I thought I’d experiment with recording live audio on my phone, using the free Dolby On app.
I was at a friend’s pub jazz gig last night and recorded the entire second of their two sets, just to see how well the app worked, really.
The result was pretty good except, because I was seated right at the front about 6ft from the drum kit, the bass drum has come out slightly too loud and occasionally a little distorted on the recording.
Any tips about correcting/mitigating this, from those in the know/sound engineering types, without losing too much of the other bass frequencies? Audacity is the sound app I use on my Windows laptop, but so far I’ve only ever used it’s basic functions.
Comments
Leave a Reply
You must be logged in to post a comment.

Oh blimey, this is tricky.
You might be able to reduce the level of the bass drum by identifying its frequency and using a notch filter to bring it down a bit – think a Graphic Equaliser. Audacity can probably do that kind of thing.
The distortion is going to be much harder to fix. The only way I can think is to capture a clean kick as a sample, if there is one, and use it to replace the distorted ones. Or use some library samples from elsewhere – how about an 808 drum machine, or John Bonham’s kick drum, or the Color Him Father break?
Tbh if you use a narrow (say, 45-65hz) notch filter and are able to get the kick dialled down low enough, the distortion on the kick’s fundamental will probably become less conspicuous, but you won’t get miracles.
Steve’s resampling suggesting might work if you were a) able to dial the kick’s fundamental down almost to nothing, and b) didn’t mind the very finicky and time consuming process of replacement. Some DAWs have drum replacement algorithms (Logic does, for eg) but they cost money.
Good advice Bob, I’ve spent (wasted) countless hours buggering around with the more adventurous parts of Audacity/Wavelab/Whatever only to find that the best thing I did before starting was to take a backup copy of the audio file and put it somewhere well away from my fiddling about.
@mike_h I’d be inclined to advise that your recording is basically toast – put it down to experimental exploration and buy yourself a proper field recorder – they don’t need to cost a fortune, they need not be much bigger than a phone, and they are WAAAAYY more capable, whatever app your phone has had installed.
I use one of these, a Zoom H2n, (about £125) and they are totally amazing:
Fully charged and loaded with a really fat SD card, you can record roughly 24 hours continuous music at CD standard. The width of the sound image is first class, and quite frankly the first time I used it to record a practice session in a large tent at a Bluegrass festival, everyone present thought the results were spectacularly good. It was sat equidistant from the guitarist, the fiddler and the drummer, but there was no overbearing part of the whole.
Mike – PM’d. Happy to help.
You’ve offered to do some of that black magic oojamawhatsit haven’t you? Couldn’t resist the challenge, eh? I hope you’ve got a good stash of coffee to keep you going, ‘cos it sounds like just the sort of voodoo-audio-malarky I’ve sworn never to attempt again!
Yes I like doing that sort of thing. It’s fun…
Thank you, sir.
WeTransferred a couple of minutes ago.
I have a good little digital recorder (Tascam DR-07) that I’ve used before with good results. I just wanted to try out the app. Results were obviously not as good as the Tascam but better than I expected, so having got a recording of some of this one-off performance, I wondered if it could be improved. The app has some pretty crude editing options but they’re not really useful IMO.
Where I was positioned (individually-allocated seating by a combination of how soon one books and pot luck) was extreme left facing the performance area (no stage) right at the front in front of the drum kit. To the (unamplified) drummer’s right were the double bassist behind the singer, then the 2 saxophonists and the keyboard player right over at the other side. My side’s PA speaker was also right in front of me and above.
My phone was on the table in front of me with the mics (bottom of the phone) pointed away from me across the front of the performance area, so as to try and avoid both excessive drumkit volume and audience chatter.
I suspect the bass drum was picked up through the table via the pub floor in addition to generally in the room.
Interesting experiment then! I’ve never thought of doing the same with my phone, mainly I suppose ‘cos the orific that is apparently the microphone is about 1mm in diameter!
