands of hours on Linux sysadmin stuff over the years, BSD before Linus was out of short pants.

No way I can afford those hardware DSP units, unless I skimp on groceries for years - even the 4-port miniDSP is pricey in my world.

Then the project Introducing DSPi | A powerful, user friendly and open source DSP for less than a cup of coffee might be of interest to you.

 

[Keith_W]

Setting aside the DSP implementation choices for now. Say the Wiim Ultra is my preamp. If my turntable is not suitable to plug in directly, it needs to go through an external box via the regular analog input. Given other analog sources, a passive switcher goes in between.

Likewise if I have multiple TOSlink devices, I may put an auto-switcher box in front of that input, and I may want a different "input compensation EQ profile" for each.

I don't know how you would automatically switch EQ profiles for each TOSLink source, nor why you would want to. A DVD player, or a DAP, or a TV, or any digital source you care to name should have perfectly linear output, unless it has some kind of built-in DSP that you can't defeat. And if you by some chance own a device that modifies sound that you can not bypass/defeat, you should throw that device in the garbage bin. Digital devices DO NOT NEED individual EQ. Only nonlinear analog devices like vinyl and tape require EQ. Why anybody would want to use these devices in a high fidelity system in 2026 is a separate discussion.

If you want to continue using your Wiim, you could use your Wiim as a source selector and pass the output of your Wiim through a PC for linear-phase DSP. You will need to figure out a way to do that, the easiest way I can think of is to use your Wiim's optical output and buy an interface with an optical input.

As for that DSPi device mentioned in the above post, that is a minphase IIR DSP device like the Wiim. It is not linphase FIR. The advantage of using that device over your Wiim is that it has more subwoofer outputs. The Wiim ultra only supports a maximum of a 2.1 configuration.

If you want to design filters for minphase IIR, I would use REW to do that. Bear in mind that biquads have to be generated specifically for each hardware device and sample rate, you can not take MiniDSP biquads and use them in your Wiim. However, the PEQ's are the same. If your device is not supported (my up-to-date REW Beta doesn't support Wiim), you will have to generate the PEQ's and enter them manually.

 

[john61ct]

Very much following. But somewhat allergic to projects under THAT rapid a development cycle.

Plus starting off with a powerful PC for learning / testing, worrying about #taps stuff, optimizing later on makes sense.

Same as consolidating storage NAS, LMS server, convolver and (maybe) a PC-based player...

 

[john61ct]

As I said a Wiim Ultra would be a pre-amp only, not doing bass management unless for just part of the overall system. Maybe good enough for room compensation EQ? i dunno yet.

My main pair needs midbass coupler reinforcement in stereo, way higher frequency range than where mono-trueSub becomes useful.

Same if I used Wiim AMP Ultra as a not-so dumb power amp on the output side.

Maybe for Center front + a "main mono" trueSub if that turned out to be a good placement, supplementary mono subs not digging so deep getting freely placed for the room modes.

minphase IIR
linphase FIR
biquads vs MiniDSP biquads

These terms do not yet have meanings in my brain...

I do not currently plan to base my DSPing on miniDSP given the cost of scaling up to a dozen channels, plus their "locked in" proprietary nature.

If the 2/4 HD is useful for testing or one specific function in one point of the system then sure, at least that's affordable

 

[3ll3d00d]

So bass management is the function category for all that.

No, bass management refers to the use of a (typically mono) subwoofer channel to handle all content below a given frequency. It doesn't require multiple subs.

 

[Keith_W]

Maybe for Center front + a "main mono" trueSub if that turned out to be a good placement, supplementary mono subs not digging so deep getting freely placed for the room modes.

As mentioned, there are a lot of ways to skin a cat. You can have no subs, or 1, 2, 3, or more. Maybe a dozen if you are crazy / rich enough. The more subs you have, up to a point of diminishing returns, the more even the in-room bass can be, and the more options you have for setting it up. If it's for a HT system (as you seem to be hinting at) I would suggest you don't use Acourate because of linear-phase FIR latency and because you can't load Acourate's filters into your AVR. Just get an AVR, use whatever room correction software they recommend, and be done with it. DSP software that comes with AVR's are typically not as flexible and tend to lock you into their way of doing things, but it's good enough for general purpose use.

These terms do not yet have meanings in my brain...

See this thread for an explanation.

I do not currently plan to base my DSPing on miniDSP given the cost of scaling up to a dozen channels, plus their "locked in" proprietary nature.

They are the industry standard. Yes, it's proprietary. But it's also widely supported. I would not worry about MiniDSP being proprietary, unless it's some kind of philosophical objection.

 

[3ll3d00d]

There's nothing proprietary about a minidsp biquads, the main difference is that one of the coefficients has the sign reversed and that's about it.

