Ah yes, there is a marked difference between what spot mikes pick up.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
"Things that cannot be measured"
- Thread starter ShiZo
- Start date
As I said, very different spectrum of the two instruments. They do not actually explain the differences in terms of bell shape and directivity, and the way they have recorded the sound with something like an almost monaural close-mic´ed arrangement (like documentary TV teams prefer it) does not even tell half of the story of how these instruments sound different in a room.
Not only that. The spot mike positioning has to take the nearfield directivity pattern of the instrument into account, as well as potential movement during the concert.
Particularly when doing a live downmix for a broadcast, the mixing engineer will compare the monitor preview of the spot mic track of the instrument and its position in the panorama of the main mics. Doing this a few times will give you a good idea how the instrument will sound in a room compared to nearfield. And these two are surprisingly different due to the bell/horn.
Well exactly, I appreciate the modern popularity and flexibility of using spot mikes and any difference in directivity of the instrument and the musicians particular movements will be particularly influential to what sound they pick up.
I have no experience with how to create a realistic impression of the sound in the far field from such inputs, my experience is old and recording in the far field.
It really depends on the instrument and the musician. Most orchestral musicians I know are primarily concerned with how the instrument sounds “out front” and that sometimes isn’t optimal for up close. One of the most valuable things I learned from my first teacher (I didn’t study with teachers until I was an adult) was the difference in his sound in the lessons room of his house versus from the mezzanine of the 3000-seat concert hall where they performed. The difference was startling—so smooth in the hall but a bit grainy and unsubtle up close. (He retired a few years ago from a major symphony, and was known for his orchestral sound.)
And my niece is a professional bassoonist. When my parents (her grandparents) decided to buy her a good bassoon as a graduation gift, the only requirement was that I be present for the transaction. So, I listened to her audition half a dozen Fox 601 bassoons. The one I liked sounded the most bassoon-like to me in the audition room. But she chose one with a noticeably buzzier sound, because it would sound more like the one I liked, but out front.
Musicians are not trained in mic placement but the recordist is often not familiar with the sound objectives of the musician. And musicians don’t always (often?) get a say. I’d much rather hear (just) a bit of the performance space than digitally added reverb. Most orchestral musicians I know are unhappy with their sound on typical recordings.
On the topic of red faces, I was at a concert of the Chicago Symphony back in the 80’s, at a tour stop in Austin. They played Mahler’s 5th, which, of course, opens with a monster trumpet solo. That solo isn’t technical and it is all about a big, orchestral sound. An oboist friend who attended with me remarked on the trumpet player’s red face, saying that player was killing himself and obviously had heart problems. Of course, he was (unknowingly) talking about Bud Herseth, who was the principal trumpet of the CSO for 53 years (and for 15 more years after this concert) and died in 2013 at age 91. I don’t put much stock in red faces, especially for instruments that use so little air as an oboe.
Rick “who got a clapped-out Toyota for graduation from college and had to buy his own tuba
” Denney
And my niece is a professional bassoonist. When my parents (her grandparents) decided to buy her a good bassoon as a graduation gift, the only requirement was that I be present for the transaction. So, I listened to her audition half a dozen Fox 601 bassoons. The one I liked sounded the most bassoon-like to me in the audition room. But she chose one with a noticeably buzzier sound, because it would sound more like the one I liked, but out front.
Musicians are not trained in mic placement but the recordist is often not familiar with the sound objectives of the musician. And musicians don’t always (often?) get a say. I’d much rather hear (just) a bit of the performance space than digitally added reverb. Most orchestral musicians I know are unhappy with their sound on typical recordings.
On the topic of red faces, I was at a concert of the Chicago Symphony back in the 80’s, at a tour stop in Austin. They played Mahler’s 5th, which, of course, opens with a monster trumpet solo. That solo isn’t technical and it is all about a big, orchestral sound. An oboist friend who attended with me remarked on the trumpet player’s red face, saying that player was killing himself and obviously had heart problems. Of course, he was (unknowingly) talking about Bud Herseth, who was the principal trumpet of the CSO for 53 years (and for 15 more years after this concert) and died in 2013 at age 91. I don’t put much stock in red faces, especially for instruments that use so little air as an oboe.
