• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

HiFiBerry Room Correction

auxout

Member
Joined
Apr 26, 2022
Messages
7
Likes
5
Hello everyone,

it took me by surprise when today I learned that HiFiBerryOS supports room correction.

This might be an interesting, somewhat tinkery budget option. Next to a microphone, it needs a DSP add-on card.
The calculations are done on a cloud server.
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,413
Likes
18,396
Location
Netherlands
I tried it once, didn’t work for me. For some reason, the measurements done with the UMIK were totally borked. Eventually abandoned it.

The version I tested did not do room correction in the cloud though. It was all local. A PI is more than capable of doing that, so I don’t see the point either. I call bullshit on the FAQ regarding this.
 
OP
A

auxout

Member
Joined
Apr 26, 2022
Messages
7
Likes
5
I tried it once, didn’t work for me. For some reason, the measurements done with the UMIK were totally borked. Eventually abandoned it.

The version I tested did not do room correction in the cloud though. It was all local. A PI is more than capable of doing that, so I don’t see the point either. I call bullshit on the FAQ regarding this.

Thank you for your fast, interesting insight! The cloud part struck me, too. Their FAQ states that they use the data to optimize their algorithm:
"Handling this on a cloud server allows us to run optimisations that the Raspberry Pi wouldn’t be able to handle, because it lacks processing power. We also use these data to improve the optimizer."
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,413
Likes
18,396
Location
Netherlands
Thank you for your fast, interesting insight! The cloud part struck me, too. Their FAQ states that they use the data to optimize their algorithm:
"Handling this on a cloud server allows us to run optimisations that the Raspberry Pi wouldn’t be able to handle, because it lacks processing power. We also use these data to improve the optimizer."
Yeah, the last part is probably the major thing.

If they improved the measurement part, I don’t know. The UX was pretty good though.
 
Last edited:
Top Bottom