Recording loud sounds

[vidmanners]

First, I want to record some loud street-drummers . . . I've heard them and they really hurt! - I tried to record them using a PalmTrack on "low"-sensitivity and No1 (of 32 levels) - but don't know what the settings mean . . . and even that overloaded. I moved about 30 feet away and that stopped the overload (a red LED), but lost some brilliance in the snares..... so I'm guessing the dynamic range needs to be covered with two mics . . . at different sensitivities, but I'm concerned that my mics could be destroyed!

The Palmtrack uses electret capsules (like most Camcorders) and these appear to overload rather easily, although they are remarkably good music quality. They survived the ordeal.

I have A Karaoke mic set and these are significantly less sensitive (moving coil, I'm sure). They are in metal tubes with XLR connectors, so I'm guessing quite reasonable quality. I tried to measure the sensitivity by plugging into the Palmtrack - it needs more gain, for a given sound-level, so I suspect this means I can record without overload, although the said mixer will be able to drop-levels the issue still remains at the microphone end - I would like this NOT to distort, e.g. like a rattle, or whatever....

I'll give that a try, but I wonder if folk here have any Hints/Tips for such audio...?
The PalmTrack is stereo, so a small mixer can bring the two mics to one stereo input. Then I can monitor it via headphones. The Palmtrack has level-meters and agc, something like a Zoom.
But I may use Manual-level for some and AGC for the others....a bit of trial and error,

Now the second recording-session is more difficult - the party finishes with Fireworks . . . there will be the usual rockets and flash-bangs, with music, but I understand the finale will be large aerial displays with explosions that fill the sky - As I recall similar events, these are LOUD, so I'm hoping my karaoke mics might do this job, recorded separately from the video, and try to fudge the sync later. To catch the Oohs/Aahs from the crowd, I can use two sensitivities for the L+R tracks.

I'll make rain-hoods to protect the mics outdoors, and dead-cats against Wind.... with something similar for the PalmTrack . . . . any other suggestions? . . . . pref. without spending any serious money.
Oh and carry a torch.
---and earplugs!
Again . . . hints and Tips, please.

[Rembrandt Rob]

Okay, you have two issues here. We'll start with the issue about damaging the mics.

If the microphone is dynamic, in other words it doesn't need an external power supply, it should be robust. (If it has "made in china" on the side, then sell it on ebay and buy a decent one like a Sure sm57 or sm58 or any of the Sennheiser range) Providing that you don't set off a firework within a yard of it or actually rest it on the drum skin, you won't damage a decent dynamic. On cheaper mics you will get a buzz or rattle in loud environments, so it's not worth the risk. Condenser mics have a more delicate diaphram but you've still got to go some to damage them (unless it is a real cheapy). Basically, if it doesn't hurt your ears, it won't hurt a microphone. (Of course dropping it, blowing into it and getting it wet doesn't count).

Your suggestion to have two settings and mix in post is the way I would also go but don't rely on the AGC. Let one be too quiet and the other overload on the bangs. Ideally dynamics for the drums and bangs and a condenser for the crowd.

If you have a high pass (or low cut) switch in on. Drums have a load of noise in the lower frequencies which we can't really hear but they eat up audio headroom.

[Rembrandt Rob]

Oh yeah, what I forgot to add...

Record in 24 bit and reduce to 16 bit in the edit. This gives you much more headroom for those loud sounds. On a digital recorder don't forget that you don't have any tape hiss, so you can record at levels a lot lower than if it were analogue. So you can have the meters (if there are any) just about moving on the crowd sounds and increase the levels in post production without adding hiss or distortion, if you have a decent microphone. That way you utilise the 24bit headroom for the unexpected loud sounds.

[vidmanners]

I've been using my PalmTrack in 24-bit as most of my "takes" are short and it has a 96-file limit (not mentioned in the Instructions) - recording onto a 4G SDHC card. This has LCD meters and will do manual OR manual with agc (which I tend to use.). It was this recorder that showed "overload" even when 30ft away - it was seriously loud. But not enough to damage hearing (the drummers used no earplugs!).
I've noticed the Karaoke mics are significantly less sensitive cw my other m/c mics, although these haven't been used with the Palmtrack; I'm not due to record them again for a while.

