ASR Acourate users

[Dathzo]

I will split your question into two answers - recording and playback.

1. For recording sweeps: your Denon X3800 MUST have an ASIO driver and microphone input capable of supporting a calibrated 48V Phantom Power microphone. You could use whatever mic came with your Denon, but it is unlikely to be calibrated, and not of high quality. I checked Denon's website, and it has no ASIO driver. So I don't think you could use your Denon to record sweeps. In this case, you will need a multichannel interface with as many DAC outputs as you need, e.g. a Motu Ultralite Mk.5 or RME Fireface UCX.

Other software (e.g. Audiolense) allows you to use a USB microphone, but there are many reports of inconsistent timing measured with USB microphones. This is because the microphone ADC is not clock synchronized to the DAC. I am starting to take the view that USB mics are good for frequency response sweeps only, and they are too inconsistent for time alignment. This is not based on first-hand experience, I formed this view from reading about complaints of USB mics on other forums.

2. For playback: in this scenario, you measure the sweeps with Acourate and playback via the Denon. In this case, you will need either:

(2a) the Denon is capable of convolution. I checked Denon's website and it does not appear to support this. The online manual suggests that the Denon only supports Audyssey. I don't know how you would even get filters generated by third party software (like Acourate) onto your device. In any case, Acourate only outputs .WAV files, and if you need it in another format, you will need more software to convert it into another format. Your Denon is likely to be incapable of 65536 tap FIR filters given that very low powered DSP chips are usually installed in AVR's. Alternatively:

(2b) you use your Denon as a multichannel DAC, and use a convolver hosted on your PC (like Acourate Convolver, Hang Loose Convolver, JRiver, Roon, etc) to do the processing and send processed signals to the DAC. In this case, your Denon needs to be recognized by Windows as an audio device, preferably via ASIO or WASAPI. Sadly for you, it does not appear to be the case for either. There is a downside to doing this however - all FIR filters introduce latency, which will cause lip sync problems if you use your AVR for video as well. You will need some way to adjust lip sync with your AVR.

Your Denon already has Audyssey built in. You could use it. The advantage of doing this is that your hardware already supports it, and it is likely to produce a "good enough" result. It won't produce ultimate quality - if that is what you want, you will need a major reconfiguration of your system and be prepared to climb the DSP learning curve. You will also lose convenience and may even lose functionality - one example is getting any convolver (not just Acourate) to process HDMI audio. For this you need two additional pieces of equipment - a HDMI splitter, and an interface card capable of internal routing. Audio goes into the HDMI splitter, which extracts the audio and sends it to the interface card. The card then routes audio from the digital input into the convolver input, and the convolver outputs via the DAC on the interface card.

Helìlo Keith,

Many thanks for the detailed answer!
The Denon has indeed Audyssey. It also has DLBC, which does a better job than Audyssey. It sounds really good to be fair, but since a while, I have been curious if Audiolense/Acourate can produce a better sound than Dirac… I guess I would need to take a different route than the AVC as you clearly explained.

All the best.-

 

[Keith_W]

Helìlo Keith,

Many thanks for the detailed answer!
The Denon has indeed Audyssey. It also has DLBC, which does a better job than Audyssey. It sounds really good to be fair, but since a while, I have been curious if Audiolense/Acourate can produce a better sound than Dirac… I guess I would need to take a different route than the AVC as you clearly explained.

All the best.-

Audiolense and Dirac are what I call "black box" DSP's. Measurement goes in, result comes out. You don't know what the black box has done to your signal. I was recently made aware that Audiolense does not let you manipulate curves, e.g. if you wanted to change the volume of a crossover, you can't do it. With all DSP software, if you feed it garbage measurements, you will get garbage results. With Acourate, everything is pretty manual and YOU have to make all the decisions ("black box" DSP makes the decisions for you). The advantage of "black box" DSP software is ease of use and speed. Most of the time the results are excellent. But if you are a tinkerer, and you think you can do a better job than automation, then it's Acourate, or REW+RePhase.

The other difference is that Dirac and the DSP built into your AVR is pretty low res. It typically has 1024 taps per channel, which limits the amount of correction you can do.

So yes, Acourate and Audiolense will potentially give you better sound than Dirac which is built in to your AVR. How much of an improvement depends on how much correction needs to be done, and the skill of the user.

 

[Carousel]

From the 4th Jan 2024 post by @Keith_W ... thank you very much for that!... I now understand the ASIO reasons why one cannot use a different DAC separate from the microphone interface when creating the active crossover filters.

For a Roon user, could it still make sense to use a separate DAC for playback? One can avoid Roon volume control… which is not always reliable and many folks question its quality. And one can avoid manual knob-twisting on the interface to change the volume.

