How accurate is the DecibelX Pro sound level meter app for iOS?
HTML-код
- Опубликовано: 19 янв 2022
- This video was specifically created to answer questions posed by students in the graduate level noise control engineering course I teach at Penn State. I assign several hands-on projects that require students to obtain a Type 2 quality sound level (standalone sound level meter device or an accurate app for mobile phone).
This video shows that while the DecibelX Pro sound level meter app for iOS looks really nice and has some excellent features, it has accuracy problems when attempting to calibrate it with an external microphone (necessary to achieve Type 2 accuracy).
Along with this video I made several other videos comparing other apps, and calibration:
Faber Acoustical SignalScope X (on iPhone and iPad) -- • Comparing the accuracy...
Calibrating the Faber Acoustics SignalScope X -- • Calibrating an externa...
NIOSH Sound Level App (on iPhone and iPad) -- • Comparing the accuracy...
Calibrating the NIOSH Sound Level App -- • Calibrating an externa...
The app is made for the built-in microphones and any external microphone connected to your iPhone would need a calibration file.
Very good video, thank you. Question: what calibrated mic is that? It comes with the adaptor to connect to the phone?
the external mic I used was a MicW i437L Omnidirectional Lightning Microphone (sells for about $170 at places like Sweetwater or B&H Video-Audio. The Calibrator is a Larson Davis CAL200 (now owned by PCB Electronics) and is much more difficult to find (sells for about $420).
@@DanRussellPSU Thank you!
Hope the app creator see this and implement a 2 point calibration.
Dear Dan thanks for the video. I have a question regarding the calibration: Is an adadpter necessary? What is the diameter of the microphone? and for which size of microphones is the calibrator? Rainhard Koch, Germany
The microphone I used in this video (i437L by MicW) was a 1/4-inch microphone and the calibrator I was using (Larson Davis CAL200) has a 1/2-inch mic opening, so the adapter was absolutely necessary. The calibrator must have a tight seal (with the correct volume) in order for the calibration to be accurate.
Will this app catch ultrasonic sounds like those coming off a dog deterrent?
Could you please recommend a mic that can be used with this app and is compatible with the calibration kit? Thanks
my preferred external microphone to use with iPhone and iPad is the MicW i437L . . . 1/4-inch mic that plugs into the lightning port. Available directly from MicWaudio.com or other places like bhphotovideo.com for about $160 USD.
Could you please provide link to the microphone used.? Also we were interested in Infrasound and ultrasound mics that are compatible with the iphone. Do you happen to know a nice software and hardware pair?
The external mic shown in this video is a i437L from MicW. If you do a Google search for "micw i437l omnidirectional lightning microphone" you'll get links where you can purchase this mic from Amazon, Sweetwater Sound, and B&H Photo Video, alf for about the same price. I believe this is a mic that works in the audio range (20 Hz - 20 kHz). I do not know of any mics that will work with an iPhone for infrasound or ultrasound. I'm also not sure that the internal electronics for an iPhone would support those frequency ranges.
Dear Dan, thanks for your video. I wonder why sound meter calibrators use 94 and 114 dB at 1KHz.
The calibration frequency of 1 kHz (1000 Hz) is used because all of the weightings (A, B, C, D) have exactly the same value at 1000 Hz. Below and above 1 kHz, the A and C weightings have very different corrections. 1000 Hz is the one frequency where there is no correction needed for any of the weighting networks.
94 dB corresponds to a sound with a root-mean-square (rms) pressure amplitude of 1 Pascal
114 dB corresponds to a sound with a rms pressure amplitude of 10 Pascals.
Funny enough, I bought an Apple watch SE and then calibrated my decibel X app to latch the Apple Watch in the A weight scale, but then switched to the z weight scale for all my hearing safety measurements. So effectively I did the same think a 1khz frequency does but for just $200.
What scares me is how loud driving a car is.
@@DanRussellPSUSo the rms pressure amplitude increases by a factor of 10 for every increase of 20 dB?
@@brettschmidt5929 yes -- every time the rms pressure changes by a factor of 10, the sound pressure level changes by 20 dB.
I have an iPad Pro, what's your guess on how accurate the iPad microphone is with this app you use?
