The best answers address the question directly, and back up facts with wikilinks and links to sources. Do not edit others' comments and do not give any medical or legal advice.
Sinc/Lanczos FFT bin interpolation on FFT-based log-frequency spectrum analyzers
I wonder if sinc/Lanczos interpolation is the best FFT bin interpolation method for "bandpower" spectrum (especially on lower frequencies part when using logarithmic frequency scale) because it approximates where is a discrete-time Fourier transform, in the best way when summation mode is set to "Sum" on my relevant CodePen project or is it? 2001:448A:3070:E3DA:7021:FEBA:971D:F9F6 (talk) 22:27, 3 June 2024 (UTC)[reply]
Isn't this more of a maths desk kind of thing? --Viennese Waltz 07:06, 4 June 2024 (UTC)[reply]
Not really. It is a question that audio engineers might be able to answer if they are knowledgeable about digital signal processing. In this context, there is no mathematical notion of how "good" a technique is. I know what FFT is, I know what interpolation is, but not what "FFT bin interpolation" is. You will not find the term "FFT bin" in a maths handbook. --Lambiam 14:46, 4 June 2024 (UTC)[reply]
@Lambiam What I meant by "FFT bin interpolation" is the same interpolation is done on FFT bins, which is necessary for logarithmic frequency spectrum analyzers since FFT have limited resolution on lower frequencies (since it has linear frequency resolution as opposed to frequency bands, which in this case has logarithmic frequency scale) and sinc interpolation closely approximates the zero-padding I believe when the interpolation is done before conversion to magnitude FFT. 2001:448A:3070:E3DA:E523:AA53:7234:600A (talk) 23:47, 4 June 2024 (UTC)[reply]
Yes. Wikipedia uses extended meanings of the useful term BIN for a partition or discrete interval in a range of values such as a Histogram bin, Data binning, a data pre-processing technique or Bin (computational geometry) a space partitioning data structure to enable fast region queries and nearest neighbor search. Philvoids (talk) 22:31, 5 June 2024 (UTC)[reply]
It's unlikely that a volunteer here will shepherd the OP in their coding project on another web site but I can give references and a worked example that will be useful.
The IEC standard 61672 for sound level meters gives much information on professional-standard audio spectral analysis.
A spectrum analysis in 1/3-octave steps implies using this bank of filters:
I distinguish two alternative approaches. 1) Bank of disparate filters, or 2) Single FFT with tailored bin allocations.
In either case the effort in the project depends on the chosen goals for resolution in power and frequency, and whether a near real-time spectrum display is required. (Performing the latter with high precision demands dedicated hardware such as a DSP or FPGA.)
1) Bank of disparate filters
The table defines 21 bandpass filters each of different widths, to be separately designed. The Butterworth bandpass design is optimised for flat response between its lower and upper half-power (-3dB) points. This means that other filter characteristics such as the out-of-band rolloff rates are neglected or "poorly" shaped. The best we can do is let adjacent filters overlap at their -3dB frequencies. Here is a worked example in Fortran to design a single Butterworth bandpass filter.
Change as required for each filter.
Band: 20 to 30 kHz
Sections: 5
Sampling interval: 10 usec
DIMENSION A(5),B(5),C(5),D(5),E(5),GRAF(2.20)
CALL BPDES(20000.,30000.,1.E-5,5,A,B,C,D,E,GRAF)
DO 1 N=1,5
1 WRITE(5,2) N,A(N),B(N),C(N),D(N),E(N)
2 FORMAT(5(13,5E14.6))
DO 3 N=1,20
DB=10.*ALOG10(GRAF(2,N))
3 WRITE(5,4) GRAF(1,N),GRAF(2,N),DB
4 FORMAT(1X,3E15.6)
STOP
END
C - BPDES
C - BANDPASS BUTTERWORTH DIGITAL FILTER DESIGN SUBROUTINE
C - INPUTS ARE PASSBAND (3-DB) FREQUENCIES F1 AND F2 IN HZ.
