McKinley Paul (asker) Mar 23, 2011:

Thank you again for all of your thoughts! I'm going to use "frequency peaks", as it seems most appropriate given the diagrams shown later in the text (with peaks at different perceived frequencies), but will let your vote of confidence be the deciding factor with regards to which answer gets chosen.

Zareh Darakjian Ph.D. Mar 23, 2011:

Not that fuzzy :) This reference makes clear that what's intended by "rais frequentailles" is "frequency components" or frequency peaks. This is clear by considerring diagrams 11, 12, 13 and of course applying our knowledge of Fourier transforms (time domain to frequency domain, in this case). My last words (for today:)

http://ww1.cnam.fr/maths//Didacticiels/Sur_Le_Traitement_Du_...

chris collister Mar 23, 2011:

Oh dear, so much fuzzy discussion. But for the record, Paul Dirac was not French, having been born in Bristol and then teaching at Cambridge.

Zareh Darakjian Ph.D. Mar 23, 2011:

FT In fact time-domain specrtrum and frequency domain spectum are Fourier Transforms of each other. It takes a whole set of time domain peaks to get one frequency domain peak (the latter is a linear combination of the former) and vice versa. So do you think that "the frequency ray" is a particular "frequency component" in the Fourier Transform? Is so, then the proper name is "frequency component", not "ray", in English.

McKinley Paul (asker) Mar 23, 2011:

I feel less certain about the term spectral lines because further on the author talks about "raies spectrales" but in a completely different context that doesn't seem at all related to the discussion about different types of dB measurements.

meirs Mar 22, 2011:

Fourier was French after all I found stuff on Fourier Transform and spectral lines - so I feel I am close. Dirac was French too - for those who remember his "delta" in the context of the transform from the T (time) to the S (spectrum) plane.

Tony M Mar 22, 2011:

If you look up the source term... ...on the Internet, you can get a pretty clear (albeit empirical!) idea of what it means.

meirs Mar 22, 2011:

I gave it my best shot If it's a line, it's spectral. ST is murky though. :-(

McKinley Paul (asker) Mar 22, 2011:

Hello again,

Thank you for the very thought-provoking discussion. I agree that the source text isn't the best... but it's all I've got to to work with. Is there any concensus at this point regarding a good term for the field of acoustics (dBSPL is given further on in the same section, so the two are differentiated by the authors) that captures the intented meaning (not necessarily word-for-word)?

With much thanks!
M

meirs Mar 22, 2011:

@Zareh OOPS - amplitude measured in which units ? numerator and denominator of the ratio must be expressed in the same units (normally power to be 10 times - or voltage or current to be 20 times - the logarithm of).

Zareh Darakjian Ph.D. Mar 22, 2011:

as previously suggested I thought we were translating the two words given to us by the asker. Of course, if dB is the question, then, in this case, as suggested above it's 10 times the logarithm of the ratio: amplitude / beam energy... This was suggested above, I would think... but that would not be an answer to the question asked. In my opinion (humble of course) we should go by what the asker wants to know...

meirs Mar 22, 2011:

@ Chris - I agree To quote William S. - "are we here to bury Caesar or..." in other words - do we translate verbatim (as far as we can) or provide technical writing services too. I don't have a clear answer.

chris collister Mar 22, 2011:

Regardless of what the ST actually says, it is attempting to define the dB. Hence any correct definition ought to do. See http://en.wikipedia.org/wiki/Decibel. It all depends on whether you want an accurate translation of the words, or an accurate translation of the meaning.

meirs Mar 22, 2011:

Frankly I wonder what the ST author wanted to achieve - explain the dB concept to whom and what for ? Not to mention what the term "énergie des raies fréquentielles" has to do (correlates with) the topic of the entire document and its intended audience.

Tony M Mar 22, 2011:

@ Meir LOL!

I don't agree with your reasoning: OK, no-one's suggesting we are talking here about actual SPLs — but that's no reason to think it might not be related in some way to sound levels (most likely, given the overall context) and that therefore it has to be e/m.

meirs Mar 22, 2011:

In context ST starts with "sans unité qui correspond à un rapport entre deux valeurs" so it cannot be about sound level - most likely about electromagnetic stuff. To Tony M - sometime I even think is terms of dB about money, like this costs 3 dB more than that, but that's a professional deformation.

chris collister Mar 22, 2011:

In fact, now that I think about it, "frequency lines" could make perfect sense if the text is talking about spectra or the frequency response of a piece of material or tissue.

chris collister Mar 22, 2011:

A dB is ALWAYS dimensionless since it is the log of something, in this case power (or energy) or amplitude squared. It is possible that this text was not written by the designer, though I suggest something along the lines of "dimensionless unit corresponding to the ratio of electrical energy to energy in the beam". "Fréquentielle" is a bit redundant here since if it wasn't alternating, it wouldn't be a beam.
In acoustics, although sound pressure per se is not dimensionless, it is referred to 20 microPa, which is taken to be 0 dB.
A "raie" can also refer to a spectral line, which may, or may not, make sense here.

Zareh Darakjian Ph.D. Mar 22, 2011:

amplitude in wave phenomena, amplitude squared is proportional to the energy. You question however, is not dB, or is it?

McKinley Paul (asker) Mar 22, 2011:

Okay, I'll have to take your word for that! They brought me onboard for the medical side of it, but I'm a bit in over my head where the physics are concerned. I really need to take a refresher!

Tony M Mar 22, 2011:

@ Paul Although your document may be about acoustics, it seems from the way they are defining 'dB' that they are perhaps viewing it in a wider context...

McKinley Paul (asker) Mar 22, 2011:

Thank you, but I am not sure that your clarification applies here, since we're dealing with the realm of acoustics rather than electronics. dB can apparently refer to a number of different things depending on the field that it's being used in.

meirs Mar 22, 2011:

Just to make sure dB is used in Electrical Engineering (electronics) to measure the ratio between two POWER levels (not "amplitude électrique"). It is defined as 10 x log (A/B) where A and B are the two power levels being compared and "log" is the 10-based logarithm. There is also a less used unit called Neper that is ln (normal logarithm) based. There is more to it - but it's too much for this forum.