Rick “who got a clapped-out Toyota for graduation from college and had to buy his own tuba
” DenneyMaybe FFT as used here means Fast Fourier Transform. That is a mathematical tool that can take a complex waveform, such as the sound of a baby crying, and break it down into a series of sine waves of different frequencies, amplitudes and phase angles with respect to the fundamental tone. When these sine waves are summed together again, the original tone is recreated. The sequence may have a very large number of terms, but when just the first few are summed together, a fairly accurate rendition of the original tone will be heard.
Reminds me of an early time in the studio (1982) with 3 french horns. Student(s) back then, teacher let us point condenser mics to the opening of the horns and it was terrible. Teacher then went and pulled a door out of the frame in the lounge of the place and placed that, at an slight angle and horizontally - behind the french horns, pointing Sennheiser MD 441's at the angled, wooden door. Tadah...
Hard to get that taught these days, I might add.
Maybe he left out the leading "p".
More specifically, an FFT is a transform on a discrete (sampled) sequence of any integer length, but best done on power of 2 lengths (although multi-radix lengths work well, too, unless the length is a prime number). It is a subset of the Discrete Fourier Transform that allows calculation to be done proportionally to n * log2(n) operations, as opposed to n*n operations (matrix multiply) in the DFT.
It is the sampled version of the Fourier Integral, but by being sampled, and thereby of limited length, it is an exact transform not only in the amplitude space, but also the power space, unlike the Fourier Integral, which is accurate in power terms (zero error on reconstruction by the inverse transform) but that has zero-energy errors (small spikes of zero width, hence zero power) in the amplitude domain at discontinuities in the original signal. This "error" is not germane to any kind of real signal, since no real signal can have energy at infinite frequency, and thereby has no actual discontinuity in the mathematical sense. "ears" due to bandwidth limiting are not this phenomenon.
The DFT/FFT, not having infinite length (Thank you Gauss for figuring this out), has no such phenomenon.
An FFT has exactly the same number of complex output terms as it has complex input terms.
Norman Morrison's book on "Fourier Analysis" is a very good way to learn this stuff. It is, however, <expletive> expensive.
Somewhat confusingly, the usual FFT on a real signal has one real output at DC, one real output at FS/2, and (n-2)/2 positive frequency complex terms, which are echoed by as many complex conjugate negative frequency terms. So, again, n samples in, n samples out is the case, but with all but two of the terms in the transformed data being complex, except at DC and FS/2. It is possible to have an "odd frequency" FFT that simply has n/2 complex terms, and the resulting conjugate terms.
There are other transforms, like the DCT,that are real to real transforms. They have more interaction between adjacent terms, but are also 1:1 and onto, or tight frames, or orthonormal, using your choice of term. One typically uses whatever is most useful in the application. In general, many transforms can have an FFT-like structure for computation, but no, not all of them. For grins, look up the 'KLT'. Have your favorite headache meds handy if you're stale on your linear algebra.
As a separate note, it is unwise to say "just the first few" are summed together. If you mean "the most energetic" you would be closer to right, but ordering in an FFT output is by frequency, so summing the first few terms via IFFT is a very uncomfortable, not very useful, form of bandwidth limiting.
- Joined
- Dec 12, 2019
- Messages
- 16,419
- Likes
- 37,270
I ... I ... never knew what made an FFT a fast FT, nor the ramifications thereof.
Now I do.
Now I do.
Holmz
Major Contributor
Cooley and Tukey may be worth a Google search.I ... I ... never knew what made an FFT a fast FT, nor the ramifications thereof.
Now I do.
Then look at the "rolling radix" implementations after you digest the factorization.
Holmz
Major Contributor
- Joined
- Dec 12, 2019
- Messages
- 16,419
- Likes
- 37,270
it's nice to know how things work.
Holmz
Major Contributor
What ever you say.
In general yeah - but there are also people that just it to work.
Say you have a cop or fireman. They do not care about Marconi and they of ADC is the handheld radio.
They just want to call for help.
But in the spirit of thread - yeah know how it works is good.
There is much to learn (for instance as in how to change an even frequency to odd frequency FFT, or into a DCT, or DST, or even MDCT/MDST, or why a Hilbert Envelope is so useful and why it exists) in that study on factorizing phase in an FFT. Because it's all about the phase, about the phase, no trouble!
Sure, those people don't have to read this thread if they don't want to. Unless of course they are at the same time claiming there are audible things that can't be measured. Can't have it both ways.
Holmz
Major Contributor
Amps can and do sound different...
Yes, absolutely.
(When they're pushed past their power limits and/or are poorly designed.)
Similar threads