C - SAMPLING INTERVAL T IN SECONDS, AND
C - NUMBER NS OF FILTER SECTIONS.
C - OUTPUTS ARE NS SETS OF FILTER COEFFICIENTS, I.E.
C - A(K) THRU E(K) FOR K=1 THRU NS, AND
C- 20 PAIRS OF FREQUENCY AND POWER GAIN, I.E.
C - GRAF(1,K) aND GRAF(2,K) FOR K=1 THRU 20.
C - NOTE THAT A(K) THRU E(K) AS WELL AS GRAF(2,20) MUST BE
C - DIMENSIONED IN THE CALLING PROGRAM.
C -
C - THE DIGITAL FILTER HAS NS SECTIONS IN CASCADE. THE KTH
C - SECTION HAS THE TRANSFER FUNCTION
C -
C - A(K)+(Z**4-2*Z**2+1)
C - H(Z)=--------------------------------------
C - Z**4+B(K)*Z**3+C(K)*Z**2+D(K)*Z+E(K)
C -
C - THUS, IF F(M) and G(M) ARE THE INPUT AND OUTPUT OF THE
C - KTH SECTION AT TIME M*T, THEN
C -
C - G(M)=A(K)*(F(M)-2*F(M-2)+F(M-4))-B(K)*G(M-1)
C - -C(K)*G(M-2)-D(K)*G(M-3)-E(K)*G(M-4)
C -
SUBROUTINE BPDES(F1,F2,T,NS,A,B,C,D,E,GRAF)
DIMENSION A(1),B(1),C(1),D(1),E(1),GRAF(2,20)
PI=3.1415926536
W1=SIN(F1*PI*T)/COS(F1*PI*T)
W2=SIN(F2*PI*T)/COS(F2*PI*T)
WC=W2-W1
Q=WC*WC+2.*W1*W2
S=W1*W1*W2*W2
DO 150 K=1,NS
CS=COS(FLOAT(2*(K+NS)-1)*PI/FLOAT(4*NS))
P=-2.*WC*CS
R=P*W1*W2
X=1.+P+Q+R+S
A(K)=WC*WC/X
B(K)=(-4.-2.*P+2.*R+4.*S)/X
C(K)=(6.-2.*Q+6.*S)/X
D(K)=(-4.+2.*P-2.*R+4.*S)/X
150 E(K)=(1.-P+Q-R+S)/X
DO 160 J=1,2
DO 160 I=1,10
K=I*(2-J)+(21-I)*(J-1)
GRAF(2,K)=.01+.98*FLOAT(I-1)/9
X=(1./GRAF(2,K)-1.)**(1./FLOAT(4+NS))
SQ=SQRT(WC*WC*X*X+4.*W1*W2)
160 GRAF(1,K)=ABS(ATAN(.5*(WC*X+FLOAT(2*J-3)*SQ)))/(PI*T)
RETURN
END
Values of B(K) and D(K) are theoretically zero in this case but show tiny rounding errors.
The second WRITE statement lists 20 points on the power gain curve of the resulting filter.
This confirms -3dB gains at F1 and F2 and you can assess the overlap between adjacent filters.
FREQ (HZ) POWER GAIN POWER GAIN (DB)
10 points on lower :::::::: ::::::::
skirt including F1
10 points on upper
skirt including F2 :::::::: ::::::::
=====================================================
The reference for the above program is "Digital Signal Analysis" by Samuel D. Stearns, 1975 Hayden.
An updated version that includes an IBM floppy disc is "Digital Signal Analysis" by Samuel D. Stearns and Don R. Rush, 1990 Hayden.
2) Single FFT with tailored bin allocations
A possible FFT specification:
Sample rate: 32,768 samples/second
Inputs: 3,276 real, imaginary components zero
Outputs: 3,276 power squared (I² + Q²)
This implements 3,275 bandpass filters 10 times a second.
