Share your in-room measurements?

[Deleted member 4708]

Does it make sense to change the target for Dirac to mimic a Harman curve? Is that supposed to be preferable to the flat Dirac curve that my system targets?
'cause if that's the case I wonder why Dirac doesn't offer that as an option (should be default option). Does Trinnov offer the Harman curve as a target?
I cannot help but notice that my Harman EQd Stax SR-009 and flat EQd Perlisten have different tonalities.

 

[ernestcarl]

Does it make sense to change the target for Dirac to mimic a Harman curve? Is that supposed to be preferable to the flat Dirac curve that my system targets?
'cause if that's the case I wonder why Dirac doesn't offer that as an option (should be default option). Does Trinnov offer the Harman curve as a target?
I cannot help but notice that my EQd Stax SR-009 and Perlisten have different tonalities.

Dunno about the better target curve...

But, speakers that measure (or have been EQ'd) to the same relative flatness will still sound different because of other things like off-axis directivity, transient decay, reflections and room characteristics.

 

[Deleted member 4708]

Nevermind my post! Here is a thread on it: https://www.audiosciencereview.com/forum/index.php?threads/harman-curve-for-loudspeakers.18782/

 

[kthulhutu]

Genelec 8341A pair with a 7360A and GLM EQ in a 10x9x8 room with a hard floor, treated sidewalls, treated front corners, treated backwall and a rug. Upper air was shelved in GLM due to it measuring weird ultrasonic buildup.
Single point measurements were taken and averaged from the listening position which is located in the approximate center of the room 125cm away from the speakers.

Not sure if it's possible to average these, so I took them from an R channel measurement

 

[flipflop]

Genelec 8341A pair with a 7360A and GLM EQ in a 10x9x8 room with a hard floor, treated sidewalls, treated front corners, treated backwall and a rug. Upper air was shelved in GLM due to it measuring weird ultrasonic buildup.
Single point measurements were taken and averaged from the listening position which is located in the approximate center of the room 125cm away from the speakers.

Not sure if it's possible to average these, so I took them from an R channel measurement

Looks very dry.
Is this system for recreational or professional use? Are you the only person using it?

 

[kthulhutu]

Looks very dry.
Is this system for recreational or professional use? Are you the only person using it?

Single person PC setup. Yes it is quite damped. I used to leave my sidewalls bare before I discovered Bacch. With Bacch I no longer feel the need to rely on early reflections and phantom images to get a sense of envelopment so I killed the troublesome slap echo.

 

[MCH]

Hi, general question. When you work on your room correction (manually, using rew+umik), what should you pay more attention to, the measurements of r and l channels separately or r+l measured simultaneously?
Thanks

 

[mightycicadalord]

I've always measured left and right separately, then averaged them and correct the average. Seems to be how all the automated ones do it as well.

 

[MCH]

I've always measured left and right separately, then averaged them and correct the average. Seems to be how all the automated ones do it as well.

Oh, never thought of that possibility. So at the end you have the same filter for both channels?

 

[mightycicadalord]

Oh, never thought of that possibility. So at the end you have the same filter for both channels?

Yeah, can't say I've ever tried individual correction on each channel, someone else will have to chime in on that.

 

[Sokel]

Couple of days ago.

 

[Weeb Labs]

Oh, never thought of that possibility. So at the end you have the same filter for both channels?

Channels should always be corrected individually. This is how most automated solutions such as Dirac Live, Audiolense and Audyssey work. We do not hear in mono and so if the channels are averaged, deviations unique to each can not be properly corrected.

 

[bennybbbx]

Hi, general question. When you work on your room correction (manually, using rew+umik), what should you pay more attention to, the measurements of r and l channels separately or r+l measured simultaneously?
Thanks

you can use a online tone generator that have the musical note frequency steps. https://www.szynalski.com/tone-generator/
then click on the note symbol (a 4) is default
Then click note steps and test if you hear the tone from middle. musical note steps is enough because if you hear music the music is record in this frequency raster with slight detune (+/- 10 cent) . It is also usefull to try out EQ settings with less group delay. I think dirac sound not good and what i see in measure is that dirac have lots of group delay jumps. see this post with measures and the eq settings i use. https://www.audiosciencereview.com/...ad-for-your-health.29127/page-17#post-1188953 if you have such a setting you can compare with longer hear if you like that more. lineaer phase in dirac or others does not help much. linear phase have in bass range lots transients problems and for room correction much db neeed because of room modes the automatic systems correct. here can hear an example at 1 min 45

 

[Chromatischism]

Oh, never thought of that possibility. So at the end you have the same filter for both channels?

Room correction doesn't use the same the filter for both channels. It uses the same target for both speakers in the speaker pair (L & R). Each will have different filters to reach that target.

 

[MCH]

Channels should always be corrected individually. This is how most automated solutions such as Dirac Live, Audiolense and Audyssey work. We do not hear in mono and so if the channels are averaged, deviations unique to each can not be properly corrected.

Room correction doesn't use the same the filter for both channels. It uses the same target for both speakers in the speaker pair (L & R). Each will have different filters to reach that target.

Ok, that's how i have done it so far and how it seems logical to me. But sometimes i am tempted to leave a peak in one channel because doing so the measurement of both channels together looks better. But i understand from your answers that i should be looking at channels individually both before and after correction. Is that correct?

 

[mightycicadalord]

Guess I've been doin it wrong lol

 

[Jukka]

If outputting L+R, the summed response is quite unpredictable for filter generation. Always separately, so you can even have a chance to get it right

 

[openbaffle3278]

Running Genelec 8250s w/ 7370a sub. GLM 4.2 just came out with their GRADE report. I'm still waiting on one K&M stand to get the right monitor off the desk. 7370a is under the desk. Listening distance is about .95m.

 

[Puddingbuks]

Monitors at different heights and depths?

 

[openbaffle3278]

Monitors at different heights and depths?

Yea, I am waiting for the second bracket for the 8000-400 stands. One was missing it in the box and they are out of stock everywhere. GLM happily calibrates and adds a .35ms delay and .3 dB attenuation to account for the differences. Not a permanent setup!