Recording up above 24kHz

SMuir

SMuir

Well-known member
Has anyone had any experience with extended frequency response microphones?

A recent effects project has had me recording monster vocals at 96kHz using a Neumann TLM170, and directly adding them to a 48kHz ProTools session (ie. playing them back at half speed). The results have been quite good, with little of that lo-fi pitch-shifted sound.

However, I'm getting greedy. I'd like to preserve even more of the high-frequency harmonics during the pitching-down process. To that end, I was wondering if anyone could reccomend a mic with a frequency response above 30kHz or so.

Thanks a bunch, folks.
 
I wonder what the purpose of recording that high could be. Not only can we not hear up that high, but there are very few speaker systems in consumers homes that can play that high. My current system will play up to 36kHz and my former system would hit 45kHz but they cost more than what only a fraction of a percent of the population could even afford if even if they were so inclined to purchase such a system.

Don't get me wrong, I am all for capturing the highest quality possible, but I don't know how you would distribute it and play it back. 192kHz sampling rates are not even really that good at such high frequencies since they don't sample enough spots to generate a reasonable approximation of the actual waveform at those frequencies.
 
Presumably the point in this case is that the audio will be slowed down...
so a 88.2khz recording played at half the speed effectively becomes a 44.khz recording, whereas if you slowed down a normal 44khz recording by half the maximum frequency that could be reproduced is 11khz.

I only know of the Sennheiser MKH800 and the sanken Co-100k, but I haven't used either of them.
 
Many, many tests have shown that people rally can not hear anything above 20kHz and for us seniors 16kHz is an achievement , thus trying to make systems to reach higher is a waste of time and money. 44.1 kHz is perfectly fine sampling rate. BUT, for slowing down there is a point in using higher recording sample rates.

I just saw a mic manufacturer's web page showing a mic reaching near 100kHz, a measurent mic turned into hi-fi/effects capture instrument. I forget the company, but it was one of the big ones (Shoeps, Sanken, DPA???)
 
While slowing down sounds are great at higher sample rates (one of the reasons the Synclaier sounds so good is that it can sample at 100kHz and thus is very smooth when you slow it down) there is not that much point in recording up there. You are likley to start getting buzzes and hums that you are then going to lower into the range of hearing. Very few FX are going to have much content up there and you may find that what you get is not what you want. So I would probably try out a really high test mic or something and record a few things before putting out money. I suspect you will start to hums and buzzes on a lot of recordings. We did some recordings at the Skywalker Foley stage and when we slowed them down we started hearing the transformer from the video monitor. It was isolated but the very high feq, stuff was getting through. Apparently nobody ever pitched anything way down so they had never heard it before. Anyway this is long winded but be-careful about what you ask for...

ANd please report back.
Good Luck
 
Good point about the monitor hum, Noiz2. We recently swapped out a couple of the monitors at the Footsteps Post foley stage, replacing them with new LCD screens. We auditioned a number of models, and found that most LCDs either hum or chirp after an hour of use. We settled on a Sony model, but discovered that they whine around 16 kHz unless you set the picture mode to "Vivid." Once we set them to vivid, the whine vanished. We ran a number of tests recording the TLM170 at 192kHz and pitching it down, and found that not much undiscovered noise showed up. Of course, we haven't checked with any high-frequency measurement gear, but we seem to be in good shape.
 
Last edited:
The only time I've ever really worried about recording stuff above 20K is if I know I'm going to pitch shift it down. Recording bats is a prime example. There is a lot going on up there that we don't hear that you CAN hear once you slow it down enough.

Along these same lines, regarding what we can and cannot hear....

Back in college, in my sound engineering class, we were talking about timbre, and overtones, and we had overtone charts of different musical instruments. My prof did some recordings with some very high quality mics (of course, I can't remember) of various instruments (flute, clarinet, trumpet, bassoon....). He would then take a recording of one of the intstruments, and then gradually filter out overtones that we supposedly can't hear. Well, the odd thing was that as he startetd doing this, the timbre of the instruments started to change, even before he got down to 20 or even 16K. I don't know how that worked, but it was weird.

-Dave
 
Filters tend to introduce phase shifts and amplitude anomalies below (or above for hi-pass) the cut-off frequency. That could be the explanation.

Or the conventional scientific knowlege is wrong.
 
Dave_Fisk said:
Well, the odd thing was that as he startetd doing this, the timbre of the instruments started to change, even before he got down to 20 or even 16K. I don't know how that worked, but it was weird.
Click to expand...

Even if we can't hear them, the other frequencies can "hear" and react to them (beat and all that).

That's why synthetic musical instruments never sound right. They're not producing those unhearable sounds.
 
HI Dave (Like myself I see you on ever other audio post forum ;-) Neve did demo's with a specially designed filter with this. IRRC his explanation was that the usual sine test (sweep a sinewave till you can hear it, stop osc and be shocked to find that your hearing has been rendered void beyond 12k ;-) was incomplete. Now if instead of a sine wave you take something with harmonics like for instance a saw wave, then applied his filter you would indeed hear differences. But then, this is just my recollection of something i read on the net somewhere. I'll ask one of my teachers at school who is totally hi-end audio engineer (http://www.grimmaudio.com) what he thinks. I kinda respect his thoughts on these matters. Ill report back :)
 
Dave_Fisk said:
Back in college, in my sound engineering class, we were talking about timbre, and overtones, and we had overtone charts of different musical instruments. My prof did some recordings with some very high quality mics (of course, I can't remember) of various instruments (flute, clarinet, trumpet, bassoon....). He would then take a recording of one of the intstruments, and then gradually filter out overtones that we supposedly can't hear. Well, the odd thing was that as he startetd doing this, the timbre of the instruments started to change, even before he got down to 20 or even 16K. I don't know how that worked, but it was weird.
Click to expand...

It probably was a junky filter with a very exaggerated rolloff. Even the best filters and eq's don't roll off perfectly accurate. If you did a low pass filter set at 20khz the rolloff curve will suck in some frequencies well below 20khz. Especially if it was dramatically attentuated. It also depends on the quality of the conversion being used.

Authenticity of real vs synth instruments has very little to do with the absence of high frequencies. If that were the case why could you tell the difference between a real instrument over a synth while listening to a monitor that only plays back 40hz-18khz? It's more about the complexity of the midrange and the positive errors injected by the person into the performance. It's like saying, "I wish I could record at 5hz, that I way I could really capture that low end in those chirping birds that we all miss.".

Now recording at higher sampling rates than 44.1khz does help in certain cases. Anytime you could capture more events within a period of time the better when you start to apply changes. Yet those changes would still be well within our hearing range. It's like shooting at a higher frame rate than 24fps. You don't do that to shoot something that's not there. You do it to get an effect within our range of vision. It's like saying, "If you shoot at 60 frames per second, you'll see a glowing aura around Burt Reynolds. That's just not possible with 24fps. He was in smokey and the bandit man!"
 
You must log in or register to reply here.
Back
Top