Did everyone see the Decware YouTube video today where Steve gives us a audio analyzer for our phones?

https://www.youtube.com/watch?v=-BNngpS1E90&t=353s

Very cool! Going to try this out for sure.

Thank you! Appreciate it
Scott

I played around with it last night.  I have 2 systems one in an untreated family room which has a Home theater along with 2 channel channel and another in an untreated living room with my main 2 channel.  Both rooms ~ 26x16 with 8-10 foot ceilings.  Noise floor at night measured 0.  Family room score was 78 and living room was 76.  I didn’t see it change much with music change but I did listen to similar music.  The scores did change slightly as I moved around the room.  Still trying to figure out how to use this tool to understand room or set up flaws and how to make changes to improve.  Fun easy to use tool.  Thanks!

I learned the app is only as good as the mic in my phone. When I tried, on an iPhone 16E, it said my ambient noise level was about 55 dB in a quiet house in a quiet neighborhood. That seems like at least 20 dB too high. So when I play music at normal volume my dynamic range is 2dB. When I crank up the 4th movement of Scheherazade, my go to demo track it says the dynamic range is 15 dB. The scores ranged from the 20s to 50. If I can pry my wife’s phone away I’m going to try it with hers.

I think Steve is just playing a joke on us.

I'm with John and James - My system has never sounded better, but I could not get a score above 50.

The app can only be as good as the phone and it's mic, as well as the particular track played.

I don't think it's a joke, and of course we all want our systems to sound amazing.  

I can't look at a data driven process without digging into the parameters of the process.  The documentation shows the four weighted categories used to calculate the score. Considering "Spectral Balance" is 35% of the score, I have to interpret this to mean only a full frequency sweep within the sampling time would provide a high score in this category.  This aligns with my findings that my score varies by 40 points just based on what track I am listening to.  One way to further test this is by extending your sample time on any track and I bet your score will improve, just because the algorithm is finding more frequencies.  If the track you play isn't even producing certain frequencies your score will be lower (my assumption).

That's about all I know - I can't speak to programming or much beyond that.  The important thing always returns to the individual being happy with what they are hearing.

All in all, this is kind of a cool tool Steve has built, and maybe the best way to think about it as a step one towards what it will be in the future and not an end-of-story finished product.  

I forgot to put the direct link to the app.
This way you don't need to pull up the YouTube video.

https://audio-analyzer-dial--decware.replit.app/

The C-9 did not effect the score. Music selection definitely does. The spread of scores in my system's loop A goes from 66 to 74. It will be interesting to see how the other 2 loops will score when time is available.

Spinning Los Lobo's Kiko fetched the following score:

74
|
03-04 05:28
|
NF 1
PK 83
DR 82dB
BAL 25.2

Couple things;

There is never a time when the noise floor will be at zero, so the phone may have noise cancellation or something running in the background.

Sound Quality Meter - attempts to put a rough measurement on sound quality.  Like your ears, it has no concept of stereo gear, rooms, speakers, or anything else.  It only knows if it likes it or not.  You can read exactly how it scores in the Technical Info in the app.  So, yes, the type of music played is a large factor.

Jazz — especially small ensemble or big band recordings — tends to have a very even frequency spread (upright bass, piano, horns, cymbals all covering different ranges), moderate dynamics, and clean recordings with low distortion. Great for Spectral Balance.

Classical — orchestral recordings fill the spectrum beautifully and have excellent dynamic range. The challenge is that many classical pieces have very quiet passages that might not stay far enough above the noise floor consistently. But during full orchestral sections, scores should be very high.

Rock — tends to be more compressed (less dynamic range) and can have heavier energy in the mids and lows with less even treble coverage. Heavily distorted electric guitars will also push up the harmonic distortion readings. It'll score decent but typically lower than the other two.

If you want to test for the absolute best score, try something like a well-recorded big band jazz album (Count Basie, Wynton Marsalis) or a full symphony at a good listening volume. That should push the dial the highest.

Steve