It’s in hand. That’s one smoking band Mike, details please.
They’re not really a band, since they don’t play together regularly apart from the drummer, Rick Finlay, a veteran of West End musical pit orchestras and a drum/percussion teacher, and Alto/Soprano sax player Jeremy Shoham, Rick’s business and musical partner.
These two organise the twice-monthly gigs at The Elephant Inn, North Finchley and all the headliners/other musicians are chosen on a per-gig basis. According to who is available.
Keyboards on this occasion were by Liam Dunachie, who will be back for the next one on Dec 4th playing organ with headline guitarist Pete Oxley. Rick Finlay and Jeremy Shoham making up a bass-less quartet.
Double bassist on Sunday was Jeremy Brown.
Anita Wardell was the headlining vocalist for Sunday’s gig. You’ll hear her after the warm-up track “The Chicken”.
An American visitor, Tom Washatka played Tenor sax and his wife Janet Planet guested on vocals in this second set. They are old friends of Anita Wardell’s who happened to be in London on holiday.
I have a Janet Planet album! Not, I should addd, the ex-Mrs Van either. Haven’t listened for ages, but I think I got it on account of some or other cover she did.
Give us a moment…..
A bit supper club but at the better end of that genre.
I checked and there are 3 googleable Janet Planets.
American jazz singer (see above), singer/dancer in Australian dance band Confidence Man and the ex- Van partner.
Have you tried descaling the flange gaskets (obviously being careful not to touch the intraocular bi-wiring)? It’s a tricky process with a constant threat of displacing the alternate polaroids but as long as you keep your dimple tweezers sterile and you’ve checked your insurance policy covers Fire, Theft and Large Explosions you’ll be fine.
My friend, Roger, had the same problem a few weeks back. He’s as well as can be expected after a rookie error attaching the reverse anodyne to the backplate annunciator.
No explosion cover, sadly.
Just trying to reduce the bass drum to a less obtrusive level, so that the distortion is liveable-with. I don’t expect to end up with something pristine from an off-the-cuff phone recording. Next time (if there is a next time) I’ll try not just placing the phone on the table and pointing it’s mics across the front of the soundstage.
Thanks for the suggestions. I’ll try notch filtering as suggested and experiment with moving a narrow range of frequencies around, see what happens.
Your phone has mics, plural? Good grief.
*looks askance at elderly Samsung device*
Have you seen how many cameras an iPhone has got? More than I’ve got eyes. And I live on the Fens.
A question for you, Mr Fenton.
As a certain amount of audio information is lost when a .m4a file is created, if one of these is unpacked for editing in Audacity or suchlike, will further information be lost if it’s then re-compressed to .m4a at the same bitrate afterwards?
I’ve often wondered that very same thing. Tis one of life’s enduring mysteries. This and the “How many guitars does a chap really need?” thread has been a source of delight, and dare I say,bewilderment.
Someone give Lodes a whack on the knee
Ah well, it depends. Are you sitting comfortably?
m4a files can be lossless, but are more usually 128, 192 or 256kbps. The m4a compression scheme is the same as mp3 but, unlike mp3, is multi-band, so is much more efficient than mp3 for a given bit rate. A 128k m4a file sounds as good as a 192k mp3 file. 256k m4a sounds better than 320k mp3.
If the encoder was any good, all the data discarded in the first encoding would be it, and any number of decoding/encoding cycles would give the same result. Sadly, it is less than perfect.
The general consensus is six cycles of encoding/decoding at 128k mp3 results in an unlistenable file.
m4a at 128k would survive more than six cycles. m4a at 256k would be many more than six.
Are you still awake? I am very dull.
What happens to the discarded data? I have a Dyson, will it just end up in there with the crumbs and pubes?
Do they go into some kind of black account, ready to be cashed in later, like the 0.5c in that (Eddie Murphy?) film? (Trading Places?)
Dull? There’s something stirring in my underpants. Seems like a very small woodland creature has awoken.