 

[john61ct]

No, bass management refers to the use of a (typically mono) subwoofer channel to handle all content below a given frequency. It doesn't require multiple subs.

I need to use bass management for multiple subs,

both trueSub (deep extension as possible, at least one)

midbass couplers up to maybe 250Hz or higher where stereo is still required.

There may or may not be more mono subs with frequency ranges more in the middle.

There is of course some overlapping requirements between "bass management" and "room compensation EQ"

but I don't see a third topic domain in there, assuming anechoic speaker compensation EQ came before.

This being a mobile use case, there are several different "rooms" involved, plus outdoors so I need reasonable efficient swapping of RC profiles in & out leaving the other domains in place

 

[john61ct]

> See this thread for an explanation.

Thanks much!

The more subs you have, up to a point of diminishing returns, the more even the in-room bass can be, and the more options you have for setting it up

It seems that optimizing for a single LP gives "tighter" results at LFs?

So if profiles can be swapped in & out, best to only use those designed for group listening or "best compromise" for say a large dance floor, are best swapped in for when it's not just me?

...

> If it's for a HT system (as you seem to be hinting at)

Music 80%, likely even for film no proprietary codecs involved, PCM input only, and certainly no AVP/AVRs

I also don't do video streaming from the internet, most use cases there is no internet at all available.

So this:

> I would suggest you don't use Acourate

does not apply?

...

> I would not worry about MiniDSP being proprietary, unless it's some kind of philosophical objection.

Yes I do feel strongly about that topic, but don't let it get in the way of practicality. The cost issue for lots of channels remains...

 

[john61ct]

This is my starting point for a "factors matrix" for classifying different DSP tools.

HQPlayer + Computer vs DSP architecture

I’ve only recently become aware of a very different digital audio system design approach, and I’m trying to sharpen my understanding of how it compares with a more conventional appliances+loudspeaker-centric architecture. For context, I’m coming at this primarily from a loudspeaker and room...

www.audiosciencereview.com

Any chance those in this thread could help answer the implied questions there about Acourate, from the audience-POV of a budget-limited noob?

I realise much may be opinion, so be it.

I think so long as the centerpoint is Acourate, post responses here?

I'll start

...

DSP PURPOSE

mostly filter design, but some overlap with REW for measurements.

convolver is separate

...

PLATFORM

Windows PC, unknown how modern / powerful required

for measurement, high-quality measurement microphone, well calibrated, ideally working through an interface with ASIO drivers. Other mic's like USB requires workarounds.

...

DSP FEATURES & FUNCTIONS

speaker compensation EQ

bass management and crossovers in general, both frequency and phase / time domain

room compensation EQ

all three best of breed?

...

SKILLZ

professional-grade tool that offers extreme flexibility

steep learning curve, especially for DSP noobs

not much automation, requires manual work

...

COST

aside from hardware ~$300 license, one-time purchase. support via forums, sometimes directly from (Sehr geehrter Herr Prof. Dr.) Ulrich Brüggemann aka @UliBru

...

FILTER TYPES

FIR (Finite Impulse Response) - latency issues, lip-sync issues with video - depend on number of taps and sample rate

minimum phase versions of filters can reduce latency, trades off time-alignment precision

 

[3ll3d00d]

I'd say this is a not uncommon situation, i.e. posting on a forum analysing what you can do to the nth degree, and it is rarely a good idea vs learning by doing. Just get any random cheap as possible solution (free software, any random cheap DSP board) and experiment with it, learn how to take measurements (without that, the whole thing is junk) and see what you really need. After some iteration it will surely settle down and solidify into a working system that changes more slowly.

As to this thread, don't think I'd ever recommend acourate for someone who has zero experience in this field. Nothing is impossible but the learning curve is brutal compared to pretty much any other software out there (no doubt something out there is harder but it's still a steep cliff to clamber up)

 

[john61ct]

Thanks for that advice. So I infer that you think the above is pretty accurate (if basic) info wrt Acourate?

 

[Keith_W]

It seems that optimizing for a single LP gives "tighter" results at LFs?

The "tightest" result will always be over a single seating position. This is especially so if you only have one subwoofer. The more subwoofers you have, the less the seat-to-seat variation. You can choose to make the result "tighter" still at one seat, or spread the benefits over several seats.

So if profiles can be swapped in & out, best to only use those designed for group listening or "best compromise" for say a large dance floor, are best swapped in for when it's not just me?

Yes, you can do that.

So this:

> I would suggest you don't use Acourate

does not apply?

As I mentioned, Acourate is specialized FIR filter design software. You can make minimum-phase FIR (less latency) or linear-phase FIR (more latency, but also more flexibility). If video lip sync is a concern, I would make a set of minimum-phase filters just for video, and a set of linear-phase filters for music.

