Talk:RIAA equalization

An article already existed at this page, created in September 2003 (by me). What happened to it? Where has all the history gone? Graham 02:29, 8 February 2006 (UTC)

All of this "new way" stuff is basically an advertisement someone shoved into this article - Maybe it should be trimmed? --69.193.129.39

The t.c. electronic Gold Channel pre-amp manual may very well have lots of useful information in it, but it is ultimately a product manual and the link doesn't really belong here, does it?Ciotog 18:19, 6 January 2007 (UTC)

Misleading claims about Stereo records

I've removed a totally misleading section that claimed that some stereo records in Europe (i.e. Decca FFSS) and America (i.e. Columbia Stereo) were not cut using the RIAA curves. I've searching the internet for any claim that could support this, and the only ones come from this product:

AMR PH-77 phono equalizer FAQ http://www.amr-audio.co.uk/html/faq_ph.html

There is no bona fide technical document that supports the claims (of stereo records not being cut conforming to RIAA) that are shown there. The other "source" of this claim is just a comment on DiscoGS:

http://www.discogs.com/Paganini-Michael-Rabin-Lovro-Von-Matacic-Philharmonia-Orchestra-The-Concert-Voor-Viool-En-Orkest-No-/release/2214757

"This LP is pressed with Decca (FFSS) EQ curve; in order to hear the full glory of this LP, it is best to use Decca EQ curve on the photo amplifier to replay it. If one uses the standard RIAA EQ curve to replay it, the body of the violin will be missing and will sound and edgy. "

I have years of experience with the LP medium (and over 3500 posts in vinylengine.com, mainly regarding technical considerations of LP playback), and have never ever heard about a Decca FFSS (stereo) curve. Known curves are here:

http://www.vinylengine.com/cartridge_database_record_equalization.php

The only non-RIAA curve used on stereo records is the TELDEC/DIN curve. I will add a paragraph regarding it soon. — Preceding unsigned comment added by Flavio81 (talk • contribs) 20:28, 18 April 2011 (UTC)

The RIAA was a poor source for info. about the curve

Some years ago, I wanted to see the actual RIAA document that describes the characteristic. I went to the RIAA Web site, and was astonished to find that they (no longer) had such a document! The only info. they could provide was a National Semiconductor application note (for electronic circuit designers) showing how to use an IC op. amp. in a phono playback circuit that included RIAA equalization. While the reliability of that information was decent, it was definitely derivative, and the lack was surely no credit to the RIAA. Nikevich 06:44, 1 March 2012 (UTC)

Stress on Stylus

About "This also reduces physical stresses on the stylus which might otherwise cause distortion or groove damage during playback.": I'm suspicious of the source for this. Actually, there is no stress on the playback stylus, aside from normal long-term wear. There is a "pinch effect", better known as tracing distortion, that is made worse with increased amplitudes of higher frequencies. The RIAA curve actually makes this WORSE, not better. (With care during cutting, RIAA makes the high frequencies better overcome surface noise, without significant added distortion, and therefore is an overall benefit to the sound quality.) So please find the source and fix this. Ohgddfp (talk) 17:38, 13 September 2012 (UTC)

Interpretation of recording curve

As you requested, I will disagree with you here, on the article's talk page, rather than your or my page. With all due respect, I disagree, Jeh. Please go to "Disc Recording Equalization Demystified" by Gary Galo, and take a long look at Figure 7, and then let's talk again. (Google search produces his article as the first hit if you use that search string.)

One thing that's not made clear in the figure is that the sine wave in the upper section is represented as amplitude, while the characteristic curve in the lower section is velocity. Since the [magnetic] pickup is velocity-sensitive and the curve also depicts velocity, it's appropriate to apply it to the amplitude (voltage) from the pickup. That's the playback curve we're all familiar with, but in terms of amplitude, the recording characteristic boosts bass and cuts treble, not the other way around. It just doesn't boost(cut) as much as the pickup cuts(boosts).

500Hz to 2122Hz is constant-velocity i.e. 6dB/octave amplitude droop. Above and below, it's constant-amplitude, except below 50Hz where it's constant-velocity again.

