Comparison of DRCs: Dirac Live for Studio, IK Multimedia ARC System 3 and Sonarworks Reference 4 Studio edition

[czt]

Hi, I'm having a bit of trouble with this With a video explaining the whole process it was much easier. I tried to search in REW where they export to WAV with timing Where is it in the REW software?

View attachment 277293 View attachment 277294

"Generator" toolbar button. (fig. 1)
The .wav loaded in "Measure", and after "Start" waiting for playback. (fig. 2)

 

[ai studio]

hi thanks
I'm having trouble with this.
I did not make it
This is my first time doing something like this


Maybe it's just possible to take a screenshot in OBS
the whole all the process and upload here or to YouTube
It was much, much easier for me
And I believe for other users too

Open Broadcaster Software | OBS

OBS (Open Broadcaster Software) is free and open source software for video recording and live streaming. Stream to Twitch, YouTube and many other providers or record your own videos with high quality H264 / AAC encoding.

obsproject.com

 

[czt]

Disregard when he sets the same values under Measure (for his dummy output), just load the saved .wav file(s) with the "From file" button before "Start". (And please do search.)

 

[landco]

You can of course be aware of EQ that's not natural (natural must mean flat) because EQ is up to taste but acknowledge that it isn't flat but tuned to taste.

Flat frequency response is not natural in a typical living space. Here it is better to use the "Harman Curve" and its varieties. I'm using a B&K 1974 Speaker target, this curve is ideal for a home environment.

 

[IamJF]

Late to the party ... but want to add my experience to this topic. @dominikz THANKS for this great measurements and documentation!

I worked with Dirac, Sonarworks and Neumann MA-1 and my conclusion is similar. With Sonarworks it always sounded different and not bad - but also not really better as without. (Good speaker in good room, probably different with a bad speaker cause even rough linearisation can be a benefit). Seeing now your test data - I know why.

Dirac is a bit of a hassle to set up (Lexicon MC-10) but works well during the process. You can also work with big Atmos setups which is great. Result is for sure a benefit and a step in a more "realistic/neutral" directioin.
(e.g. my older way cheaper Yamaha amp also delivered a "different" sound with some benefits but no comparison to Dirac which is clearly a step forward)

And Neumann MA-1. Downside - only works with Neumann DSP speakers. But these are pretty good anyways ;-)
The software is still under development, maximum of 2.2 is suported. But the measurement itself is easy to do and stable. It's a little clunky to manage presets - but that's with Dirac too. The result ... is great! A clear step in the direction of "confidence". I used this with monitors for mixing, KH750 + KH80 for a friend in a relatively bad room, KH120ii in my strongly dampened room. And in both cases these setups reminded me to my main speakers - there was some "truth" in what you hear, you could reliably mix on these setups.
As far as I searched for comments it seems to bring better results as Genelecs system - but I never tested or compared it, so this is only silly hearsay, don't take it to serious .

 

[Flak]

heads-up

Black Friday - Dirac

Don't miss out on Black Friday 2023 at Dirac on November 24. Treat yourself to better audio with Dirac Live during Black Week.

www.dirac.com

 

[bennybbbx]

here is no test of the new arc 4. test it there is a 9 day free demo and i think it is the best of all speaker corrections when set correction type to broad(is less correction) and natural. it is possible to set it to linear but in both modes it have only 0.1 ms latency. so guess it work then in mixed mode. in sharp(much correction) mode linear sound better as in natural but all and all the natural and broad mode give the best results. to hear if a speaker correction is good, play a mono song and when it sound smallest then this is best correction between left and right speakers. I can with arc 4 also use diffrent correction for left and right speakers and sound better as combined.

when hear much bass i get no feeling that the reverb of bass come more from left side . I think it is a huge step forward in speaker correction that it use broad mode. before arc 4 i never get good souning automatic speakercorrection, but now it is good for me and asymetric room. see screenshot what it correct in broad mode. the bass also low bass is a little more in compare to headphone but sound good. this screenshot show only R channel before and after to see better that it not correct so much. maybe it can enhance even more when have more steps between sharp and broad

broad correction type

sharp correction type

 

[DWPress]

Room symmetry is skewing things in favor of your right channel for room gain but things match up after 1.5kHz. Try limiting your correction to just 1.5kHz and things might sound better than what you're happy with now.