My background is in electronics and yr comments about bits isn't quite how I understand it. To get the best reproduction you should invoke as many "bits" as possible, hence the Hi-Lo switch on my Palmtrack and the built-in attenuator/amplifier . . . by having a low-level recording you are in danger of reducing the fidelity, since increasing the volume doesn't create more data . . . I suppose it closest to "pixilation" when you Enlarge (amplify). Of course a low-setting does give greater headroom, if your objective is to record the loud parts, whereas turning up the gain during quiet parts means it's going to overload later-on, hence the idea of a 2-channel/different gains although this only works in "manual" - as AGC works against you.

+ Thanks for yr comments re the Euro-mics - I may to get one. However, I've read good vibes (SoS magazine)on the Prodipe budget m/c mic - quite a bargain at £35, but the cable arrangement is far from clear.

[Rembrandt Rob]

No, you haven't quite got the difference between 16 bit and 24 bit right. You're confusing bits with samples.

The samples are the number of times the sound is digitized per second. This is what affects the quality the most. Video tends to record 48,000 samples per second (CDs have 44,100 samples). The bits are in effect the loudness. So, with 16 bit you have about 66,000 levels (can't remember the exact number) between silent and full volume. With 24 bit you have about 16 million levels. So, with 24 bit you can record a larger range of levels than with 16 bit however...

Since video usually uses 16 bit sound, you have to convert 24 down to 16 bit but most software can do it without distorting.

Recording at low levels with a decent mic is the way to do it digitally. The worst thing which can happen to digital audio is that it's recorded too loud. Unlike analogue, once you've got 100%, then there's nothing you can do to recover the information. Recording at a low level is not so much of a problem since, as I explained, in a digital recording there is far less background noise and, as such, it will take amplification well.
Providing that you have a decent microphone.

Edit: I don't know the Prodipe range but if SoS says they're good... I've read the review and it does look like excellent value for money (although they're French but, I suppose, at some point we can overlook that!) and they look like a copy of the Sennheiser 825 which is a superb mic for its relatively low cost.

[vidmanners]

Thanks both of you.
+ I had a moderate success (yesterday) recording those loud sounds (Drummers), although it was in a "soft" arena (so no relections, the sound was somewhat quieter - but I took the precaution of wearing closed-headphones with good surround-pads.) . . . The microphone came out of the odds-box + it appears to have recorded without overload, although the "quality" I'll leave to those with a studio-set-up and credentials. (So I'm making them a short CD).

paulears:
No sure I understand the "pad" description - do you know if this is a simple resistive divider, or perhaps something more exotic?
For it seems to me that if the condenser/electret is exposed to excessive movement (loudness), then cutting down the electrical output is not going to solve the resulting non-linearity issues (er, or clips). So if these are "pro" mikes then I suspect the capsule is not hitting its peak - the only danger is the i/p stage might - - - hence the divider.
I've been looking at Electrets (which require a small battery supply), since these are very small (typ. 10mm Dia). yet appear to have a near-flat response 20-20kHz (according to Mfr data). However, many suggest they are "difficult" on high spl. This being the case, an attenuator is not going to solve the problem, but I'm comforted by the thought that the capsule front-end (diaphragm) isn't the problem.
Mixer:
Soon, I shall be making my own mini-mixer and maybe phantom; I will need a wrecked phantom-powered mic to experiment with. The other issue is providing the nominal 48v- not too easy from a NiMH 3.6-0-3.6v PSU, but watch this space!

Those pads for Phantom . . . the problem is that they effectively "divide" the phantom supply - so a capacitively coupled pad will eliminate this, allowing the DC to pass, whilst reducing the sound signal. I'm sure you know this, so thanks for the warning. However, it probably better to have a built-in pad, as anything "external" is subject to breakage/loss/etc. However, IMHO the best place to drop in an attenuator is at the far-end of the cable, so any microphonic noises (caused by wire-insulation rubbing) will be proportionately less.

Rembrant Bob . . . thanks - Prodipe are indeed Fr, but the SoS article was nearer to Editorial, not a review using the thing. Prodipe claim the Chinese dynamic-inserts are regarded as exceptionally good, although others have pointed out the spec is +-10db - which says quite a lot. However, the available rare-earth magnets in China means this is likely the explanation - in the same way in-ear phones and loudspeakers have improved by leaps and bounds.

[vidmanners]

Thanks for that, Paulears, - - - now I've had a chance to play the CD on some modest gear, without the PC present and it's not bad, but somehow lacks "Umph"*. The Band-leader's vocals come through, boosted a little by me.
Being there, I know how tummy-loud it was and that doesn't come over on my CD - I guess the mic behaved itself there being a few surprise transients present and not one waveform was clipped.
Also the playback set-up is "modest" so no massive headroom...

