0
\$\begingroup\$

Imagine someone used @channel in a Slack channel with 100,000 people. How many watt-hours are used to play the Slack notification on the laptops for these 100,000 people? This is a bit of a silly calculation, so some handwaving is fine.

Assumptions

  1. Every person has the same laptop speakers with the same volume settings.
  2. Each laptop plays the same notification sound for 1 second.
  3. The content of the sound clip can have whatever properties make the calculation easier. Need it to be a constant frequency? That's fine.
  4. The sound was measured in a fancy anechoic chamber by a not-so-fancy Android app called Sound Meter. I believe the app measures Sound Pressure Level (SPL) in decibels. The result was an uncomfortable 80 dB for the duration of the 1-second notification, measured from 1 meter away. I'm not sure how the measurement relates to different frequencies in the signal, thus Assumption #3 allowing for handwaving.

In my simplistic understanding, you can calculate the watts used by a speaker given A) the Sound Pressure Level (SPL) output in dB, and B) an efficiency rating of how well the speaker turns watts into sound. From there, you'd use the watts and the time duration to calculate the watt-hours. Is this correct?

I've provided the details on A in the assumptions, though I'm hoping an answerer to can fill in the gap on B with the numbers from some arbitrarily chosen laptop speaker out there. I don't know the terms to Google to find such details on any speaker.

\$\endgroup\$
5
  • \$\begingroup\$ Just to be clear, you are interested solely by the power used by the speaker, right? Because if you want to include CPU, audio drivers/DACs, or even storage, network, etc. then it becomes a lot harder… \$\endgroup\$
    – jcaron
    Commented Apr 3, 2023 at 21:41
  • \$\begingroup\$ A typical shelf speaker has sensitivity of 80 dB SPL at 1m at 1 watt. So 1 watt would be consumed by one speaker in one second or 1/3600 hours. But laptops don't have shelf speakers, and even the infrastructure of providing a powered-on laptop with powering the internet connectivity across continents for the duration of the testing will make the calculations useless. \$\endgroup\$
    – Justme
    Commented Apr 3, 2023 at 22:01
  • \$\begingroup\$ For a rough estimation you could use the nameplate rating of the speakers and the SPL of some loud music turned up to high volume, then compare that to the SPL of the notification \$\endgroup\$ Commented Apr 3, 2023 at 22:10
  • \$\begingroup\$ @jcaron, yes I'm just interested in the delta in power consumption between normal operation of the laptop and the moments the speaker is playing. I think you and Justme are correct that the rest of the system and infrastructure dwarfs the cost of powering the speakers briefly. \$\endgroup\$ Commented Apr 4, 2023 at 13:32
  • \$\begingroup\$ In any case, I think "Sensitivity" was the word I was missing to Google the numbers I need, thank you @Justme I'm happy to accept some answer with that explanation \$\endgroup\$ Commented Apr 4, 2023 at 13:35

1 Answer 1

0
\$\begingroup\$

The only thing missing is the relation between power fed into speaker and resulting sound level.

This is called speaker sensitivity.

The sensitivity will vary between speakers, but a typical average loudspeaker might have a sensitivity of 80 dB measured at 1 meter when driven at 1 watt.

Which neatly converts back your measurement setup to 1 watt.

On the other hand, laptop speakers may be optimized for sensitivity so you might have sensitivity of 90 dB. Which means you only need 0.1 watts to produce 80 dB SPL.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.