To obtain the overall level in any bandwidth, sum the squared levels of each FFT bin in the band, divide by the number of bins, and take the square root. Where the FFT bin -3dB points do not match the desired 1/3-octave steps, share the bin powers across borders as in this example:
FREQUENCIES in Hz
1/3-OCTAVE FILTER (example)
Band limits ADD THESE FFT BINS Weight
Center Lower Upper Center Lower Upper
125 112 110 105 115 30%
120 115 125 100%
130 125 135 100%
141 140 135 145 40%
More analysis filter resolution at low frequencies will need FFT process of longer sound streams than 0.1 second while analysis to higher audio frequencies will need a higher sample rate. Pursuing both these aims will increase demand on the FFT computation, give a slower result or need more expensive hardware. I do not think that alternative filter designs from the analog world offer a shortcut. Philvoids (talk) 23:25, 5 June 2024 (UTC)[reply]
June 5
The dead and France
Does anyone know if the country of France does and did (c. 1976) require incoming persons to have a passport, regardless of status as deceased? In reference to an event up for discussion at Ramesses II.
Temerarius (talk) 18:50, 5 June 2024 (UTC)[reply]
@Temerarius: There is some information at Agreement_on_the_Transfer_of_Corpses and Travel_document#Laissez-passers. There are two main treaties, from 1973 and 1937. France has signed the later one but Egypt signed the earlier one, so not sure which would apply in the case you refer to. Perhaps the mummy was not in a coffin that met the standard required by the treaty and so some other arrangement needed to be made. You might want to check for French newspaper reporting on the subject at the time it reportedly happened. RudolfRed (talk) 21:25, 5 June 2024 (UTC)[reply]
Good tip, know any good French search terms for it? Temerarius (talk) 01:40, 6 June 2024 (UTC)[reply]
And it is disputed in this France24 article]. To me it doesn't seem plausible that France (who wanted Ramesses to come) should have insisted on something like a passport. Queen Elizabeth II. famously did not have a passport, yet they let her into the country repeatedly. It seems obvious though, that Ramesses' visit was accompanied by a lot of paperwork, certainly including documents with identification, detailed description and presumably photographs of the mummy — the Egyptians should have insisted on that, after all they wanted their Ramesses back and not some random Jean-Jacques, and if possible undamaged, too. Call one of these documents a "passport" if you like, but the idea of a standard-format passport seems rather ludicrous — except maybe as a publicity stunt, but then I guess we would have found pictures of it. --Wrongfilter (talk) 08:33, 6 June 2024 (UTC)[reply]
The comparison to Queen Elizabeth II is particularly fitting given that, like her, Ramesses was a head of state. (What would they put in the "citizenship" field? I doubt Ramesses would be very happy at the suggestion that he was a citizen of a republic, and "He of Sedge and Bee" might cause confusion at border control.) Proteus(Talk) 12:23, 6 June 2024 (UTC)[reply]
The idea that the head of a state is not a citizen of said state seems ludicrous to me. Of course, the entire concept of royalty also seems ludicrous to me.--User:Khajidha (talk) (contributions) 12:27, 6 June 2024 (UTC)[reply]
Your UK passport was the Queen's letter that you may travel. So she didn't need one. Practical royal prerogatives like this don't extend to cadavers.
Lambiam (and any others who didn't click through), according to Talk:Ramesses II § A mummy with a passport (redux), the article added a claim about mummy passports in November 2006, so any "RS" statements that postdate the unsourced addition to the article – such as history.com 2013 and National Geographic 2018 – may very well be the result of WP:CITOGENESIS. We'll need a pre-2006 source to establish veracity. Folly Mox (talk) 12:19, 6 June 2024 (UTC)[reply]
In 1974, experts discovered that Ramses II's skin was being destroyed by a mysterious infection. The royal mummy was flown to Paris three years later for medical treatment. International regulations required him to have a passport, which gave his occupation as "King (deceased)". When he arrived in France, a team of conservators successfully cured the infection, which turned out to be a fungus...