The thinking is to get a multichannel DAC like an exaSound, an Okto or a Topping DM7. After the filters are made with an RME Fireface 802 cabled to the speaker amps, the cables are switched over to the DAC, the filters are loaded into Roon and the DAC is used for playback. Testing this with an exaSound S88 works. The exaSound s88 is nice as one can choose whether to send from Roon over USB or over the network. If one uses the network, then the volume can be changed from a tablet. Though, yes, the exaSound is much more expensive than the Okto and Topping combined.

Any feedback from folks using a separate DAC for playback? Any advice for why else this might be a good idea, or why not, or what else to try? Thanks much!

 

[Keith_W]

I use a separate DAC for playback. Measurements are taken with an RME Fireface UC, and playback is via a Merging NADAC MC-8 (the 8 channel version). The reason I did this was because back then, I believed that DAC's made an audible difference, and I specifically chose Merging because I believed in DSD. I don't believe this any more.

There are issues with using separate DAC's for measurement and playback. I have a multichannel active system, with each driver having its own amplifier. All amps are XLR except for the tweeters which are RCA. The Merging outputs a lower voltage through RCA than it does through XLR, whereas the RME outputs the same voltage in all channels. This means that a measurement which is flat when taken through the RME will show a drop in tweeter volume when played back through the Merging. I had to adjust the tweeter gain on the RME by taking a dozen sweeps.

I hate having to swap 8 cables over when I want to change from measurement to listening. It is a pain in the arse and having two sets of 8 cables creates a snake pit that you wouldn't believe. On top of that there is the potential for plugging cables into the wrong channel and blowing up your tweeters.

Since I already own all the equipment, I persist with it. But if I were to start over, I would get a Merging Hapi. It can have up to 32 DAC outputs and 16 mic inputs on a single device (via option cards), and is infinitely expandable via Ravenna. I could buy 10 Hapi's and have 320 DAC outputs if I wanted to. Right now, the easiest way for me to get more DAC channels is to sell the RME and buy a Merging Anubis. It has 6 DAC channels and 2 mic inputs, and can be added to the NADAC via Ravenna. But to me, having too many boxes sitting around is rather inelegant so I might just spring for the Hapi instead. Decisions, decisions.

And BTW, I don't use Roon. I use JRiver. Its volume control is absolutely reliable.

So: regardless of your reason or motivation for using a separate DAC for measurement and playback, of course it is possible. I just don't think it's a good idea any more.

 

[Carousel]

That is very helpful information. Thank you. Was just playing with the equipment I have handy today… though I have started considering some Merging equipment… and have a question.

Has anyone successfully used Acourate Convolver to send to an exaSound DAC via the exaSound ASIO?

When I use Acourate Convolver to send 24-88 to an RME Fireface 802, the RME dials the buffer size up to 2048 samples. But with the exaSound ASIO, Acourate Convolver reports only 512 samples and the sound is all chopped up.

 

[Keith_W]

That's a question you should ask Dr. Uli in the Acourate support forum.

 

[Carousel]

He told me to ask exaSound whether there is a way around that it seems their ASIO does not have a setting available to up the buffer. I have asked and am waiting for a reply. Was wondering if anyone else had figured it out already here. Or maybe I have to use a different convolver etc. Do not want to post the question across forums while already waiting on a reply from an OEM. But figured it would not be too rude to ask the question in one logical place. I only have so many days Easter holiday to play and they are going... PS: He uses a Merging Hapi himself.

 

[Keith_W]

You can try a different convolver. JRiver and Roon have built-in convolvers. I believe Foobar2000, which is free, has a convolver plugin. There's also CamillaDSP, which is free, but with the caveat that it is rather difficult to install and get working. And you can download and try some other convolvers for free, e.g. Hang Loose Convolver.

 

[Paul W]

Any thoughts or experience with driver linearization for an MTM style speaker?

 

[Keith_W]

Any thoughts or experience with driver linearization for an MTM style speaker?

I haven't tried it, so I don't have direct experience. But I can't imagine that it would be a problem. Do you have a more specific question?

 

[Paul W]

Yes. Wondering about best distance from the mic and the thought process.

(BTW, we had a wonderful time in your country about 2 months ago.)

 

[Keith_W]

When in doubt, measure at various distances and repeat. I measured some speakers which I have never seen before today. They had open a pair of baffle woofers per side, and I have never measured open baffle woofers, or a pair of woofers for that matter. So:

I had all kinds of doubts, e.g. would the pair of woofers cause driver lobing, what about the out-of-phase back wave (I had no choice but to do an in-room measurement), and so on. So I took measurements 20cm, 50cm, and 100cm from the driver plane on axis (red, green, brown respectively). I elected to use the green measurement for correction, but in hindsight I should have not done a nearfield correction of the woofers at all. When in doubt, leave them alone. Not to say it's not worth experimenting, if you don't mind spending the time.

 

[3ll3d00d]

Any thoughts or experience with driver linearization for an MTM style speaker?

I would measure with both drivers active quasi anechoically and mic facing the tweeter to start with