The difficulty is how you are going to configure your system for video and for music. For video, a very specific architecture is needed. Even with minphase FIR, there is still some latency from buffers, and you will need lip sync. Therefore the video MUST pass through software which can delay video for lip sync - e.g. JRiver, Kodi, and some others. If you want to watch a DVD, your PC will need a DVD drive. If you want to watch TV, you will need a TV receiver card. And your software needs to support accepting input from the DVD drive and TV receiver. There is no software that I know of that can handle video streaming (e.g. Netflix) and provide lip sync on Windows. Maybe if you watch Netflix in a browser window in JRiver it might work, but I don't know the answer. @3ll3d00d might know, he is our local JRiver expert. The only device that can do that are AVR's.

For music, the solution is much simpler. Player and convolver on the same machine, output to DAC with enough channels, and you are done.

In my signature, there is a link to my REW eBook. In the same Google Drive folder, there are also two free Acourate eBooks. I didn't really write them for beginners, its more a collection of Acourate recipes, and the writing style is rather terse and information-dense. But it gives you an idea what Acourate can do.

 

[Keith_W]

DSP PURPOSE
mostly filter design, but some overlap with REW for measurements.
convolver is separate

Correct.

PLATFORM
Windows PC, unknown how modern / powerful required
for measurement, high-quality measurement microphone, well calibrated, ideally working through an interface with ASIO drivers. Other mic's like USB requires workarounds.

Windows PC: Acourate will run on pretty humble hardware, and the filters can run on something as humble as a Raspberry Pi.

For measurement: I wouldn't even try to use Acourate with a USB mic. The inconvenience and complexity is not something you want to deal with as a beginner. XLR mic and interface is by far the best option.

DSP FEATURES & FUNCTIONS
speaker compensation EQ
bass management and crossovers in general, both frequency and phase / time domain
room compensation EQ
all three best of breed?

"Best of breed": I wouldn't say so. If you only use Acourate's room macros, it's just inversion over a target curve. ALL filter design software does that. Its real strength is its flexibility. You can make it do whatever you want. If you are a beginner, your result will be the same as any other filter design software on the market. But if you are keen to delve into what is possible with DSP, IMO Acourate is the best tool.

Filter design software exists on a spectrum. On one end of the spectrum, you get automated software that tries to hide the arcane nature of DSP from the user. On the other end of the spectrum, you type the equations in yourself (e.g. Matlab or Octave). Acourate is in-between. There is some automation in Acourate, but those things are really macros - they execute Acourate's modules in a specific order and with a friendlier interface. If you wanted to, you could skip all those macros and execute each module yourself.

SKILLZ
professional-grade tool that offers extreme flexibility
steep learning curve, especially for DSP noobs
not much automation, requires manual work

Re: the learning curve. I bought my Acourate license more than 10 years ago. I was a complete beginner. I used Mitch Barnett's book and followed every step religiously. It is possible for you to follow a recipe and obtain a very good result. At the time, I found some of Acourate's design decisions to be absolutely baffling. It forces you to make your own test signal and read your own impulse response for timing measurements - other software gives you timing information instantly. Why do this? Because it forces you to think about timing, and what you are really measuring. Nowadays, when I see an automated timing result, I question what was measured.

FILTER TYPES
FIR (Finite Impulse Response) - latency issues, lip-sync issues with video - depend on number of taps and sample rate
minimum phase versions of filters can reduce latency, trades off time-alignment precision

For minimum-phase, the trade off is not time alignment precision. It is all the issues that accompany minimum-phase DSP. With minphase, the phase is locked to the amplitude. If you change the amplitude response, you change the phase response. For a HPF/LPF, each filter order introduces 90deg of phase rotation. This is why some crossover types don't sum properly. You don't get such problems with a linphase filter - filter orders do not introduce any phase rotation at all, so all symmetrical linphase HPF/LPF sum properly, regardless of filter Q or order. Minphase filter design is slightly more complex because of this.

Linear-phase lets you manipulate both independently. It does that by delaying the impulse (so that corrections can happen BEFORE the impulse), which is what causes the latency. It can do things unheard of in nature, like linearise the phase of a driver, even across the roll-off regions.

 

[john61ct]

Thanks much for all that

For video, a very specific architecture is needed.

We've had no "TV" since the 80's, whatever that might be nowadays, nothing via antenna or "cable", no discs

We do not stream video, only download files then play them. Whatever software does the job is fine.

Music SQ really is the priority.

Do not assume there is "a player" running on any PC.

Whatever convolver, will work with input from any sources including analog. Let's just say for illustration, everything goes through the same pre amp stages, then the convolver "outboard processor" just before feeding the power amps.