Wow, in the dummy they used the same photo I attached the factoid to. Funny. In the link it looks like they were using "passporte" a bit informally to describe the documentation. Good work all! That's a fun investigation.
The spelling is passeport. Placed between scare guillemets, the term was obviously not meant to be taken literally. --Lambiam 09:46, 11 June 2024 (UTC)[reply]
How do I find out a date of death (UK) for person involved in an article I'm writing
I've been writing an article on a local history event in the 1970s and I'd like to check if some of the people involved are still alive. The UK online archives seem to be of the wrong date range or behind paywalls - what do people use on Wikipedia for looking up dates of death?
LicenceToCrenellate (talk) 17:37, 6 June 2024 (UTC)[reply]
Thank you! I'll dig out some details tomorrow and drop you a message on your talk page, if that's okay? I dont have a lot of information about them, but I'd like to give you as much as I can. LicenceToCrenellate (talk) 18:13, 6 June 2024 (UTC)[reply]
That's fine; if I find anything I will either edit the article directly or will let you know on your talk page. Hassocks5489 (Floreat Hova!) 18:33, 6 June 2024 (UTC)[reply]
If you know where the person is buried, you may be able to get the information from findagrave.com -- SGBailey (talk) 06:59, 13 June 2024 (UTC)[reply]
Re: Do the IBM Watson videos with the different robots still exist, or are they gone?
Over a decade ago, I once saw some videos that featured several fictional robots that were each interested in doing something that Watson could already do. In each video, one of the robots was interviewed. For example, one video showed a robot talking about how it wants to help improve air quality in Beijing, to which the interviewer responds that Beijing's smog problem is "a bit more complicated than that". In another video, another robot talks about how "Back in my day, vacuum tubes were" and I forget the rest of what it said, but I hope you get the idea. – MrPersonHumanGuy (talk) 15:36, 1 May 2024 (UTC)[reply]
Relisted to generate a more thorough discussion and clearer consensus. Relisting comment: I've seen this done with RfD discussions, so I don't see why I shouldn't try to do a similar thing with my unanswered Reference Desk question. (though I can see how this would be annoying if someone were to continuously repeat questions that have already received plenty of responses) Please add new comments below this notice. Thanks, ~~~~
If no one can answer my question at the moment, then I need to know that. Total silence isn't always a helpful way to get that point across.
My best guess is that nobody else scrolling through this page was familiar with the particular videos I was trying to describe. – MrPersonHumanGuy (talk) 21:20, 9 June 2024 (UTC)[reply]
It is, however, not more than a guess. I guess that nobody who saw this question knows for certain that nobody who saw this question can answer it, so no one of us can reply, "No one here can answer your question." One might reply, "Perhaps nobody seeing this question can answer it." This might be added to any question that has gone unanswered. A bot might even automatically add this to any question when it is posted, so we don't need to do this. --Lambiam 07:07, 10 June 2024 (UTC)[reply]
I thought the question was odd. I don't recall ever seeing these videos, and for the most part, I keep on top of some of this stuff (I take a few months off from popular culture and media every year for my own health, so I do end up missing some things). Was this video only seen in China? Viriditas (talk) 20:05, 10 June 2024 (UTC)[reply]
I'm not sure how any RefDesk editor can be expected to know that other editors don't know the answer to a question. Alansplodge (talk) 11:53, 12 June 2024 (UTC)[reply]
It could have been. Digital displays were common by then. I doubt you'll find a reference for such a trivial question. Shantavira|feed me 08:20, 10 June 2024 (UTC)[reply]
By 1980, digital watches were cheap enough that anyone could buy one. I got my first one in 1979. By 1983, they were so common and cheap that they began coming out with digital watches that also had games on them like PacMan. So, not only could he have had a digital watch, he could have had a digital watch with a soccer game built into it. 75.136.148.8 (talk) 15:11, 10 June 2024 (UTC)[reply]
