Skip to content
HomeRecordingSetup
Menu

Contact

Corrections, questions about a calculation, and partnership enquiries all come here. If you are reporting a wrong figure, please include the page and the value you believe is correct. Specification and acoustic data goes stale when manufacturers revise a product, and a correction from someone who owns the box is worth more than another round of desk research.

Your message is delivered by email and is not added to any mailing list. If you want the build guides emailed to you instead, the capture form on any complete build page does that.

What gets a reply and what does not

Corrections always get a reply, and if the correction is right the page gets updated. That is the most useful message anyone can send.

Questions about a specific calculation or a specific build usually get a reply, especially if they reveal that a page is unclear rather than wrong, because that is a page problem worth fixing.

Requests to review a product, place a link, or publish a guest article do not get a reply. Nothing on this site is placed in exchange for payment, and the recommendation logic is documented on the about page.

Before you write about a calculation

Two of the formulas on this site get questioned regularly and both are worth checking against before sending, because in almost every case the disagreement is about which figure is being quoted rather than about the arithmetic.

Room modes. The axial mode formula is the speed of sound divided by twice the dimension, so 1130 divided by 24 gives 47.1 Hz for a 12 foot wall. This calculator lists axial modes only and says so on the page. Tangential and oblique modes exist, are weaker, and are deliberately omitted because listing the full modal set of a bedroom produces a table where the useful information is in the first nine rows. See the room mode calculator.

Latency. Buffer latency is samples divided by sample rate times 1000, so 128 samples at 44.1 kHz is 2.90 ms. That figure is one direction. The number people compare it against is usually a round trip, which is roughly double plus 2 to 6 ms of converter and driver overhead. See the latency buffer calculator, which reports all three.

Elsewhere on the site