Sometimes one engineer does all recording and mixing. However, it is the producer's job to make sure everything is optimized in an overall coherent way. That starts with the song, arrangement, instruments, mics, recording medium, and so on down the line. If anything is done poorly at any one stage, it often can't be fixed later.
Theoretically, yes. For an infinitely thin string. For real strings, the effective length of the string is often shorter for higher harmonics.
Last edited:
Hee! Don't worry Bonsai, you were just thinking about music instead of test tones! Nothing wrong with that.
- Jim
Wow! Analog FIR (sort of)! I never heard of that trick before. Mark, do you know any examples of commercially available recordings where this was done?
- Jim
45 degrees off axis is pretty severe though, esp as the 1976 WW article was referring to on axis dips in response. But intrigued me to revisit some of the Zaphaudio designs. Interestingly some of those are designed with off-axis listening in mind* and across the board they show how flat you can get off-axis if you think about it and design wisely.
*fully aware that those versed in the art will scoff at why I wasn't aware of the advantages of optimising for 15degrees off axis.
Thanks Jan that's what I remember seeing. Take the 48k file for instance bin 3 and 4 imter-modulate into bin 7 (f1 + f2) and is masked because there is an input tone there so I just have a different understanding of exactly what is trying to be achieved. This is not true for the 44.1 file (for bin 3 and bin 4). My understanding was the Excel numbers were generated by some algorithm and you picked the closest match for your FFT size and sampling rate. If you allow unlimited orders of distortion for all 31 frequencies there simply are not enough unique bins.
EDIT - on some thought an algorithm that guaranteed >some< bins were always empty would be useful ( I guess?).
Last edited:
If you have high enough resolution FFT (I usually use 128K or 256K) and a suitable window pretty much everything has its own bin I would think, making the issue of finding optimum frequencies less important. I have used as high as 16M but you go out for lunch while its taking a measurement and then frequency stability becomes a real issue.
Hi Scott, I miss them also. It's not the mics. They sounded like that live!
Those film studio choruses, the Robert Shaw Chorale, Roger Wagner, all had one thing in common - the singers were really good and even if they were not classically trained, probably all could sing classical repertoire.
Robert Shaw, when asked how he got his wonderful results, down played his own talents somewhat when he said something like, 'I find the best singers I can, and let them sing.'
He and those directors of that time all let their artists sing with their natural vibrato and formant and the blend is that gorgeous sound.
I can say this confidently because I sang in a very small amateur choir in the 1960's under a talented American conductor who thought Shaw was the bee's knees and in a very short period we started producing the same sort of sound. (Not as good, though).
Here ya go Shaw, 1948. Noise and all, still sounds good.
https://www.youtube.com/watch?v=siYf8YIlY3Q
1956, this is exquisite
https://www.youtube.com/watch?v=Oc22Vfzl1Kg
Living Stereo
https://www.youtube.com/watch?v=Mbf5CuabIlY&list=PLD02C2991130D9839&index=30
I just stumbled over this and you get it because Doris is a doll and in 1954 I think my wiring got connected right listening to her. The back up is Norman Luboff singers.
https://www.youtube.com/watch?v=Lc5xknlOl3c
Did someone already point out that with FIR filter eq you can manipulate the amplitude response and the phase response completely separate. You can have say a 100dB/oct roll off yet flat phase response throughout that.
Jan
I was indeed under the impression that IM tones would fall in empty bins, that that was the whole deal with the special algorithm. I never actually looked at it the way you did (which is yet more proof that you are much smarter than I am of course ;-).
I do remember that there was a limit to the number of harmonics this would show, depending on FFT size and such.
I have a folder of more than a dozen with this type of files, if anyone is interested.
Jan
Re frequencies matching FFT bin centre frequencies, there is a linked Excel file at: Software tools>Getting the numbers right>1st post.
George
George
Quite so, but it was in the context of the brainblurbs on harmonics nice to say it is easy as 1-2-3, in all my humble perplexity.
Yes, I know what you mean.
They are selling strong chandeliers here as well so you can swing from them with one arm whilst scratching your armpit with the other.
After that, you go out and buy 2 litres of warm beer and the Sun for a relaxing morning sitting on the doorstep.
😀
Last edited:
- Status
- Not open for further replies.