 

[czt]

And try to decrease the low range until the before and after is in overlap (for natural speaker cut off), Ctrl + Left click on trim, and you can test the combined l/r mode for asymmetry.

 

[bennybbbx]

Room symmetry is skewing things in favor of your right channel for room gain but things match up after 1.5kHz. Try limiting your correction to just 1.5kHz and things might sound better than what you're happy with now.

the screenshots i do show only right channel before and after correction, i write in text before that only R. my left speaker is louder in bass and the right speaker is more near wall. strange. but happen with all speakers can see in top of screesshot. grey L R text mean it is disable to show. i do this to show that it is better not correct too much. When show all 4 curves it can not see anything. here i show the before L and R curve. limit the correction range give no better results. it is not easy to hear best setting. it can happen that it sound great with lots of clarity(not thur boost high) but after longer hear(more than 5 minutes), ears get tired and strange feeling. this happen when correct full also with phase linear.

I use a desktop and the correction seem made for desktop. I choose this

here is before coreection both channels

after correction

this room config i use. i have keyboard on desktop, mixing i can do in DAW. I have also treatet my room with basstraps basotect and 5 cm plates . i do it with 7 measures . strange is in my room i always get a huge gap at 500 hz which sound not so nice. but boost sound also not good. i only can simulate such a gap on headphone which sound not so good. i thry with plates, but nothing help, also plate on desktop

EDIT: Another thing I noticed is that the latency display in studio one doesn't change when you switch from natural to linear. You have to turn the plugin off then on again then you can see it properly. In linear mode the arc 4 then has a latency of 42.7 ms. So full linear in the bass too. I still think it doesn't sound that good in lineaer mode. But it could also be because the level changes at 63 Hz and the valley that I have on the left becomes even larger. But it's amazing that the Arc 4 is the first automatic speaker calibration thing that sounds better than my hand-tuned one in the mini DSP.

 

[bennybbbx]

I think speaker correction compare need also do with same microphone.
Because i see never in my life a test of measure microphone i thought a 50$ microphone is good enough. I want know it and i buy a beyerdynamic MM1 which cost ~230 $ and use it in arc 4.
in arc 4 is suggest to use measure level you hear normally. I like hear 72-75 db and neighbours can not hear. with the arc microphone and sonarworks microphone(which cost ~80$) i get not good results at this level. i measure with 85 db or more. now with the beyer i can measure with 75 db and the strange bass that on left side sound as more reverb is less, also clarity and location is increase . brain try always correct that it sound good, but in compare what find before good sound now not so good when longer hear. In the FR the beyer record ~1 db more level from 150-900 hz. so the correction reduce this level in compare to other microphones i have.

I have do screenshot when run measure with speakers off so it measure FR of ambient noise. the arc mems microphone can see as a reference because such mems microphones have very few production deviations. because there is no FR correction for the mems microphone i do a correction that it is simular to the sonarworks long time ago. sonarworks have much more noise. but mems create noise at 50 hz. for rew it is best do calibration and add always the sensivity parameter, because every microphone have diffrent gain. beyer give 4 db more gain and reduced noise . that in bass is noise more is because my PC fans rotate with 700 rpm do some noise and the street too do noise in bass

and distortion of speaker get also lower with the beyer.

btw i buy now a esi unik05+ it have ribbon tweeter, have strong driver and a bass plug is attached to close the port, and distortion is not increase. it is smaller as the HS5 and have this valley in bass the HS5 have not. I can with the mm1 microphone strong correct and get good sound with much low bass. speakers are place same as hs5 on foam pads that are tilt. measure is with open bass port . my conclusion is . when use speaker correction a good measure microphone is very important . if the measure microphone for 300 or 800 $ get even more better sound, i did not verify. but i guess the mm1 is not much topable.