In addition, take a look at http://www.steveespinola.com/lathetrolls/resources/RIAA-curve_for_disc_recording.pdf . (Said to be from the Westrex 1700 electronics manual.) The first words are "The recording characteristic is defined as the radial velocity of the modulated groove...", followed by the graph you see everywhere. But this graph is velocity vs frequency, not amplitude vs frequency. There's also a table of dB vs frequency. Everywhere the frequency doubles or halves, the dB change is less than 6. Therefore, when you superimpose the magnetic playback pickup's inherent -6dB/octave characteristic, you get a composite curve that falls with frequency instead of rising. That's the amplitude version of the Recording Characteristic; bass is recorded at higher amplitude than treble. — Preceding unsigned comment added by Vintage Dave (talk • contribs) 17:15, 28 April 2014 (UTC)

This change by you is wrong. You wrote that bass frequencies are increased and high frequencies attenuated during the disc cutting process.

In the book Handbook for Sound Engineers: The New Audio Cyclopedia by Glen Ballou, a large chapter is devoted to "Disk Recording and Playback", written by George Alexandrovich. He says that if disk recording is attempted without equalization, then the "low-frequency excursions would take all the space. The high frequencies would be of such a low amplitude that during the playback, high frequency signals could be very close to the noise level of the system. ...[I]n the early days... the cutting head sensitivity was decreased at low frequencies so that the amplitudes in midrange and at high frequencies could be recorded at higher levels... [T]he playback amplifiers were adjusted to boost the low frequencies to compensate for the losses introduced in recording."

So you see that bass frequencies are decreased during the disc cutting operation. The RIAA curve also increases the amplitude of high frequencies as the disc is cut. Please stop trying to guess for yourself how it works. Binksternet (talk) 18:09, 28 April 2014 (UTC)

Bass velocity is decreased (and treble increased), true, but bass *amplitude* is increased, and treble amplitude decreased.

Did you look at the Recording Characteristic chart? Every octave change enumerated there lists a change less than 6dB. The Recording Characteristic is tailored for a flat system response when the pickup is a magnetic type. All magnetic pickups rise 6dB per octave. Therefore, to disentangle the pickup from the Characteristic, you have to multiply the Characteristic graph by a curve consisting of a straight line at 6dB drop per octave. You can also apply this on a spot basis to the chart - just subtract 6dB from every octave difference. Every place you do this, the net change per octave is negative, not positive, which means that as far as groove excursion (amplitude) is concerned, the Recording Characteristic slopes down to the right, not up to the right.

If the entire frequency range was recorded constant-velocity, then the high frequencies would disappear in the noise and the low frequencies would take up too much groove space, as you say. That is why the recording curve goes constant-amplitude above 2122 and below 500. (The CV region below 50 is a late addition to mask rumble, which became a problem when platter speed was reduced from 78 to 45 and 33, and system bass response improved in the hi-fi era.) Since the two CA regions are separated by a CV region (500-2122), their respective amplitudes will be different. But since the CV region equates to falling amplitude, the highs are smaller groove excursion than the lows. The magnetic pickup emphasizes the highs more than the recording curve attenuated them, which is why the playback curve slopes down: to compensate for the pickup AND the recording.

Everything written about velocity-responding cutters became obsolete when the Feedback cutter was invented. All cutting systems since then have translated amplitude at the input jack to amplitude at the groove. Nevertheless, the recording curve was - and is - specified in terms of velocity, while the playback curve describes amplitude. Taken that way, the magnetic pickup's velocity response is tacitly included. But if you take recording from an amplitude point of view, you have to unwind the pickup response out of the equation, and the resulting recording amplitude curve slopes down. I'm not guessing, it's just math. — Preceding unsigned comment added by Vintage Dave (talk • contribs) 18:40, 28 April 2014 (UTC)

I found another reference. See http://frank.pocnet.net/other/AWV_Radiotronics/Radiotronics_1965/1965_08_AWV_Radiotronics_30_08.pdf . This is "Radiotronics" magazine, volume 30 number 8, August 1965. See article "Inside The Ceramic Cartridge", on pp142-149, or pp2-9 in the pdf. (Never mind the title; the part we are interested in is about the recording side, not the playback side.) The paragraph headed "Equalization" begins with "Recordings are made today using the RIAA recording characteristic shown in curve A of Figure 6, wherein the vertical coordinate represents stylus velocity." Velocity! It goes on, "If now the same curve were plotted in terms of stylus displacement, then the same equalization takes the form shown in curve B." Curve A is the familiar upward-sloping curve, and curve B slopes downward.

