B&O Tech: BeoLab 90 – How to set up Active Room Calibration #64 in a series of articles about the technology behind Bang & Olufsen products
10 Responses on “B&O Tech: BeoLab 90 – How to set up Active Room Calibration”
We watched your video as we performed some measurements. The is a little difference in microphone placement described in the Beolab 90 manual (at least the copy I downloaded a year ago) and your video. You seem to be placing the microphone at the listening position first. Then higher behind and thirdly lower and in front. The manual describes the second measurement at ear level to the side and in front and the third is at the same level, to the other side and behind.
I am guessing the difference is minimal or have you changed your mind about how the measurements are best performed?
The question about height relates to whether you want to allow for listeners of different heights… If it’s always just you, and you are always in exactly the same place vertically (in other words, you never slouch…), then there is little advantage in changing the vertical position of the microphone. If, however, you expect to be telling a friend that is shorter or taller than you “just have a seat and listen”, then you want to get the measurements to cover the height difference. You won’t make the ARC result worse by changing vertical position – but you can make it better. I’ll change the manual to reflect this recommendation. (If we get a little geeky, essentially, what you’re doing by changing the vertical position of the microphone for the 3 measurements is to give a better representation / compensation for the vertical modes in the room around the listening position.)
Great video! Much better than reading a old fashioned manual! ;-)
Just made me wonder if there is any chance to see, in the future, a Room Compensation algorithm embedded in the BeoSystem 4…
Another point: don’t you already get several measurements of each speaker level while doing the Room Compensation procedure? Why do you still have to input the speaker level manually when creating a preset?
Of course, I cannot comment on anything outside the realm of currently-available hardware and software…
Regarding your question regarding automatic setup of Speaker Distance or Level using the ARC measurements: Of course, this is technically possible. However, doing so would introduce a User Interface or procedural issue. For example, we need 3 measurements for the ARC filter calculation for one zone. Which of those three measurements should be used for the distance or level? Or should we use 2 or 3 of the measurements to arrive at a best solution – or none? The only “correct” answer is to ask the user when the measurements are performed, which makes the measurement procedure “clunky”. The cleaner path is to separate the ARC measurements from the distance and level and make those independently adjustable.
The latest edition of the Technical Sound Guide for the BeoLab 90’s is now available online at this link.
You may notice that I’ve included your suggestions regarding height recommendations when doing the ARC measurements.
I’ve also added a “Tips and Tricks” section that may help with the setup procedure.
Well, I would use the average of all measurements as reference.
But the most important point is that you auto-complete the preset with this measurement while still allowing the manual adjustment for final fine tuning…
This way the ordinary customer won’t have to have a SPL metering device or learn how to properly use it (although this is quite straight forward to do).
Unfortunately, experience has shown that this will give you the wrong answer in enough cases to make it unreliable. This is primarily because the intention behind the room compensation measurements is different than the intention behind the Speaker Level measurements.
If everyone were doing the measurements the same way (e.g. for “one-chair-no-friends” OR for “the-whole-sofa”) with the same directivity setting in the loudspeakers (e.g. Narrow, front, long latency) then I might agree with you. However, when we’ve tried this, it has not worked similarly enough in these two cases to come up with an answer we’d use. There are many reasons for this, relating to a combination of the directivity of the loudspeakers, the contribution of the reflections, the distance to the speakers (and therefore the frequency-dependent critical distance of the room), the separation of the loudspeakers, and the distance between the microphone positions (and probably more stuff….).
Ultimately, we decided that, for now, it’s better to do nothing (and assume that the customer to NOT do the SPL measurements, leaving the Speaker Levels at 0 dB) than to put in what is probably the wrong answer.
I appreciate the updates to the Technical Sound Guide and since my comfy chair has a recline function I am going to redo the ARC to allow for some change in height.
Of course the other option is the have presets for various degrees of slumping.
But mostly I plan to sit back and enjoy the music.
Now that I have had time to play around with ARC and presets etc I have lots of combinations and set ups. More than any one needs, but all is part of the experience.
Something I learned:
Since I could not work out a way to play one speaker on its own I could not measure sound levels from each speaker separately.
Susan Vega (Tom’s Diner) to the rescue……I find that adjusting the speaker levels until she is in the middle works for me. Only needed for locations well “off centre”.
To be completely honest, that’s exactly how I do it. In the end, I find that doing it with Tom’s Diner (or the evening news) gives me a better result than if I do it with measurements…
– put in the Distance measurements first (using a laser thingy)
– put in a rough guess for the levels
– have a listen to Suzanne Vega
– Tweak the levels a little
– have another listen
– tweak the distance and level of one speaker (usually the closest one – but that’s just so that I drop levels instead of increasing them – this is just and old habit that started in my days as a recording engineer in an analogue studio. Attenuation is usually smarter than gain in that environment for SNR reasons…)
– Listen for a day or two and then have one last tweak, if necessary…