atteched is the calibrartion file i use for the arc mems . I get this together with the arc 2 and arc 4 i buy upgrade for 100 $ without microphone. if somebody also have a arc microphone he can use my cal file and compare his microphone to this. i have set the gain knob in middle of the scarlet focusrite 8i6o. then it is what in cal file stand

Sensitivity -22.0dbFS

 

[Basilderat]

I approach DSP a little differently. I prefer to listen to DSD recordings, so software DSP won't work for me. For DSD I've measured the room with the MMM method, then use a Harman DBX231 31-band GEQ to flatten the response curve as best I can.
I also listen to FLAC files from time to time, so from there I also used REW to generate room curve filters which I apply in Foobar via the FabFilter Pro Q3 VST plugin.

 

[Snarfie]

Just wanted to share a few of my random thoughts on automatic room correction that I was mulling over the past few days...

IMHO one of the main issues with practically all current automatic room EQ software solutions is their default target curve. As you probably noticed, it is typically flat (Sonarworks, MathAudio, ARC System) or close to it (slight tilt in Dirac Live).

My problem with this is that I feel many users will at least initially just use the default target, and it doesn't seem likely to me this would fit many (perhaps even most) real-world scenarios. Of course, in my tests I did find that the flat target may work OK in nearfield with flat on-axis response loudspeakers, but I'm not sure how widespread that scenario is in reality.
Different loudspeakers will have different on-axis response, and especially if listening in farfield, louspeaker dispersion as well as room characteristics factor-in and one needs to adapt at least the tilt of the curve to get natural sounding results.

So I wondered why none of the DRCs I tested try to automatically generate a target curve shape based on the measured in-room response?
E.g. perhaps they could do a linear regression of the measured in-room response above the transition frequency to build a flat but tilted target better matched to the room and loudspeaker.
Another option would be to simply follow loudspeaker in-room response above the transition frequency, and use e.g. 1dB per octave bass boost below it. I'm sure there are other/better options that one could use.
Sure, some loudspeakers may measure so uneven that it could be very difficult to estimate a reasonable target curve - though probably in that case a flat target also wouldn't perform much better.

Any thoughts on this?

Seems to me like it could be a simple way to provide better out-of-the-box results for many casual users.

Mabey a bit late but for my understanding Wavelet used for head phone corrections has a functionality/silder where you can linear could change the amount of correction regarding the used corrected target curve ( Harman curve) to reach a sort of maximum (personal subjective) transparancy.
Would that be a solution for Farfield listening with speakers. Besides FR correction could also phase/time correction be addressed the same way ? Something that will be a less or no problem with headphone correction what i understand. I could imagen that the found flat target curve contains also phase/ time correction which you probably would always maximize. By changing linear only the FR both could be or will be out of sync.

https://imgur.com/Qa89SX5

 

[DWPress]

Besides FR correction could also phase/time correction be addressed the same way ?

Yes, and with active systems each individual speaker component can be phase adjusted L/R + sub(s). Check out many of the room correction discussions or Amir's video for more information.

 

[dominikz]

Mabey a bit late but for my understanding Wavelet used for head phone corrections has a functionality/silder where you can linear could change the amount of correction regarding the used corrected target curve ( Harman curve) to reach a sort of maximum (personal subjective) transparancy.
Would that be a solution for Farfield listening with speakers.

In my opinion the following is the optimal solution for farfield listening:

1. Optimize loudspeaker/subwoofer placement and crossover parameters to minimize the main SBIR dip (link). Using a properly integrated subwoofer can help a lot with this.
2. Measure each audio channel response using the moving microphone method in REW (link), and apply variable smoothing to the measured responses.
3. Use a downward-sloping line as a target for EQ in REW, roughly matching the slope of the in-room measured response. In my case the slope is approx. -0,6dB/oct, but this depends on both how reflective the room is and how wide or narrow the loudspeaker radiates (directivity).
4. Use REW to calculate correction filters for every channel. My suggestion is to use only negative-gain filters (you can do this by limiting individual and total max boost to zero), and to only calculate the correction below approx. 350Hz (link). The intention here is only to knock down room resonance peaks, while response dips should already be addressed in the first step.
5. In case response correction at higher frequencies is needed, e.g. in case of loudspeakers with poorly tuned stock tonality, use the EQ profile from spinorama.org above 350Hz, and the correction calculated in the previous step below 350Hz.
6. Apply the calculated PEQ parameters in your choice of HW or SW and enjoy!

