Freesound Forums

Hey there!
With my more professional field recording kit I made my first serious (yet experimental) recordings of a metal door.
I uploaded the sounds into the following sample pack: http://freesound.iua.upf.edu/packsViewSingle.php?id=3405

I'd like to have some comments/critics on its production about what I could have done better. So far I have identified on my own:
- I could have used a little more gain (there's still plenty of headroom left)
- The footsteps could have been avoidable; But since they only happen between the 'critical' events, this is a minor point.

I'm especially interested if there were any other/better mic positions -- or are these situation dependant and not bound to rules of thumb?

I would definitely try using a cardiod near coincident or coincident mic setup. This will help give more directness to the sound you are going after. In the recordings you posted you have a lot of the "ROOM" sound in the recording. Getting less room in your sound will make the recording a lot more useful for yourself and others. For example if you needed to add this sound to a film or something and it took place in a very different space such as a small room or outdoors the reverb from your recording would sound very out of place. The drier the better (for the most part) is a good rule of thumb to follow. Remember with a dry sound you can always add reverb to match a space, but you can never remove it.

If you are stuck with Omni's i would just try and move the mics much closer. This will skew the balance of the direct to reverberant sound in your favor.

Maybe try again and post the results to see how they turn out. Good Luck and happy recording!

Also in regards to your comment on having 'plenty of headroom left', I don't think I would identify that as a problem. You should have plenty of headroom in every recording. Optimal recording levels generally are about -12 to -18 dB on peaks. That way, when you mix it with a lot of other sounds (say in a soundtrack) you don't have heavy attenuation on all your channels.

i am gonna have to disagree with tweeterdj. The leaving 12-18 db of headroom will increase the noise floor when you bring it up to maximum in a mix. Most of the time attenuating the signal lower will not cause a problem but adding gain to a signal that was recorded with too much headroom just raises the noise floor along with the signal, therefore a noisy signal. If you record with very little headroom and you would like a quieter sound in the mix all you have to do is attenuate the signal, with no problems.

Thank you for your replies so far.

In sum, the more of the original sound is recorded, the better. This also counts for gain and for any effect that the surroundings add (reverb...). Interestingly, I considered the gain, but not the room

Then again, in some way I did want to test the recording of the room (which is also why I made a stereo recording); I bought the two omnis for ambience (room) recordings, so in this respect they performed quite well (I'd hope so - for their price ) but if it were for the door alone (the essence of the whole sound), I understand your point.

The only alternative I currently have is an Audio-Technica 835b, which (in comparison to the Audix) is quite noisy - or I yet haven't used it properly.

ejfortin
i am gonna have to disagree with tweeterdj. The leaving 12-18 db of headroom will increase the noise floor when you bring it up to maximum in a mix. Most of the time attenuating the signal lower will not cause a problem but adding gain to a signal that was recorded with too much headroom just raises the noise floor along with the signal, therefore a noisy signal. If you record with very little headroom and you would like a quieter sound in the mix all you have to do is attenuate the signal, with no problems.

please google 'gain structure'. that is all.

I tend to agree that having a louder recording at the start will be more beneficial.

You don't want to push your gear too loudly as even before clipping etc your equipment won't be working as well but apart from that crank it up a bit.

And yea, put the Omni's close to the source. They don't suffer the same bass increasing effects as other mics.

Mike

tweeterdj - can you please show me your reference? I have never heard of such a thing. Please post your reference here.

I'm definitely with ejforin on this one. 12-18 dB is too much headroom. You don't really need to leave any more than 3 dB. When I'm going through a sound library, the first thing I look for is recording that is loud. I want the noise floor to be as low as possible. This is because all of the noise from all of the different recordings compound together as I drop more an more sounds on top of each other. By the time I have all the sounds I need stacked, there's a ton of rumble underneath everything.

Figuring out a proper gain structure is only difficult in situations where you have a long and/or complex signal chain, like in a large studio with lots of outboard gear or a big PA rig. Also, the entire point of having proper gain structure is to keep the signal to noise ratio as high as possible (i.e. the noise floor as low as possible)...so, no offense, but I really don't see the point you're trying to make, tweeter. Like ejforin said, the more headroom you leave, the higher your noise floor is gonna be when you crank the levels later, and the worse the recording is going to sound. You pretty much always want to have the mic running as hot as possible when you're recording in order to get the cleanest signal.

By the way, I did google "gain structure" just to make sure I wasn't full of crap (it happens lol), but I'm pretty sure I'm correct in what I'm posting.

Cheers,
Will

thanks for the backup on this thetruwu. Lots of people on this site are still learning the craft and it's important that those people aren't led astray in regards to the audio theory discussed on the forums. Yeah, not really any facts behind tweeterdj's reasoning.

no prob, ej.

Just to clarify, I'm not trying to come off as a prick, but way quiet recordings are one of my biggest pet peeves.