Here, I did the math for you. The first dB column in the table below is taken directly from the RIAA Recording Characteristic document. The sentence before the table is "The following chart states the R.I.A.A. recording characteristic:". I don't know how much more unambiguous I can get.

The second dB column is my contribution, dB in velocity translated to dB in amplitude by taking into account the magnetic pickup 6dB/octave slope.

You can get a 6dB/octave slope (centered at 1kHz) with

   y = 20 log(f/1000)

with f in Hz and y in dB.

30Hz worked out as an example:

log(30/1000) = -1.523 * 20 = -30.46
-18.61 - (-30.46) = +11.85

f	dB (velocity)	dB (amplitude)
30	-18.61		+11.85
50	-16.96		+9.06
70	-15.31		+7.79
100	-13.11		+6.89
200	-8.22		+5.76
300	-5.53		+4.92
400	-3.81		+4.15
700	-1.23		+1.87
1k	0		0
2k	+2.61		-3.41
3k	+4.76		-4.78
4k	+6.64		-5.40
5k	+8.23		-5.75
6k	+9.62		-5.94
7k	+10.85		-6.05
8k	+11.91		-6.15
9k	+12.88		-6.21
10k	+13.75		-6.25
11k	+14.55		-6.27
12k	+15.28		-6.30
13k	+15.95		-6.33
14k	+16.64		-6.28
15k	+17.17		-6.35

Notice how it flattens out above 2kHz? That's the high-frequency constant-amplitude region. Increasing velocity, but constant amplitude. (The 50Hz-500Hz CA region is harder to make out, because it's between two inflection points.)

If you can point me to a free copy of the Alexandrovich article, I'd like to read it, but I think I've made my case. Thanks for spurring me on to the extra research.

<edit> Sorry, I don't know how to make tables in Wikipedia, and I have to go to a meeting RIGHT NOW. Three columns: Hz, dB (velocity), dB (amplitude).

Velocity is not the target; amplitude is. The whole point of the curve is to reduce bass amplitude during cutting, and increase it during playback. Also to increase treble amplitude during cutting, and decrease it at playback. Velocity is a means to that end, part of the electronic design of the circuitry. There is no chart, no table, no list of numbers you can bring to this page to convince me of your position. Your position is directly contradicted by the standard audio engineering texts. Binksternet (talk) 21:24, 28 April 2014 (UTC)

The following are replies to Vintage Dave (talk · contribs), not to Binksternet...

Re tables: Well, one thing you can easily do on a talk page is precede each line with a space. Then it comes out looking like code, and it obeys your line breaks and even honors tabs. I've taken the liberty of reformatting your table above in that way, and I think I fixed your calculation for correct display also. Yes, tables take some learning, but simple ones aren't too hard. You might try reading this for starters. In the meantime, could you at least learn to sign your talk page posts? (That's a link, please click it.) It's not difficult. Anyway...

You claim: One thing that's not made clear in the figure is that the sine wave in the upper section is represented as amplitude, while the characteristic curve in the lower section is velocity.

That's odd. I always thought it is simply what you get if you check the freq response of an RIAA eq network. i.e. level vs. frequency over the entire curve. Looks like that here, in fact.

You claim: The Recording Characteristic is tailored for a flat system response when the pickup is a magnetic type. All magnetic pickups rise 6dB per octave.

You are missing a fundamental point: The cutting head is also an electromagnetic transducer, but operating in the opposite mode, so its output falls by 6dB per octave. This effect is therefore nullified in the total path from cutter head input to pickup output!! So claims that the recording curve is the inverse of the playback curve PLUS compensation for the rising-with-frequency output of the playback pickup are bogus. The recording curve need only be the inverse of the playback curve. The playback curve slopes down from left to right, as you say; this can be verified by anyone with a turntable and a non-RIAA-equalized high-gain input; the playback will have too little bass and too much treble. The recording curve therefore slopes up from left to right, with the result that the bass groove excursions are lower than they would be if the recording curve had not been used.

We can debate this for weeks here, but what matters on Wikipedia are reliable sources. The page by Gary Galo looks to me like a self-published source, or close to that. I see no way to interpret the statements in the mainstream sources that have been cited here and in the article as meaning anything other than, e.g., "the bass groove excursions are lower than they would have been had the recording curve not been used." Furthermore, all RS's state that that (reducing the groove excursions) is one of the primary reasons for the recording curve's existence, the other being to boost treble levels up above the groove noise. I imagine that both Binksternet and I could come up with a dozen sources that say the above, in almost those words. And neither of those goals would be met by a recording curve that slopes down—regardless of what's going to happen at playback time. So you're going to have to come up with a better source. Several better sources, in fact. Jeh (talk) 21:51, 28 April 2014 (UTC)

Dave here. Thanks for the formatting.

