TOPIC:

Question: when setting up Dirac Live for nearfield monitoring (i.e. speakers close to the listening position), is it better to angle the measurement mic towards the ceiling, rather than pointing between the speakers, so as to capture more of the room acoustics and not just the speaker response?

Some background...

I’ve been using a minidsp DDRC-22A hardware box running Dirac Live for about a year now. Works great and I’m very happy with the results. My set up is a small control room (used for music production), involving a nearfield set up - i.e. a pair of speakers located quite close to my listening position, approximately 1.6m distance.

I’ve always followed the minidsp Dirac Live recommendation for a 2-channel stereo set up, which is to position the measurement mic on-axis (0 degree calibration file), pointing between the speakers. I’m now wondering whether I should instead be using the 90 degree file and point the mic to the ceiling?

My reasoning is that due to the close proximity of my nearfield set up, could it be that the using the mic on-axis means that I’m measuring more of the ‘speaker response’ than the ‘room response’ during the Dirac Live set up process, which may lead to sub-optimal results? Would I be better using the mic vertically to capture more of the room, given that the distance between the mic and speakers is only 1.6m when measuring?

I’m assuming that the Dirac Live manual is geared more towards home hi-fi / listening room environments, in which the speakers are positioned further away. Hence my question.

Grateful if anyone from the devteam or with experience of nearfield set ups could advise on what might be the best approach.

Many thanks!

@Skol303

I believe that we already answered that question before directly... MAybe some community members here will share more..
As far as we're concerned, directly pointing at the speaker is still the right way. Only if you have a 360 like speaker arrangement (i.e. 5.1/7.1) would we see some interest in having the mic pointing up, i.e. able to hear all speaker with the same angle of incidence.

Our 2cts.. Feel free to send your feedback from trialing..

DevTeam

devteam wrote:

As far as we're concerned, directly pointing at the speaker is still the right way. Only if you have a 360 like speaker arrangement (i.e. 5.1/7.1) would we see some interest in having the mic pointing up

Thanks for confirming. I’ll continue using the mic on-axis as you recommend.

I asked Dirac for their recommendation for using the UMIK-1 with Dirac Live RCS to set up a 2 channel stereo system. This is their response.

"By pointing the microphone upwards you will get the same coloration on the sound coming from the backward/side/front wall as the direct wave from the speaker. This is important in the filter deign process so we don't do any wrong decisions. If you point the microphone to the speakers instead, the coloration of the room will differ from the direct wave, this could result in filters which are overcompensated in certain frequencies. For our room correction system I would always recommend to point the microphone upwards.

UMIK-1 instructions from MiniDSP.

For Dirac Live applications (e.g. DDRC-88A, nanoAVR DL), you should always use the 90deg file."

Seems pretty clear to me. Not sure why the 0 deg. recommendation keeps coming up. Both the Dirac Live & MiniDSP instructions agree that the 90 deg. orientation is what Dirac Live software is designed to process.
Kev

bugeyed wrote:

I asked Dirac for their recommendation for using the UMIK-1 with Dirac Live RCS to set up a 2 channel stereo system. This is their response.

Interesting, thanks Kev. I asked the same question to Dirac Live just a few days ago and this is the response I received:

“In an ideal world the mic orientation of an omnidirectional mic would make no difference... calibration files try to correct the response of "non perfect" omnidirectional mics but they are not perfect either as they ideal at one angle only. Since the magnitude response of the microphone is (slightly) different for every angle its calibration curve is only valid for a single angle... that raises the question, which angle to chose? One may point it towards the speaker if mostly interested in the direct wave of speakers in front of the microphone (i.e. a stereo set-up) or point it towards the ceiling in case of home theater set-ups with ceiling speakers."

MiniDSP do indeed recommend using the 90deg file for products designed for surround sound systems (the DDRC-88A and nanoAVR DL as you mention). But their recommendations for 2 channel stereo applications is to point the mic on-axis (0-deg file), as follows:

UMIK-1 set up instructions:
Which calibration file should I use and where to point the UMIK-1?
We provide two calibration files to be used depending on your application.

- For stereo system (e.g. 2ch dirac live, single speaker measurement), use the 0deg file and point the UMIK-1 at the speakers
- For multichannel system (E.g. 5.1/7.1) or a surround application where multiple speakers are spreadout around the room, use the 90deg file and point the UMIK-1 at the ceiling.

DDRC-22 series user manual:


SHD Dirac Live series user manual:


“For stereo systems, we generally recommend pointing the microphone between the two speakers”.

bugeyed wrote:

Seems pretty clear to me. Not sure why the 0 deg. recommendation keeps coming up.