For me the above process always resulted in what I consider natural sound reproduction, so I personally don't see a need for an additional control to reduce the level of correction.
But other people may of course disagree, so I guess why not have more options.

Besides FR correction could also phase/time correction be addressed the same way ?

Personally I have yet to hear consistent benefits of dedicated phase or impulse correction when doing overall room EQ. I can however see some benefit of phase correction in a couple cases:

• Basic delay/phase alignment of the crossover between mains and subs.
• Precise phase correction in case a summed channel response exhibits a phase cancellation in the bass which doesn't otherwise exist in individual channel responses.

Apart from those two cases, IME even when the difference was audible in a blind test, it was usually worse with phase correction since there can be audible pre-ringing with severe phase correction. In some cases the difference was audible but it was neither better nor worse, and in most cases it was not audible at all. So at least so far I've decided general phase correction is simply not worth the effort for me. Note that I'm not the only one skeptical of loudspeaker phase correction (see this article in AudioXpress).

I could imagen that the found flat target curve contains also phase/ time correction which you probably would always maximize. By changing linear only the FR both could be or will be out of sync.

Note that any minimum-phase EQ will apply both magnitude and phase correction to the input signal.

 

[landco]

Gentlemen, use Cavern EQ and forget about the primitive 30-band Harman equalizers, as well as the long and tedious assembly of filters manually using REW, RePhase, etc.

 

[czt]

I use 2 cycle FDW excess phase inversion with REW (the very long filter don't bothers me) and never want to listen without it.

 

[dominikz]

To expand a bit on the original topic of DRC comparison, recently I've also had a chance to experiment with the WiiM Room Correction (RC) feature.

First I have to say that I'm absolutely impressed that WiiM is continuously developing and enhancing these kind of advanced functions, and even introducing them retroactively to older products!
The SW support side seems quite incredible too - e.g. I found a bug with RC function early in my tests, reported it to WiiM, and they delivered a fix in a couple of weeks. Amazing!

How it works​

Back to topic, here's how WiiM RC works in principle:

• You open the WiiM Home App, select the setting for the WiiM device in question and choose "Room Correction"
• You have the option of editing some basic settings like:
  • target curve,
  • frequency range where correction is calculated and applied,
  • max gain and Q,
  • smoothing level,
  • whether to correct the subwoofer range (i.e. below the sub crossover frequency) or only above it,
  • whether a single or multiple sweeps are used to measure the in-room response,
  • and recently also the option to include a mic calibration file for external microphones (i.e. not available if you use your built-in phone mic).
• You then start the process by playing a test sound through the speakers, after which you have to option to select "stereo" or "individual channel" correction - basically whether the same PEQ filters are calculated for and applied to both channels or each channel gets their own PEQ filters. There is a limitation at the moment where you can only use the "stereo" correction option if WiiM Bass Management is also enabled.
• You then point the phone mic (or external USB mic, if you have one) towards the loudspeakers and record some sweeps.
• WiiM RC then calculated the correction PEQ filters, and lets you save them with a custom name (so you can have multiple ones).
• If needed, you can switch between RC-calculated PEQ presets and even edit the filters manually in the EQ settings view in the app.

More information on the process (with pictures) can be found in WiiM support pages.

The results after WiiM RC can absolutely be good, but as always rely a lot on which settings you choose, and how 'flat' your microphone response is.

The default WiiM RC settings are OK but IMHO not ideal, and some IMHO important options are missing (more on this at the end of the post).
In addition, many phone built-in mics will surely have a terrible frequency response, potentially degrading the response - e.g. my Samsung S23+ is pretty bad.