You (Jeh) said,

"You are missing a fundamental point: The cutting head is also an electromagnetic transducer, but operating in the opposite mode,
so its output falls by 6dB per octave. This effect is therefore nullified in the total path from cutter head input to pickup output!!"

I am not missing any point.

If you drive a velocity cutter with constant-amplitude input, you get flat velocity, i.e., 6dB/octave falling amplitude.

If you drive it with 6dB/octave rising amplitude, you get 6dB/octave rising velocity i.e. flat groove amplitude.

But listen to this: the RIAA curve rises at less than 6dB/octave.

Therefore, if you drive a velocity cutter with RIAA amplitude, you get... falling groove amplitude!

Which of these points are you denying?

Velocity cutters were made obsolete by feedback cutters. A feedback cutter is amplitude-responsive. It has motional feedback. The amplifier and feedback network apply whatever power is needed to achieve the desired amplitude at any given instant. To make it cut the same as a velocity cutter, it needs a different curve. Since it's amplitude-in-amplitude-out, to get the same groove as the velocity setup, you need... falling input amplitude!

Please see the Radiotronics article. Please see the RIAA Recording Characteristic Specification from the Westrex 1700 manual. Both of them say unequivocally that the recording curve is a velocity curve. I quoted them. I'll quote the RIAA doc again:

APPENDIX C
THE R.I.A.A. FREQUENCY RESPONSE CHARACTERISTIC FOR DISK RECORDING

DEFINITION   The recording characteristic is defined as the radial velocity of the modulated groove as a function of the modulation signal frequency, with constant
signal voltage applied to the recording channel bus.

It's a velocity curve. Are you going to deny that?

After the DEFINITION comes the SPECIFICATION, in terms of time constants. After the SPECIFICATION comes a Chart. I transcribed it above; it's the first two columns. After the Chart comes the upward-sloping graph we're talking about.

It's a velocity curve. To transform it into an amplitude curve, you must subtract 6dB/octave.

The Radiotronics article does just that, transforming "Curve A" into an equivalent "Curve B" for amplitude. Curve B is the same as my third column in the table, which I too derived from the Recording Characteristic by subtracting 6dB/octave.

The RIAA Recording Characteristic slopes up in terms of velocity, but down in terms of amplitude. Amplitude is the curve you use on a feedback cutter. And that's the curve for interpreting groove amplitude, no matter which cutter is used.

We are partly in agreement. You said,

"The whole point of the curve is to reduce bass amplitude during cutting, and increase it during playback."

Well, if the default were constant-velocity (flat system response with velocity cutter, magnetic pickup, and no EQ), then relative to that, the RIAA Recording Characteristic does reduce bass amplitude - by not increasing it as much as straight CV would have.

It reduces bass velocity a lot. Amplitude... Nope. Relative to 1kHz, bass is actually increased. Look at column 3 above.

Now return to Curve A, the familiar one. It's flat (CV) below 50Hz, down (CA) from 50 to 500, flat (CV) again between 500 and 2122, and down (CA) beyond that. Regarding that last, again, relative to straight CV, going CA does boost the highs - by not attenuating them. But relative to 1kHz, all the treble grooves are smaller. Faster, yes, but smaller.

All this also agrees with the Galo article, distilled as his Figure 7.

You said,

"You claim: One thing that's not made clear in the figure is that the sine wave in the upper section is represented as
amplitude, while the characteristic curve in the lower section is velocity."

Sorry, I didn't make my reference clear. I'm talking about Galo's Figure 7. It's a composite, showing the groove you'd get from an RIAA-recorded frequency sweep on top, and schematic-form (no 3dB-point rounding) recording and playback curves on the bottom. He marked the vertical scale "Amplitude", but that applies only to the groove, not the recording curve. That curve has velocity as its (unmarked) vertical scale. (Since the pickup transforms velocity to amplitude, the playback curve has amplitude or velocity as its vertical scale - they're equivalent.)

Thanks for the title change. It's obvious I'm just getting my legs here.