My iPad Pro shows a huge 30db quieter than my iPhone - both with decibel X. Not sure why 🤷♂️
It's good that you demonstrate how inaccurate the app is at higher SPLs, but you could also show how it performs side by side with a Class 1 SLM, to demonstrate comparative measurements at around 50-60 dB.
That is a good idea . . . I don't happen to have a Type/Class 1 meter handy at the moment (those are more expensive than I can afford, though my department might have one in our labs). I made this video to share with students in the Noise Control Engineering course I was teaching in Spring 2022 -- I had required them to purchase a Type/Class 2 sound level meter or an external microphone and an App that could be calibrated. They used these meters/apps for several measurement projects during the semester. A number of students wanted to use the DecibelX Pro app (because it has a front end that looks really nice) and I posted this video to provide some caution about its limitations.
how accurate is it?
I am trying to capture and record my noisy upstairs neighbor with stomping, walking hard, and pounding noises using the dB X app on my iPhone 12. How do I remove background noise so I can hear the loud stomping? How do I read it to prove to my landlord it is excessive noise preventing me from having a peaceful enviornment? I really need help please!
I second this. Environmental noise is a problem.
I suppose the app is not the issue. The issue could be on the ADC or the mic itself not being linear at high SPL.
the external mic is very linear (I've tested it). And I've used the same mic with two other apps (NIOSH sound level meter and SignalScopePro by FaberAcoustical) that do not have this problem at all . . . thus it appears to be the app.
@@DanRussellPSU Hi Dr. Russell, Thanks for your reply. I watched the video again, and noticed excessive harmonic distortion in the FFT plot when you bumped up the calibrator setting to 114. This could be an indication that the signal that the app is receiving is near clipping. Maybe the app is adding some gain since you mentioned that it works fine with other apps. Like someone else mentioned below, it might work just fine with the built-in mics that have less sensitivity.
@@waynepark I'll repeat that using exactly the same microphone, same calibrator, same iPhone with two other apps (SignalScope X and NIOSH) did not have the level errors that DecibelX Pro shows, so it is something with the app. (I've updated the text description to include links to the videos where I compare those two apps). I"m not saying that this app isn't a good app for general purpose use. But, I am saying that it is not acceptable if you need Type 2 level accuracy that conforms to ANSI and ISO standards. I created this app because I teach a graduate level noise control engineering course and require my students to obtain a sound level meter that meets Type 2 specifications (either a standalone sound level meter or a combination of external microphone and app). Several students asked about DecibelX Pro because it is extremely affordable. However, this app does not have a calibration option, and as this video shows (compared to the other apps I tested using the same equipment and the same procedure) this app does not have the accuracy required to satisfy Type 2 specs across the range of sound pressure levels.
Unfortunately, there is no way to calibrate the built-in mic so there is no way to know how well the built in mic would perform this same test -- and the DecibelX app does not have a calibration feature. Also, Apple has built in quite a bit of signal processing and circuitry into the audio system to enhance the audio quality for voice (using it as an actual phone) -- and the FaberAcoustic SignalScopeX and NIOSH apps had to "undo" the built-in Apple signal processing and filtering in order to get their apps to conform to ASNI and ISO standards. I'm guessing that this is not the case for DecibleX Pro, which is only costs a couple of dollars. NIOSH has government funding and FaberAcoustic apps are expensive.
It has some really, really nice features (I often use it myself for quick tests). But, I would recommend against using it for anything that requires a high level of accuracy (like consulting, or legal issues, or project assignments for a graduate course in a graduate acoustics program).
How do you know the app is at fault? Couldn't it be the mic or the calibrator?
because I've tested the mic and the calibrator several times in my laboratory, and this video was posted at the same time as a couple of other videos demonstrating other iPhone apps that work properly and don't demonstrate the problems this app shows.
@@DanRussellPSU Thanks for the explanation, I think you should mention that kind of thing in the video. People coming in and watching this have no idea what other testing you've done or videos you've posted at the same time.