Testing RC​

Let me give an example of how using a different mic influences the response after WiiM Room Correction:

Red dotted line shows the uncorrected response. Note that this is an ultra-nearfield scenario so the target response should be almost flat.
The orange dashed line represents the response after WiiM RC when the built-in S23+ mic is used to measure. You can see the lows are boosted a bit.
The blue solid line represents the response after WiiM RC when an external USB-C earbud microphone is used to measure. You can see the lows are boosted by a lot - result of the microphone significant response roll-off in the lows.
The best-looking response (green solid line) is what I got when I used a calibration file with the same USB-C earbud microphone.
You can read how I calibrated the earbud microphone against a Cross-Spectrum Labs calibrated Dayton Audio EMM-6 measurement microphone in my thread on the WiiM Forum.

These were the RC settings I used in the app:

And this is how the correction results looks in the app:

What is missing?​

While WiiM RC in its current state works pretty well, I do see a lot of room for improvement in the future.

I summarized these in the related WiiM Forum thread, but I'd like to mention them here for completeness:

1. A pre-gain control is missing, which could be used to reduce the level going into EQ to avoid digital clipping; especially when positive-gain EQ filters are used!
2. Individual channel room correction when bass management is enabled is missing.
3. At the moment a single control defines both maximum positive and negative PEQ gain. Separate controls for the positive (boost) and the negative (cut) gain range of room correction filters would allow deep cut filters, but not strong boosts - which saves amplifier headroom and avoids temporal ringing.
4. The capability to import custom correction target curves is missing. At the moment there are only three predefined curves are available (B&K, Harman, Flat).
5. Perhaps the built-in microphone of the WiiM Voice Remote could be used in the Room Correction feature. This might be an easy way to more consistent correction results for many users without access to calibrated measurement microphones.
6. Variable smoothing option is missing. This would reduce the risk of over-correction in mid and high frequencies, while still allowing precise correction in low frequencies.
7. At the moment only sweep measurements with simple smoothing are supported, which is also not the best choice for corrections at mid and high frequencies. I'd like to see the option to use MMM instead since MMM gives much more consistent results between attempts and averages out a lot of the perceptually irrelevant reflections.
8. The default RC setting values could IMHO be improved to provide better correction for most users out of the box. Mainly to use a downward sloping target curve (B&K is fine), limit the correction range to below ~500Hz, only allow PEQ cuts (no boosts), and use some kind of variable smoothing.

Note that these only apply if you use WiiM automated Room Correction. You can still use PC with REW and a calibrated measurement microphone to manually measure your system, calculate there the PEQ parameter values against your preferred target curve in REW, and manually type these PEQ parameters in the WiiM PEQ function. So there is a way around the current RC limitations.

While not strictly part of RC, IMO the following features which are missing today would take WiiM RC a step further:

1. Simple tone control (individual treble-bass control, or a single "slope" control) to easily tune overall system tonality to taste (post-RC).
2. REW PEQ setting file import function.
3. A way to backup and restore configuration (including RC and EQ settings).
4. An automated loudspeaker correction function which could directly import PEQ presets based on anechoic loudspeaker measurements from e.g. the amazing spinorama.org maintained by @pierre. I guess something along the lines of AutoEQ app functionality for headphones. (link to related thread on the WiiM forum)

Summary​

All in all I have to say I'm quite happy with how WiiM RC works, especially given that it was added for free after many of the WiiM products were already in the market for quite a while. At the moment I'm still using my own REW generated correction filters loaded in WiiM EQ, but I can see myself moving to it if a few of the missing features are added in the future.

Hope this will be interesting to some!

 

[landco]

You then point the phone mic (or external USB mic, if you have one)

This is a lie. WiiM do not suppotr USB mic

 

[dominikz]

WiiM do not suppotr USB mic

It does, and the proof is in my previous post (screenshot from Android version of WiiM Home app v3.1.3 Room Correction function):

Actually, I successfully used both the Samsung USB-C headset microphone and a standalone Audio-Technica AT2020USB+ condenser USB microphone with the WiiM Home app. The USB microphone of course needs to be compliant with Android/iOS requirements for USB audio.

This is a lie.

The unwarranted ad-hominem accusation is not appreciated. Reported and blocked.