Zooming in on some of the better images, it looks a bit like a Swatch, which was launched in the spring of that year, but it's hard to be certain. Alansplodge (talk) 18:15, 10 June 2024 (UTC)[reply]
It looks like an analog watch to me. If I were a referee I would not like a digital display without seconds pointer that rounds off to the nearest whole minute and I get too much real soccer to need more pretend soccer game.Philvoids (talk) 10:16, 12 June 2024 (UTC)[reply]
I agree it's analogue. It is much quicker and easier to read an analogue watch. It's difficult enough to keep an eye on the action at the same time as timekeeping. Shantavira|feed me 10:53, 12 June 2024 (UTC)[reply]
"It is much quicker and easier to read an analogue watch" Wouldn't that vary from person to person? --User:Khajidha (talk) (contributions) 18:49, 12 June 2024 (UTC)[reply]
Why do you assume that a digital watch wouldn't indicate seconds? I had a digital watch back then that could show hours, minutes, and seconds. And could also display the date or function as a stopwatch. (Not to mention that it could also transform into a toy robot.) --User:Khajidha (talk) (contributions) 18:48, 12 June 2024 (UTC)[reply]
I made no such categorical assumption and like not the provocative Straw man questioning. As a referee watching the conduct of 22 soccer players I would be merely irritated by a watch that keeps flashing ..53 54 55 56 57 58.. seconds, that threatens to turn into a toy robot or that shows phases of the Moon, whether you could buy such watches in 1983 or not. Philvoids (talk) 08:46, 13 June 2024 (UTC)[reply]
June 11
Monetary policy, attempt #2
OK, now that I've clarified where this question is coming from, here it is again: if the Federal Reserve was abolished, what effect would it have on the money supply? My educated guess as a non-expert is that it would greatly decrease the money supply -- is that correct? 2601:646:8082:BA0:E19D:4C11:C950:1DA8 (talk) 03:44, 11 June 2024 (UTC)[reply]
Money is not created by the Federal Reserve, which is a system, but by banks providing credit to customers. I assume that "abolishing the Federal Reserve" also removes the banking regulation and supervision assigned by law to the Federal Reserve, and thereby the capital requirement that is meant to ensure that banks do not overextend the credits granted and remain solvent, that is, can honour their obligations to depositors. As the brakes are removed, the immediate effect is most likely a huge influx of newly created, easy money. The FDIC will be unable to keep providing deposit insurance, which depends on the assumption that the banks are regulated. Some banks will fail spectacularly and take others with them, and as depositors see their deposits are not safe they will want to see cash – bank runs will ensue. As banks go serially belly up and the economy contracts to a depression that becomes world-wide, the money that was abundantly available will dry up. --Lambiam 05:50, 11 June 2024 (UTC)[reply]
The abolition of the Fed would only arise under extreme political circumstances that cannot be ignored in favor of a purely economic answer, except in some fantasy world. In that world, the author would decide what would make a good story. In the real world, the most likely outcome would be a combination of hoarding and skyrocketing inflation, arising from a comprehensive loss of faith in both institutions and the currency. DOR (ex-HK) (talk) 17:08, 11 June 2024 (UTC)[reply]
There are people who actually advocate the idea. Which is better, to ignore them, because they live in a fantasy world, or explain why this is maybe not the best idea, as long as the economy is still based on money? --Lambiam 19:22, 11 June 2024 (UTC)[reply]
Sorry for the late reply (I was busy with other political stuff which mainly had to do with defending Israel and denouncing Hamas sympathizers as the terrorists they are) -- yes, I'm asking this question because me and my followers want to formulate an argument against abolishing the Fed and in favor of keeping it, but we want to make an argument based on national security considerations rather than purely economic ones! (And for how the money supply affects national security, see Balance of trade and Autarky, first and foremost!) 2601:646:8082:BA0:180E:97CC:ADFB:D01C (talk) 08:45, 16 June 2024 (UTC)[reply]
June 12
Tarot PDFs
Hi, I'm looking for tarot decks by renowned illustrators to print. Hebrew letter inclusion a plus.
