Hot answers tagged signal-processing
7
This is actually really straightforward given one caveat:
Your recording hardware and software needs to have a frequency response which includes the range you want (either very low or very high)
Almost all professional recording software will let you frequency shift - either directly, or by speeding up or slowing the playback of the waveform. Even free ...
6
You should use an expander when an outright gate will sound really unnatural (and that is not the desire). For example, if your vocalist has a really distracting breathing sound when idle, an expander will reduce that without completely eliminating the attendant ambiance that makes a track sound coherent (which is why it's important to record a minute or ...
6
To expand on Dr Mayhem's great answer:
It is not possible to hear a frequency that you cannot hear.
It is possible can transform a frequency that is too high or low to hear into one that you can, and you do that by shifting the recording's frequency as Dr Mayhem describes.
You won't be hearing the original frequency, but you'd be hearing a sound at the ...
5
You're right in that when your digital signal "clips" at 0dB, it's an absolute maximum limit. 16-bit audio has 2^16 possible volume levels, and once you get to 2^16, there's no higher number to express your sound, so it maxes out, or clips.
With analog signals, your "clipping" limit is less well defined. The term distortion in an amplifier just refers to ...
5
I assume you're talking about what I would call a lowpass filter sweep: a lowpass (hi-cut) filter starts out cutting out most of the signal, and gradually raises the cutoff frequency until the whole signal is heard.
I assume you're using some kind of DAW software like Reaper or Pro Tools or something like that. Generally, what you'll do is insert a lowpass ...
5
The "cocktail party problem" is actually a famous example in signal processing, and there are several different Blind signal separation or Source separation algorithms to solve it in different conditions. Are you doing this for homework in such a class? I've got an example of one on my website, but this statistical method (ICA) wouldn't work well for you ...
4
I haven't seen that movie, but one way to achieve this effect is with a binaural microphone. you would have to record and add breathing and swallowing noises later, as I am sure they did in the movie. Be warned, while binaural recordings work amazing well on headphones, they do not fair so well on speakers. This is why they tend to be used more for research ...
4
Key to this is the harmonics of the note. If you just play a 440Hz perfect sine wave it doesn't sound very exciting at all, but add in some of the harmonics and it starts to come alive.
If you add in the waveforms at 3x and 5x the fundamental, or at 2x the fundamental you will get a very different tonal effect. Varying the relative amplitude of these also ...
3
So, the fundamental difficulty to understand here is that each individual sound is made up of many many composite frequencies. When you play A440 on a guitar, it sounds different from A440 on a piano because the piano has vastly different upper frequency content. You said "the waves", but this is a little bit misleading: The waveform you see when you record ...
3
A couple possibilities:
Ring Modulation. For an example, see the sample audio for Mickey Delp's ring modular module.
Vocoding — use a noise source (or some other sound source of ambiguous pitch) for the carrier. For an example, see the track "Uranium" on Kraftwerk's album Radio-Activity.
2
I'm not 100% sure I understand your question, but here's what I think you need to do:
Add some compression. This will bring down the peaks closer to the average level. It will only affect the peaks, not the rest of the audio.
Normalize. This will raise the level of the the whole audio clip to the specified level.
The end result will be less variation ...
2
"...from the vocoder, where you can just play the keys and get the source tuned to the carrier frequency?"
That's not what a vocoder actually does. A vocoder analyses the rough frequency spectrum of a (source) signal, and uses this spectrum as a kind of equalizer envelope for the processing of another (carrier) signal. If the carrier had a very even ...
2
With analog equipment, if the signal level rises above the maximum input level of the equipment, it will begin to distort. The corresponding output level may or may not continue to increase, depending on the equipment. At some point, you will begin to damage the equipment.
Equipment that has Line-Level inputs will typically not accept levels much beyond ...
1
There are probably many different solutions to this problem. One thing you don't need to do is write more software. There is plenty of great software out there that will do the job. Hiring a local audio engineer who has experience with sound design and recording will probably save you a lot of time, money, and headache.
But if you want to do it yourself, or ...
1
There are two main concepts that I think are key here. The first is learning about signal paths and what the functions of each piece of equipment does along the way. The second is understanding the signals themselves and how they work. In video, this can be particularly important since many signals are actually the same on different cables. In both ...
1
It is impossible to repair clipped signal, since when clipping occurs, part of the original signal is eliminated and cannot be restored. The phenomena is described in the below image:
However there are few commercial clicks\pops removal apps available which could improve the signal. The following list of software appear in Wikipedia:
Sony Sound Forge
...
1
I did a little bit of searching and I was unable to find anything similar. That being said, if they're discontinuing a product, there's a good chance that a new version will be released. Failing that, your best bet is the RECEPTOR - I'd buy one myself if I had the money.
The problem with devices like these [And my guess as to why they're rare] is the number ...
1
I'm not sure you will be able to keep the same texture but eliminate pitch.
We recognize sounds as having a pitch when that pitch is stable enough to recognize. Taking your drum example, drum sounds often sweep down in pitch (kicks and toms) or are enveloped white noise (snares and high-hats). In either case, you don't have a constant, identifiable pitch. ...
1
(If this is too elementary please forgive me. Not meaning at all to talk down to you!)
Say you have a "noisy sine wave" on your oscilloscope. The main part of the sine wave is the low-frequency part, and the noise (the "fuzziness") on the sine wave is the high-frequency part.
The oscilloscope sweep would be your time domain.
Low-pass filters let the low ...
1
(I submitted this answer to the SO version of this question, I assume one or the other will be closed, not trying to pad rep or anything)
You might look into the LMS algorithm by Widrow and Hoff. You could apply it to each of your mixed recordings, using man.wav as the desired reference and woman.wav as the disturbance signal. There's also a way to use it ...
Only top voted, non community-wiki answers of a minimum length are eligible
