Room correction or speaker correction?

[Dvddvd]

I read that room correction programs installed in HIFI amps and AV processors can do wonders to set ups? Dirac, Audyssey, Room perfect, Arc etc

Anybody use them?

Are they correcting the room or the speakers? Or are both the same really?

They measure the output and then put filters to get a preferred frequency curve and then you can adjust to your liking, I suppose it's EQ with the added bonus of altering phase, levels etc

 

[PuritéAudio]

A touch of EQ below a few hundred hertz can be very valuable , low bass room modes are very difficult to treat passively , EQ is available in playback software, DACs and loudspeakers themselves.

Keith

 

[Tony_J]

I use Dirac as part of my active setup and it makes a significant (positive) difference.

 

[tuga]

I use REW filters to cut off a couple of marring peaks generated by my listening room below 100Hz.

P.S. I would add that I am correcting the response of my speakers in this specific room and for this particular speaker and listener location.

 

[Tony_J]

P.S. I would add that I am correcting the response of my speakers in this specific room and for this particular speaker and listener location.

Unless you do the REW etc. measurement anechoically, that is of course always the case.

 

[rabski]

I used the basic audyssey on our AV setup when I replaced the sub. There is one (large) sub in a corner, and although I could have done it 'by ear', it would have taken ages. Audyssey took one 'sweep' to kill off an unpleasant boom and it now sounds more than good enough.

I've ever bothered in the 'proper' system, because my room is fortunately very benign so I've never felt the need. If I did, I wouldn't worry about using any of the systems. Unless I'm wrong, they only really attack low frequencies (which is where the ssues usually are).

 

[Dvddvd]

I read that most of them don't do much with the bass? I do have at present a miniDSP which is a tweekers paradise. I bought the added plug in as I had three subs at the time( it will eq four subs)

 

[Tony_J]

Unless I'm wrong, they only really attack low frequencies (which is where the ssues usually are).

With Dirac, you can choose the frequency range over which the eq is applied and give it a target curve to aim at, but the significant difference (IMHO) comes from the fact that it doesn't just adjust levels to smooth out dips and humps, it also corrects phase/timing issues. With my current setup, I am using it over the full frequency range of the speakers.

 

[GJO]

Have been playing around with my speaker's as they have the bass drivers on the side and I have always used them firing inwards,the first graph is firing in,the other firing out,was surprised at the difference,but the Dirac software sorts it out fine,undecided as yet what I will settle on..

A friend of mine is currently trialing the Dirac 2.0 beta program and it seems to be quite an improvement,will get to hear it next week sometime.

 

[MartinC]

Speaker output correction and what might be termed 'room correction' are very different things. 

For speaker output adjustment you need to make a time-gated measurement that allows you to exclude all boundary reflections and so just see the direct signal from the speaker, like I did here:



Measurements made at the normal listening position without time gating show the sum of the direct signal and the reflected signals. What it makes most sense to correct for on these is low frequency response. In the following ASR thread that is worth a read, Floyd Toole is suggesting this be done up to about 500 Hz for typical domestic rooms. Others suggest the (lower) Schroeder frequency.

https://www.******************.com/forum/index.php?threads/an-enticing-marketing-story-theory-without-measurement.7127/

I haven't quite got to doing this myself yet but I think what makes most sense is to combine the above two processes: fix room mode problems at low frequencies but only make speaker output corrections at higher frequencies.

Edit: the reason not to start making 'corrections' to room measurements at higher frequencies is that in doing so you will be corrupting the direct signal from the speaker. Floyd Toole's suggestion is that if any changes are made at higher frequencies then they should be broadband like traditional tone controls, rather than trying to address particular peaks.

 

[MartinC]

I do have at present a miniDSP which is a tweekers paradise. I bought the added plug in as I had three subs at the time( it will eq four subs)

Do you also have a microphone?

 

[Dvddvd]

Yes the calibrated Umic-1

 

[tuga]

For speaker output adjustment you need to make a time-gated measurement that allows you to exclude all boundary reflections and so just see the direct signal from the speaker

As far as I know gated measurements are unable to exclude low frequency boundary reflections, which is why both Stereophile and Hi-Fi News never plot information below 300Hz in their CSD measurements:

https://www.stereophile.com/content/measuring-loudspeakers-part-two-page-5

Alan Shaw comments on pseudo-anechoic measurements here:

https://www.stereophile.com/content/harbeth-m401-loudspeaker-manufacturers-response

 

[MartinC]

Yes the calibrated Umic-1

OK, so you can make some measurements if you're curious then. As you've got the kit you can do the speaker comparison you discussed in your other thread as well.

One tip: for any measurements you take, only have one speaker running at a time.

 

[MartinC]

As far as I know gated measurements are unable to exclude low frequency boundary reflections, which is why both Stereophile and Hi-Fi News never plot information below 300Hz in their CSD measurements:

https://www.stereophile.com/content/measuring-loudspeakers-part-two-page-5

Alan Shaw comments on pseudo-anechoic measurements here:

https://www.stereophile.com/content/harbeth-m401-loudspeaker-manufacturers-response

Yes, absolutely. If you look at the thread I linked to you'll see there is a low frequency limit.

(The minimum frequency where there is any data is at the reciprocal of the time window length.)

 

[MartinC]

@tuga fortuitously the region where you can't make in-room gated speaker measurements is also the frequency range where actually room modes are dominant such that you wouldn't want to  .

 

[MartinC]

I posted this on @Dvddvd's other thread a while ago, but the other limitation of gated in-room measurements is that the frequency resolution is relatively low (being equal to the reciprocal of the time window length).

Making measurements outside is one way to improve things, with how high you can support the speaker and the microphone becoming the limiting factor. Well, as well as the weather and how weird you mind your neighbours thinking you are  .

 

[Nopiano]

I think the answer to your questions: ‘Are they correcting the room or the speakers? Or are both the same really?’ is they aren’t exactly correcting either, but are more about adjusting the effect of the speaker within the room (in the location the user has chosen).   

As mentioned already, some speakers can adjust themselves (like the Kii and D&D that Keith sells), but these are usually rather beyond the budget of most home listeners.  

As well as the systems already stated above, there is also Linn’s proprietary Sound Optimisation system, included in all their DS and DSM digital streamers and preamps.  Even the cheapest models gained this retrospectively, for the cost of a free software update.  Unlike those that measure the actual speaker/room output, this adjusts the speaker output based on modelling of room dimensions, construction, and the locations of doors and windows, etc.  

https://www.linn.co.uk/technology/space-optimisation

I wasn’t too convinced at first, but after living with some tweaked settings for a few months, I was recently surprised how poor it sounded when switched off.  My speakers have since been measured, so I’ll be creating a new profile soon.   

 

[tuga]

I wasn’t too convinced at first, but after living with some tweaked settings for a few months, I was recently surprised how poor it sounded when switched off.

This is my argument against A-B comparisons.