So perhaps not as clear as it may seem, as there appear to be some contradictions - and hence my confusion

As it happens, I've since run some REW measurements sweeps in my space (a heavily treated mix room) and the frequency response results for the 0-degree and 90-degree calibrations are almost identical, but that may be because my room has extensive acoustic treatment / low reflections. So in my case I won't be re-calibrating Dirac Live using the mic at 90-degree axis, but in a room with no acoustic treatment the choice of angle for the mic may of course have a more significant influence on the final result (?).

For reference - comparative room measurement with Dirac Live engaged: blue taken using UMIK-1 mic on-axis (0-degree); red shows measurement with mic angled vertically (90-degrees).

I think that the overriding fact is that Dirac recommends the 90 deg. mic orientation! MiniDSP also recommends 90 deg for Dirac software. Dirac Live RCS software is built on the basis of the microphone "listening" while pointed at the ceiling. Dirac & MiniDSP both recommend this position. I think your last statement is the most important thing. Regardless of whether you have a 2 channel stereo setup or multi channel surround sound, Dirac should be listening to the room to get the proper results from their correction software.
Kev

bugeyed wrote:

I think that the overriding fact is that Dirac recommends the 90 deg. mic orientation! MiniDSP also recommends 90 deg for Dirac software.

I think there may be some confusion. MiniDSP recommends 90 deg for Dirac software when setting up a surround sound system; but they recommend the 0 deg angle/calibration when setting up a two-channel stereo system. This is clearly stated in their product user manuals (examples in my post above).

The Dirac Live user manual has this to say about angling the measurement mic:
"Direct the microphone upward, pointing to the ceiling, to get the most omnidirectional recording of the room response, or towards the speakers depending on the particular microphone and its calibration file (if any). Please follow the instructions from your supplier."

Which kind of equates to: “Point it one way, or the other, depending…” Hence my reason for contacting them to confirm. The response I received ("point it on-axis for stereo") contradicts the answer you received from them yourself ("point it vertically"). And if we "follow the instructions from our supplier", as suggested, then for miniDSP products we're recommended to point the mic at 0-degrees.

As I understand it, Dirac Live (and all DSP software) is concerned with "correcting" the speakers and not the room. So to me it seems logical that we should be most concerned with the direct speaker response at the listening position (specifically for a stereo set up). But I started this thread because I wondered whether the direct response in a nearfield set up - i.e. monitors very close to listener - would skew the results, and perhaps the mic should be angled vertically in typical nearfield studio set ups. To which both the miniDSP and Dirac teams have replied: keep the mic on-axis. Clear as mud, right?

Not trying to challenge your view here. Ultimately, if it works - it works. And thankfully, in my case it works. But nowhere in the miniDSP documentation does it recommend pointing the mic at 90 degrees when setting up Dirac Live for stereo systems. Quite the opposite. Hence my request for some clarity.

Previous reply from devteam on this topic:
www.minidsp.com/forum/dirac-series-suppo...for-dirac-live#27758

When I said it seems pretty clear I obviously meant that it isn’t clear at all. I have sent a query to Dirac, again, & included your differing response from them. Hopefully Dirac will, once & for all decide what their recommendation is!
kev

Thanks Kev!

Let’s hope we get a clear answer. Third time lucky

Skol303 wrote:

bugeyed wrote:

.
As I understand it, Dirac Live (and all DSP software) is concerned with "correcting" the speakers and not the room. .
Previous reply from devteam on this topic:
www.minidsp.com/forum/dirac-series-suppo...for-dirac-live#27758

While waiting for a second response from Dirac I noticed the above statement that you made. My understanding is that Dirac Live designs filters to correct for unwanted "room" interaction. Obviously they can only do this by altering the speaker response, but their focus is on room correction. Dirac states: "By pointing the microphone upwards you will get the same coloration on the sound coming from the backward/side/front wall as the direct wave from the speaker. This is important in the filter deign process so we don't do any wrong decisions. If you point the microphone to the speakers instead, the coloration of the room will differ from the direct wave, this could result in filters which are overcompensated in certain frequencies. For our room correction system I would always recommend to point the microphone upwards."
In your nearfield setup it may not matter since you have effectively taken the room out of the equation. Not to mention your room treatment. In my setup the room has minimal treatment, mainly at first reflection points & various objects working as diffusers. The Dirac Live filters (90 deg) have made quite a big difference in balancing the left right tonal balance, spatial queues, center image clarity & low frequency modes. BTW I am running standalone Dirac Live 1.x with beta AU and VST plugins. I use the VST plug in with JRiver Media Center 24 on a Mac Mini. I have a high end DAC so the standalone software is my preferred choice,
Kev

^ Good point about individual room set ups. “One size” perhaps does not “fit all”.

I started this thread to question whether using the mic on-axis for nearfield set up of Dirac Live might skew the results a little (ie. direct speaker response being too dominant due to short distance between speakers and mic at listening position).