Vintage Dave (talk) 00:03, 29 April 2014 (UTC)

I'm sorry, but what you have written above is what WP calls original research, or perhaps "original synthesis". WP does not publish original research. WP publishes information that can be verified directly from reliable sources. Not by saying "well, that curve is really showing so-and-so, and we have to take into account", etc. If you find a WP:RS that says, in so many words, that the groove excursion is wider with the RIAA recording curve than it would be without it, please cite it. (btw: Since one of the stated purposes of the recording curve is to decrease the required groove pitch, that is what most, I think, would consider an extraordinary claim, and it's going to need extraordinary proof to back it up.) Likewise, if you can find a published graph of the curve that shows the recording curve sloping up from left to right as you claim it does, cite that. Otherwise I'm afraid you're wasting your time here, as your claims, reasoning, derivations, and calculations don't mean anything as far as WP's standards of verifiability are concerned. That may seem harsh, but that's how WP works. Jeh (talk) 05:01, 29 April 2014 (UTC)

 if you can find a published graph of the curve that shows the recording curve sloping up from left to right as you claim it does, cite that

I already did. Two sources, already cited.

Here is the first source, cited for the third time: RIAA Recording Characteristic Specification from the Westrex 1700 manual. It states clearly that the Recording Characteristic is a velocity, there's a graph, and the graph slopes upward to the right.

Here's the second source, cited for the second time: Radiotronics article. (The relevant part is the paragraph titled "Equalization", and Figure 6.) It states clearly that the Recording Characteristic is a velocity, there's a graph, and the graph slopes upward to the right.

Please read these sources.

Vintage Dave (talk) 15:35, 29 April 2014 (UTC)

What is more telling is the description from the Radiotronics article: "If now the same [recording] curve were plotted in terms of stylus displacement, then ... curve B." Which slopes down to the right. So they're deliberately using a curve that results in greater displacement (greater groove width) than one would get without any preemph? The trouble you're facing is that that contradicts a great many sources, already cited, which state that it is necessary to apply EQ during recording to reduce groove width. How do you answer that? Jeh (talk) 16:11, 29 April 2014 (UTC)

Is your "no-preemph" case constant-amplitude cutting, or constant-velocity? Since you like velocity cutters, I figure it's CV, but I'll cover both.

Curve A describes groove velocity, as the article clearly states. Curve B is the exact same pre/de-emphasis, only expressed in terms of displacement instead of velocity.

They are the same EQ. Curve B is what it is. It slopes down. What conclusion can be drawn that's neither absurd nor disingenuous?

RIAA cuts larger bass grooves than CA.
RIAA cuts smaller bass grooves than CV.

Do you deny it?

Both statements are true. They do not contradict each other. It's not a simple, straightforward, black-and-white situation. RIAA EQ is a compromise. To understand why, it's necessary to work out the problems arising from the various simple approaches, and steer a path between them.

Let's go to a third source, second citation. The first time I cited it, you dismissed it as "self-published", but it's not.

Galo, Gary A.
"Disc Recording Equalization Demystified".
ARSC Journal, Volume 27, Number 2 (Fall 1996), pp 188-211.
Journal back-issue index here.
Association for Recorded Sound Collections.
The biannual ARSC Journal is devoted to research
on sound-recording history,
preservation and restoration of sound
recordings, record and book reviews,
and much more. You can obtain membership
information from Peter Shambarger,
Executive Director, ARSC, PO
Box 543, Annapolis, MD 21404-0543.
Later collected in "The LP Is Back!", paperback, 160 pages, Audio Amateur Publications, 1999, ISBN 1882580214.  Available at Amazon.
Contains References and Bibliography.
Cited by three ARSC papers and one AES paper.
53 Google hits on the exact title, including citations by other papers and technical manuals worldwide.
Google Scholar link (1 of 4) to "The LP Is Back!", Chapter 14 here.

"Neither a constant-amplitude nor a constant-
velocity recording characteristic
can meet all of the requirements, yet
both have unique advantages.
[...]
However, constant-velocity
recording has two inherent problems.
First, since the recorded amplitude
rises as the frequency decreases, groove
excursions at low frequencies become
too large. This not only makes the record
difficult or impossible for the playback
stylus to track, but it also limits the playing
time on the record."

There. Galo points out that bass (a) is too big, and (b) takes too much space. When straight CV is used.

"The second problem with constant-velocity
recording is that relative to the low
frequencies, the highs are recorded at an
extremely low level, potentially lower
than the surface noise of the record."