I suspect my m/c mic is just too cheap to have a decent LF response - perhaps I might try a passive bass-boost, if there is a next time.
Looking at the waveforms in SoundForge (free with Vegas Studio Production Suite 10), it needed much work on the levels, esp at the start; but later I got cunning, shifting the gain higher when they started (tapping their sticks on the drum-cases.). With PalmTrack's LCD level indicators I could see the recording in progress. (so why is this Model discontinued?)
Perhaps I do need a decent mic for recording Drums. However, for my vids I'm more likley to need mid-freq instruments.
With these Drums, Palmtrack's internal Electrets overload at all settings, except zero (that's Off), why would you want a zero-setting? I've recorded a few blanks thinking it was just quiet.

Also some attention to mic-positioning would help, along with a song-list in advance.

I note in ads. nearly every mic is suitable for drums, . er, and everything else. Also I don't get the drum Pick-up (sorry can't recall Term), -this is a detector that activates the PC when the drum is struck . . . why doesn't the waveform do that? Can the Drum pick-up be that much faster that it matters? Also why isn't the recording running anyway . . . unless this is a start-signal - why not lose a few seconds, running?
When I'm recording I'm more concerned about the battery, than running low on memory.

*Of course I don't have a commercial CD of drumming to compare, that would provide a bench-mark. However, being realistic, my mic is rubbish at LF.

[vidmanners]

Expect to go (Nov.5th) to a couple of "Fireworks Displays" and thought I'd take the PalmTrack...Huh. (One on 6th).

Any ideas for recording severely loud bangs, from fireworks? Some of the "mortars" are really loud, do I put a karaoke mic in a tupperware and play about in SoundForge?

I suspect the PalmTrack is just not going to do the job with its electrets, so I won't try...
((This thread was started with a "Loud-sound" issue caused by a loud band of Drummers. That was resolved using an insensitive m/c mic and the Palmtrack on LOW setting #1 ))
+ Fairly sure there were no overloads, but it didn't reproduce well - probably something to do with the playback speakers being well below solid-bass. Not too bad in my AKGs with closed backs - as worn at the Venue as defendgers.... but I looked very "pro"

Another idea is to make a "mute" in reverse, ie an outer tube with acoustic foam/carpet in which I burry the mic....I believe the "whole tube" will have no reflections, so it should not "colour" the sound, but any sound entering will be attenuated by the time it gets to the mic . . . just wondering as I have some 6" (150mm) Dia rigid plastic tuning doing very little else..

Maybe I should buy some Bangers - for experiment, beforehand?

[Rembrandt Rob]

Don't forget that with a decent microphone and recording digitally you can always increase the volume in the post production without distortion. If you get the recording wrong and the signal is overloaded, you are stuffed. So, it's always best to record at a level which errs on the side of safety by being a bit low.
Fireworks and pyrotechnics give off a shock wave and lots of low frequency waves which you feel, as much as hear, which is why they are so impressive "live". These are very difficult to record on standard equipment and virtually impossible for the average television to reproduce. Thes e low frequency sounds also eat up recording bits, meaning that you lose details in the high frequencies which are more important for reproducing fidelity. So, in the recording, it's better to sacrifice the low frequencies than the high ones. These low frequencies are in the range of wind and handling noise, so if you leave the low frequencies in the recording, you will also get a lot more handling noise and wind sounds.

If the PalmTrack has attenuation I would switch it on, wrap it in a little bit of foam to cut out wind noise and record ambient sounds with that. Be aware that foam, if it's too thick (and carpet) will reduce the high frequency sounds a lot. Not good!

You are far better recording "clean" sound at the venue and then adjusting it in post to get the "boom" effect.

[TimStannard]

Thes e low frequency sounds also eat up recording bits, meaning that you lose details in the high frequencies which are more important for reproducing fidelity. So, in the recording, it's better to sacrifice the low frequencies than the high ones.

Just had an "ah-hah" moment, Rob, because of this useful comment.

I'd wondered why mics often have a LF cut switch. I had previously assumed it was simply to reduce wind and/or handling noise (which, of course it will do to some extent) and I had always ignored the switch on the basis that I'd prefer to capture as much as I could and roll-off the lows in post, where necessary.

Now I realise my misunderstanding. Thanks.