From recent testing using REW, I’ve discovered that it makes literally no difference whether I use the 0-degree or 90-degree mic calibration: the acoustic measurements appear identical (frequency response, decay, ETC, etc). So to answer my initial question, I don’t think it matters which mic angle/calibration is used my case for Dirac Live set up: the filters are responding to the same acoustic information either way. But that’s specific to my room and perhaps others like it - ie. environments that are designed to ‘remove’ or certainly reduce the effects of room acoustics through extensive acoustic treatment.

Certainly, when measuring room acoustics, many acousticians recommend angling the mic upwards to capture a more balanced ‘picture’ of the room - eg. as a basis for designing acoustic treatment. So there’s plenty of evidence that supports using the mic that way (90-degrees).

Ultimately, the confusing/contradictory advice we’re getting from the devs at miniDSP and Dirac Live may just stem from there being no clear “right or wrong” answer: that it depends on the individual circumstances of the room. Even if so, I think both companies should provide more explanation on this topic as for me, it is quite fundamental to how their products perform.

As mentioned, the response from miniDSP has always been “point the mic on-axis between the speakers for 2 channel stereo set up” (as seen above in the thread), but I’ve questioned that myself. Hopefully the folks at Dirac can shed some further light on it

Regardless, if using 90-degree calibration produces good results in your own room, then that’s obviously the “right” way.

Exactly! in my latest query to Dirac I included the fact that MiniDSP & Dirac support are providing conflicting advice.

Kev

bugeyed wrote:

in my latest query to Dirac I included the fact that MiniDSP & Dirac support are providing conflicting advice.

Whilst we're awaiting a response from the Dirac HelpDesk, I've done some background reading into how the Dirac Live software functions, which might shed some light on the subject.

Here's a brief summary of the key points relevant to using the measurement mic on-axis or vertically. Bear in mind I've had to interpret this information myself, so it may not be correct...

1) Dirac Live's approach to impulse response correction (time domain) is concerned with the direct wave and very early reflections. So that's the sound coming from the speakers, with minimal influence from the room itself.

"Dirac Live does not address late-reflections since they are too position-dependent to address well via processing. Late-reflections can add a sense of spaciousness, but those who do want to rid themselves of late-reflections are advised to use acoustic treatments for that goal" (link to source).

2) After addressing the time-domain behaviour, the software then attempts to correct late reverberation in the room that causes "coloration". It corrects this in the frequency domain (so that's not just the direct sound from the speakers, but also the reflected sound in the room).

Here's a summary of the measurements used by Dirac Live in its calculations:

• Room Resonances/Standing Waves: low frequencies, large wavelength, late sound, all directions.
• Misalignment of Drivers: mid to high frequencies, direct sound, same direction as speaker.
• Early Reflections/Cabinet Diffraction: mid to high frequencies, early sound, same direction as speaker.
• Late Reflections/Room Reverberation: mid to high frequencies, late sound, all directions.

So what does this tell us? Well, it doesn't suggest a clear preference either way for how to position the mic...!

We know that Dirac's time/impulse correction is based on the direct sound, which supports pointing the mic on-axis between stereo speakers... but the 90-degree calibration file supplied with the UMIK-1 mic mostly compensates for this and captures a sufficiently clear picture of the direct sound (at least clear enough for this purpose). From my own testing the variance between the direct field on/off axis is around 1dB at high frequencies, so quite negligible.

We know that late-reflections are discounted by the impulse response correction, but instead are dealt with by frequency domain correction, which supports pointing the mic vertically. Doing so would help to capture late-reflections from the side/rear walls (as well as the direct speaker response), so using the 90-degree mic calibration for this purpose would seem sensible if your room has no acoustic treatment (i.e. in rooms where late-reflections are not dampened).

Conclusion?

It appears to me that using the mic on-axis is ideal for time domain correction (as it concerns the direct speaker response); whereas using the mic vertically may lead to more accurate frequency domain correction (as it concerns late-reflections in addition to the direct response). So take your pick!

Personally speaking, in an untreated room I think I'd opt for using the mic vertically (as mentioned by Kev in this thread). In my own treated room, using the mic on-axis seems appropriate due the late-reflections being heavily dampened by acoustic panels (hence the frequency response measuring almost identical no matter how mic is pointed).

Still, it would be good to get some clarity and consistency on this topic from the experts at miniDSP and Dirac Live, as their own recommendations are somewhat contradictory.

References:
www.dirac.com/live-home-professional-audio-info
www.dirac.com/dirac-blog/2016/7/1/4-types-of-distortion
www.audioholics.com/room-acoustics/dirac-live-update

Thanks Skol303,
Nice summary. I am still waiting for a response from Dirac. I am thinking I may need to poke them & make sure they received my query. On the other hand, maybe it's been escalated & we will get a definitive answer from them.
Kev

Skol303 wrote:

using the mic vertically may lead to more accurate frequency domain correction (as it concerns late-reflections in addition to the direct response).

You appear to be assuming that late reflections are primarily horizontal. But reflections are in all directions (aren't they?)