Temerarius (talk) 15:12, 12 June 2024 (UTC)[reply]
There are loads of images on Commons that are identified as being "Tarot". The image quality varies a lot, but it's a free option to start your search. Matt Deres (talk) 19:55, 12 June 2024 (UTC)[reply]
June 13
Enemies with wallrunning in single-player first-person shooters
I already saw the scrapped Pilot AI for Titanfall 2 single-player campaign many times and I wonder why that such thing were cut from TF|2 campaign and are there other first-person shooter single-player games where players have wallrunning ability and there are also enemies that have same abilities as you including wallrunning (much like the Evolved from Prototype 2 has similar set of abilities to the player)? And what is most difficult part for wallrunning in enemy AI to implement in a hypothetical singleplayer campaign of third Titanfall installment? 2001:448A:3070:E47D:A9C7:4A5A:80E5:29EF (talk) 11:14, 13 June 2024 (UTC)[reply]
Sign standards
ISO 7010 gives standards for signs. It includes prohibition signs (red border circles, most with a diagonal strikeout); Warning signs (red border triangles); Mandatory signs (blue circles); and emergency signs. Is there a standard for "Permissive signs"? (Eg the oppositite of "No dogs" being "Dogs are permitted" rather than "Dogs are required"? -- SGBailey (talk) 17:04, 13 June 2024 (UTC)[reply]
ISO 7010 does not explicitly use the term "permissive signs", it does standardize a category of green "safe condition" signs that convey permitted or allowed actions. These signs typically have a green background with a white symbol or text. The design principles for these permissive symbols are outlined in ISO 3864-3,[1][2] --136.54.106.120 (talk) 20:44, 13 June 2024 (UTC)[reply]
A green-background sign of a person showering does not mean "Showering permitted here", just like a sign EXIT does not mean one is granted permission to leave. These signs are meant to facilitate safety actions in case of emergency situations. Using signs that are similar in design to convey permitted or allowed actions not related to safety will dilute the intended signalling function of the ISO 7010 standard. --Lambiam 06:47, 14 June 2024 (UTC)[reply]
In free countries, at least, the lack of a warning or prohibiting sign would imply permission. ←Baseball BugsWhat's up, Doc? carrots→ 21:10, 13 June 2024 (UTC)[reply]
However there are some situations - buildings, locations - where a sign could contradict assumptions, saying something like "photography permitted", "firearms permitted on premises", "designated smoking area", or "nude beach". Just in case anybody was being too meek. Card Zero (talk) 22:06, 13 June 2024 (UTC)[reply]
In those cases, there is no lack of a sign. I'm reminded of an old story: A guy's looking for a parking space. He sees a bunch of parked cars. He asks a cop, "Can I park there?" The cop says, "No." The guy says, "What about those others?" The cop says, "They didn't ask." ←Baseball BugsWhat's up, Doc? carrots→ 22:49, 13 June 2024 (UTC)[reply]
ISO 7001 includes a sign for "smoking allowed". It doesn't seem to have one for "dogs allowed". PF 015: Smoking area or smoking allowedAlmostReadytoFly (talk) 14:07, 17 June 2024 (UTC)[reply]
Welcome. What is your question? RudolfRed (talk) 17:52, 15 June 2024 (UTC)[reply]
Yes, sorry I was not precise. If you look towards 12 o'clock in the picture (upwards) there seems to be something that looks like an electronic scoreboard, that's my impression. Or maybe it actually wasn't there and still needed to be installed? Because you can't really see much. Thank you.