That convolver running on a box separate from NAS / music server / control points / renderer (if any)

 

[Keith_W]

We've had no "TV" since the 80's, whatever that might be nowadays, nothing via antenna or "cable", no discs

Great! In that case, a PC + interface is all that you need. Your choice of Roon / JRiver / Foobar2000 / Kodi / etc. + a convolver of your choice, all running on the same PC. Note that JRiver and Roon have built-in convolvers, but those are a bit basic. As for the interface, choose one from a reputable company (I recommend Motu or RME) that has as many output channels as you need.

The only question is whether Acourate is suitable for you. There are many options, Acourate is only one of them.

 

[3ll3d00d]

The difficulty is how you are going to configure your system for video and for music. For video, a very specific architecture is needed. Even with minphase FIR, there is still some latency from buffers, and you will need lip sync. Therefore the video MUST pass through software which can delay video for lip sync - e.g. JRiver, Kodi, and some others. If you want to watch a DVD, your PC will need a DVD drive. If you want to watch TV, you will need a TV receiver card. And your software needs to support accepting input from the DVD drive and TV receiver. There is no software that I know of that can handle video streaming (e.g. Netflix) and provide lip sync on Windows. Maybe if you watch Netflix in a browser window in JRiver it might work, but I don't know the answer. @3ll3d00d might know, he is our local JRiver expert. The only device that can do that are AVR's.

Whatever convolver, will work with input from any sources including analog. Let's just say for illustration, everything goes through the same pre amp stages, then the convolver "outboard processor" just before feeding the power amps.

tl;dr, if you want high latency audio filters for video playback, plan on using jriver as it's the only software i know of that has the required functionality (and only for file playback not streaming).

longer version.... modularity provides flexibility but it invariably incurs a cost, in this context the costs are typically increased complexity (more things to control/configure/go wrong) and reduced efficiency and/or optimisation opportunities. For video, you need to maintain sync between audio and video which means you need predictable latency that fits within the video pipeline latency budget (projectors can be relatively high latency) or you need a player that is aware of the time spent in each pipeline and delays one or the other accordingly. The latter means the player either is or hosts the convolver & is aware of the filter latency (aka using the jriver convolver or a convolver vst that reports its latency via the vst api e.g. hang loose).

for live playback, you can definitely do this via a computer with appropriate hardware and stream 4k + hdr through it (albeit I had to write the software to do that as there was nothing else out there that could handle both audio and video) but there's currently no software (that I know of) that will delay the video to allow arbitrary length filters to work. I don't think it's that hard to implement but I don't need it (I use mds based crossovers with IIR on top for live playback) so never bothered to implement it.

 

[john61ct]

Video is such a low priority, cross that bridge later, after music is solved...

Meantime, put up with lips out of synch NP, or

switch over to no-dsp paths, whatever

 

[AudioJester]

Looks like I need a new calibrated microphone. Any recommendations?

 

[Keith_W]

Looks like I need a new calibrated microphone. Any recommendations?

Cheapest option: Behringer ECM8000 calibrated by a lab, e.g. Cross Spectrum Labs. Google "microphone calibration" to find services in your area. Call / email them and ask if they can calibrate an omnidirectional condenser microphone for the purpose of loudspeaker/acoustic measurements with a 0deg and 90deg incidence. Make sure you ask if the calibration file is supplied in inverted or non-inverted format. It's usually non-inverted, meaning that you have to invert it yourself (Acourate) or the software will invert it for you (REW). But if it's already inverted, you will have to invert your cal file before you use it in REW! Disadvantage of a Behringer ECM8000: poor microphone sensitivity. It's about 10-15dB less sensitive than my Earthworks M30. Alternative to ECM8000: Dayton EMM-6.

A note about cheap "calibrated" microphones: they may not be individually calibrated. Usually a few mics are selected from the batch and tested, then the calibration is applied to the entire batch. Depending on manufacturing tolerances, your mic may deviate from the calibration. How to test: enter a different serial number which is likely from the same batch (e.g. 1234567 and 1234566) and see whether the cal file is the same.

Cheapest "decent" option: iSemCon EMX-7150 or Beyerdynamic MM-1. I have no experience with either of these mics.

Luxury option: Earthworks M23 or Earthworks M30. The difference is that the M23 measures up to 23kHz, and the M30 up to 30kHz. I used to think I wasted my money on an M30 when I could have bought an M23 instead, but no - apparently the minimum-phase nature of microphones means that they may not be phase accurate at high frequencies (applies to all mics not just Earthworks). If the HF limit is higher, the phase distortion is pushed higher in the freq range. JJ made this point somewhere on ASR (sorry, I can't find it).

Ultra-luxury option: Bruel & Kjaer 1/4" free-field or diffuse-field microphone.