@@farleygibbs6320 -- fair point . . . I made this particular video (as one of several) while I was teaching my graduate level noise control course online and initially published it as "unlisted' only for my students to view. I had assigned them several hands-on projects that required them to obtain a sound level meter that would perform to a "Type 2" level of accuracy, and several students were asking about this particular app. Later, after the course ended, I republished it as "public in case it might be useful for other people. This particular video (from the set of three?) has become more popular than I expected. Had I made this video with an original intention of posting it publicly to my channel, I probably would have added more explanation of what I was doing, and why, and how this fit into the broader scheme of things. I thought I had put some of this explanation in the text description below the video -- but I realize that most people watch videos without reading the text description to get the context. Maybe I can expand the text to provide a more clear explanation.
Dan Russell
@Farley Gibbs - I've changed the text description to better explain why I made app and provided links to the other videos where I do the same for two other apps (NIOSH Sound Level Meter and FaberAcoustics SignalScope X).
I wouldn't really say how "inaccurate the app" is at high dB levels. I would put it more down to the limitations of the phone's physical microphone and not necessarily an app problem. If you think about it, phone mics are intended for voice and there wouldn't be much merit in a mobile manufacturer to install industrial grade microphones capable of measuring construction level noises.
Also notice how when you set your calibrator to 114 dB, the phone app is showing a consistent 111.1 dB level. Whereas when you set it to 94 dB, the app was jumping around between 94.7 and 94.8 dB. This makes me believe the phone's microphone maxed out at 111.1 dB and even if you had applied 130 dBs the app would've read 111.1 dB. Just a theory.
But I'm actually really surprised how good the app was out of the box with no calibration. It was within 2 dB accuracy. That's as good as many dedicated SPL meters 👌🏻
Thanks for the comment . . . but your assumptions are incorrect about a couple of points.
First -- for this video I was not using the internal phone microphone -- but was instead using a type 2 professional quality external microphone. So the limitations of the internal microphone are irrelevant for what I am showing in this video.
Second -- there are several professional quality sound level meter apps (Faberacoustical and NIOSH are two) that compensate for the internal electronics and which as a result do not have the flaws of this particular app. These other professional quality sound level meter apps correctly display the correct 94 and 114 dB levels regardless of which level is used to calibrate the mic and app. This app apparently does not
Thirdly -- this it is not a problem with the internal microphone being maxed out at 111.1 dB (and I wasn't using the internal microphone anyway). I have used this phone and external microphone to accurately measure levels well above 114 dB (corroborated with a standalone professional grade sound level meter) on several occasions. However, when I did those measurements, I was using a different app. This particular app is just not accurate, even after "calibration."
This particular app is useful if you want to get a quick idea of how loud a sound is. And it has some cool looking visual features. But, it is not a great app if any kind of accuracy is needed. The reason I recorded and posted this video was to answer questions from some of my students who were required to obtain an app and external microphone for several hands-on projects for a noise control course I was teaching during COVID. This app was popular because it was cheap, but I wanted to show my students that it has limited accuracy and they needed to be careful when using it for the projects I had assigned them.
@@DanRussellPSU I was too quick to comment after watching the video and realised what you're saying after downloading and testing the app myself. Appreciate the detailed response and might check out those 2 apps you mentioned! Cheers
Is this any application?
Some of the other apps I have used (NIOSH SLM and Faberacoustic SignalScope X) appear to be much more consistent across the range of levels and frequencies. This particular app looks really nice, and several of my students wanted to use it for class projects, so I made this video to check out how consistent it was.
@@DanRussellPSU hi, will u suggest this application for a research? since the accuracy is +-2db
@@tanyarawat6367 this particular app would be OK for general measurements and it has some rather nice user features. For undergraduate homework projects where the goal is to learn about sound levels, or to make general measurements about sound sources it would be fine. But I would not recommend it for research (especially not something for a graduate level project). For research or consulting---something that requires publication or a submitted report to a sponsor---you would want at least a Type 2 level of accuracy and I'm not confident that this app can provide that level of accuracy. It is a nice looking app, but I would never use it for my own research projects.
Zexi Li relying on this app for honking trucks inside her home above 10floors.
Anybody who uses an app for a reading probably isn't doing highly technical work and doesn't mind a reading somewhere within 5dB.