-Will

Alright guys, did you look at the sample pack? He recorded them at 96K, 24 bit. Therefore he doesn't have to worry about having hot levels to maximize his s/n ratio or his bit depth. In fact, I've found several places on the web that say recording at 24 bit you should leave 12-18 dB of headroom. And 12-18 dB is not "way quiet" either, not when the dynamic range at 24 bit is 144 dB.

At any rate, this belongs in a different thread, rather than hijacking an existing one.

Tweeterdj your reasoning isn't sound. If you feel thetruwu and me are incorrect then please send us your "several places on the web". PLEASE PROVIDE LINKS. You have to understand, what you are saying goes against tried and true audio theory.

I'd like to second that pet peeve! The quieter the recording the less you can do with it in the end. It really limits the functionality of a sound.

I don't want to be the guy to say moderation is everything but it pretty much applies to this thread.

You don't have to worry about low sound levels with 24bit to the extent that you did when people tended to work at 16bit.

You also need to consider that your pre-amps and converters won't perform as well in the red (pre- clipping).

However you should record at a fairly high volume to keep the signal to noise ratio low.

Mike

Tweeter, I don't care if you're recording at 16/44 or 24/96, you try to get the levels within 5 dB of 0. This is what they teach in school, this is what they do in studios. Maybe field recording is different, but I really doubt it.

I'm sorry to come off like this, but I don't want people who are just starting out to pick up bad habits, *especially* since I may be using sounds that they record in films that I work on.

Anyways, I'll keep an open mind about this if you or anyone else can find some references that say otherwise. I've looked, but I haven't found any.

-W

In particular for field stuff, I feel comfortable at about 5-6db of headroom at the loudest point if possible. (Though many outdoor environments dont even come up to that)
I try leave about -5 or -6db in the studio recorder as well, but I'm much more careful about it there and its hard to 'soundcheck' certain things in nature, but a band can be 'encouraged' to do so.

thetruwu
Tweeter, I don't care if you're recording at 16/44 or 24/96, you try to get the levels within 5 dB of 0. This is what they teach in school, this is what they do in studios. Maybe field recording is different, but I really doubt it.

I'm sorry to come off like this, but I don't want people who are just starting out to pick up bad habits, *especially* since I may be using sounds that they record in films that I work on.

Anyways, I'll keep an open mind about this if you or anyone else can find some references that say otherwise. I've looked, but I haven't found any.

-W

Matey, I think you should start listening to what people are telling you from experience and stop being drawn into the pigeon holes that college teaches you.

This is becoming an increasingly boring forum post because you don't understand what you are talking about.

If you work at 24bit you have more resolution than you need because it has more dynamic range than human hearing. Even recording at lower volumes than Tweet is talking about you still have WAY more dynamic range than CD quality.

It's important to maximise levels at 16bit because otherwise you can find yourself working with a recording that represents 12bits for example. You had to keep volumes high to keep audio fidelity.

Learn what you are talking about before you are critical of other people.

Mike

mike T3 - Again your reasoning doesn't work for me as well as other professionals that are a part of freesound. Recording hot levels is just a basic practice that is used throughout the industry. If I turned over a bunch of low level recordings as part of a sound design project I would not be in the business for long. Low level signals are not accepted by industry professionals. It is taught in colleges and other audio institutions because it is a good practice. Getting in the habit of recording low level signals is a bad idea, and should be avoided.

I took a look at your site and most of the music I saw seemed to be sample based. I am sure the commercial libraries you are using wouldn't be successful if they had noisy / low level samples recorded for their libraries. Also, you say "stop being drawn into the pigeon holes that college teaches you". Did you attend a college audio program??? It says on your site you attended "Alchemea, College of Audio Engineering". I am sure there they didn't tell you it is a good idea to record at levels peaking at -12 and -18. There is no way you can tell me this is a good idea. Oh, and I AM TELLING YOUR FROM EXPERIENCE recording levels this low is bad. End of story.

Read this forum post it's good.

http://forum.cakewalk.com/tm.asp?m=480244&mpage=1&key=How%2CMuch%2CHeadroom%2CMain%2CBuses&#480244

I didn't say 'don't go to college' or 'I didn't go to college' I said don't be drawn into a set rules that college (or anyone else) teaches you. You have to set rules in order to teach people, but don't live and breath by them. ie I MUST RECORD ALL MY AUDIO AS LOUDLY AS POSSIBLE. It doesn't work like that.

My reasoning isn't flawed. Consider what you know about the difference between 24bit and 16bit to be.

Your posts are really frustrating. At what point did I say noisy samples were acceptable?

Stop being defensive and listen. Please don't be so incredibly arrogant to write 'End of story' at the end of your posts.

Mike

All I am going to say is that in the professional audio world low level recordings are not acceptable.