Is there such a beast?
I just did a spectrum analysis of two of my reference tracks, and they both had a frequency cut off at about 17khz. I don't mean attenuation, I mean a vertical line. How is that done? I certainly can't emulate it with EQ.
Comments
audiokid, post: 379163 wrote: I don't use one but a multiband co
audiokid, post: 379163 wrote: I don't use one but a multiband compressor/limiter or side chaining might do what you are asking ? Vertical line... hmm looking forward to others chiming in.
Well that's the probably nearer to what I want, but a multiband compressor centres on one specific frequency - I'm looking for a shelf, i.e. "everything over 17khz"
I see this often these days. Many folks are having tracks record
I see this often these days. Many folks are having tracks recorded by other folks at others' Studios & home studios. And he keeps stuff flowing easily across the Internet, folks are utilizing 128 kb per second, 44.1 kHz sampling, 16-bit MP3's. At 120 kb per second, these MP3's all cut off dramatically sharp i.e. brick wall filtered at 15 kHz. This means very little makes it beyond 15 kHz. For many situations, this is quite all right as nothing is transmitted for television or radio over the air beyond 15 kHz. In fact they all use brickwall filters at 15 kHz because they have to. And one of the most utilized of all recording microphones is the venerable SHURE SM57/58 whose response drops off dramatically after 17 kHz. Regardless of that, it is virtually the staple microphone of the industry for all amplified guitar amplifiers, most everything on a drum kit, most every vocal, most everything. So don't be mixing with your eyeballs but when your ears. You might want to complain to God since we were not given flat response hearing. Our hearing is less sensitive in high and low frequencies at lower volume levels then at higher volume levels. Our hearing response varies with loudness levels. And that's also one of the reasons why some people have a tendency to sing or perform on the flat side of notes when their headphones are too loud. Then pitch is perceived at a higher frequency and the results is a performance consistently under the pitch. So understanding the propagation of sound and the equipment that also reproduces it is of paramount importance in being able to make quality sounding recordings to begin with. And many of us filter our mixes of certain instruments & singers to reduce overall bandwidth such as high pass and low pass filtering. This is done to restrict certain instruments & vocalists to sit in a certain area of the listenable spectrum. Not everything is supposed to be 20-20,000 Hz. Only the equipment is supposed to be that way that we are recording upon. And our microphones are the largest palette of colors that we have to work with. It ain't the preamps snort equalizers that generally gives a recording its remarkable color. It's the selection of the microphones utilized that has more to do with your recording than any of the input equipment in which you are plugging it into.
Just to reiterate, MP3 at 128 kb per second features a 15 kHz cut off. At every incremental increase to 160, 190, 224, 256, up to 320 kb per second high frequency response is extended incrementally. So at 320 kb per second, you have full response out to 20 kHz where the normal Nyquist brick wall filter is always utilized at 20 kHz. 22 kHz for 48 kHz sampling and why some people feel that it is better sounding at 48 kHz than at 44.1 kHz sampling. But then the math doesn't work out properly when trying to downconvert to 44.1 kHz sampling from 48 kHz sampling. It's a little like getting a PI in the eye at 22 divided by seven which has no known outcome.
Now I have also been working on simple ways to recover from that 15 kHz cutoff. Essentially, most everything between 15 kHz and beyond is mostly made up of harmonic content rather than fundamental content. Meaning that this is mostly more of a loss of high frequency harmonic content. If one has the software in their multitrack software package, one may choose to high pass filter from 1 kHz. This will cut off everything below 1 kHz. Then you go into that section that allows you to change pitch without changing tempo. You then pitch up a complete octave. This creates a semi-distorted alien chipmunk track. You render this out as a separate file. Then you place it next to your original file which is undamaged, unmolested in any way. You then add in some of the chipmunk track either in phase or with inverted phase. One will improve the high-frequency response the other will kill it. You are effectively adding second harmonic distortion of all frequencies above 1 kHz. Of course the Nyquist filter still cuts everything off at 20 kHz for 44.1 kHz sampling. But you have added back in some of the high-frequency content lost from the 15 kHz cut off of 128 kb per second MP3's and this also includes low bit rate versions of other file formats as well such as .WMA, etc.. This may bring back some of the lost sparkle or it could make things sound like termites ate your mix?
Try anything that works
Mx. Remy Ann David
audiokid, post: 379196 wrote: hehe, I figured it had to be somet
audiokid, post: 379196 wrote: hehe, I figured it had to be something different than the obvious. How does this work for a LPF brick wall limiter?
Just that they have really steep low pass filters in converters to remove all frequencies down to the Nyquist frequency. Mp3 encoders do the same thing for a somewhat different reason, data reduction.
Here's what the spectrum of an mp3 looks like:
I don't use one but a multiband compressor/limiter or side chain
I don't use one but a multiband compressor/limiter or side chaining might do what you are asking ? Vertical line... hmm looking forward to others chiming in.