Something that puzzles me for a while. After calibrating a 7.1.4 room, using GLM 4.1 with cloud calibration, I fed full bandwidth pink noise to all speakers at the same time. Then I started to solo the speakers individually one by one. I found that some speakers sounded different in their response at the listening position. Not measuring, just listening.
At first I thought it was because of my ears position towards the speakers, for example the rear speakers and so on, but if I turn and face the speakers directly it´s the same. Even one of the front speakers sounds different than the other two.
I assume that one of the goals of calibration and using GLM is to ensure that all speakers are tonally matched frequency wise as much as possible, phase aligned etc minimising room acoustics interference on what we hear at mixing position. Assuming that this is correct, why do some speakers sound different? The grade report shows overall great response from the room.
Any explanation for this? Thanks!
Comments
Sorry for the delayed response, Paulo! Although we of course monitor this forum, please always contact us directly at support@genelec.com if you have a problem and need direct support. You'll usually receive a reply within 2 working days. What you describe has the following cause: GLM can compensate for influences like boosts in the frequency response, but notches will remain - for physical reasons. And you can of course hear a change in tonality if there are significant notches that are different for some speakers. Just have a look at the green curves in GLM and GRADE that show the resulting frequency responses at your listening position. But in general, the overall similarity of the speakers significantly improves after calibration. I'll open a support ticket for you to have a closer look at your setup's results. Greetings!
It's October 10th and Genelec hasn't responded to your question. That's a bit concerning.
Where's the support?
Hi Marcel, no problem. I´ll follow the ticket. Thanks.
Please sign in to leave a comment.