Straight CV is also a problem for treble.

"The particular advantage of constant-amplitude
cutting is that it works well at
low frequencies. It holds groove excursions
to a reasonable level, since the
recorded amplitude does not increase as
the frequency drops."

RIAA EQ keeps the bass to a practical level - but later it's made clear that it is still boosted relative to midrange.

"It [CA] also minimizes
the high-frequency noise problem, because
the recorded amplitude will always
be higher than the surface noise of
the record."
"However, there is no free lunch here,
either. Recording high frequencies at
these levels will also cause stylus-tracing
distortion [...]
Wide excursions at high frequencies
also cause cartridge/stylus tracking
problems.

"Maxfield and Harrison opted for a hybrid
recording characteristic that used both
constant-amplitude and constant-velocity
cutting to best advantage."

He's referring to early 78 electrical recording, which used CV at the bottom and CA at the top.

"The constant-amplitude characteristic
limits the groove excursions in the bass
region, thus minimizing stylus-tracking
problems and maximizing playing time."

There is the stated reason for using CA at the bottom. The result:

"Relative to the constant-velocity region,
the bass frequencies are attenuated."

That is, if CV were employed all the way to the bottom, bass would be larger than it is. CA prevents that. But nevertheless, compared to midrange, bass amplitude is boosted:

"The constant-velocity characteristic
between the bass and treble turnover
points then functioned as a transitional
region between the higher amplitude in the bass
and the lower amplitude of the treble." (Emphasis mine.)

There are two CA regions, one between 50 and 500, and one from 2122 on up. The 50-500 is louder than the 2122-15k, because there's a CV region between them, in 500-2122.

"As electrical recording advanced, it became
possible to extend the high-frequency
response of the recording well
beyond Maxfield and Harrison’s 6kHz
limit. In order to prevent surface noise
from burying the extended high frequencies,
later electrical 78s were cut with
the characteristic shown in Fig. 6. Here,
the constant-velocity characteristic was
not used up to the highest recorded frequency.
At a predetermined transition
point, the treble transition (or treble
turnover) frequency, the cutting characteristic
became, once again, constant amplitude.
This switch back to a constant-amplitude
characteristic is often referred to as
treble preemphasis, but this term is misleading.
The treble region is boosted
only in terms of recorded velocity—the
amplitude is still lower than it is below
the treble transition frequency." (Emphasis mine.)

Now a description of the modern RIAA curve.

"Below 50Hz (actually 50.05), the
characteristic is constant velocity. The
bass and treble turnover frequencies
have been set at 500Hz (actually 500.5)
and 2.122kHz. [...]
The reason for the switch to constant
velocity below 50Hz is related to low-frequency
noise in the playback
process. As the turntable speed progressively
decreases, from 78 to 33-1/3 rpm,
turntable rumble and record-warp noise
also become lower in frequency."
[...]
"This minimizes
low-frequency noise problems, including
turntable rumble, record-warp
noise and AC power-line hum."
[...]
"RCA Victor
decided that this frequency [Columbia's 100Hz] was too
high, since it could result in excessively
wide groove excursions at the lowest
recorded frequencies. Hence, a frequency
of 50Hz was accepted as a sensible
compromise between reasonable
groove excursions and noise levels."

RIAA EQ reduces bass groove excursions, even though they are bigger than with CA. It's a compromise, especially below 50Hz, where we put up with wider and wider excursions in order to mask rumble.

"Electrical recordings are not cut with a
flat frequency response, but conventional
descriptions of the recording characteristics
used for 78-rpm and long-playing
records are often misleading. Disc
records are not cut with the bass attenuated
and the treble boosted, at least not
in terms of recorded amplitude. Electrical
recordings are made using a combination
of constant-amplitude and constant-
velocity cutting."

Vintage Dave (talk) 18:28, 29 April 2014 (UTC)

After all the effort you've put me through, I think it's not unreasonable to ask that you meet the same standards, and quote the chapter and verse that is making you fight me so fiercely. Justify your resistance. Vintage Dave (talk) 18:34, 29 April 2014 (UTC)

One down.

[Cut and Thrust: RIAA LP Equalization By Keith Howard, Stereophile March 2009 http://www.stereophile.com/features/cut_and_thrust_riaa_lp_equalization/index.html], one of the sources you base your position on, actually backs mine. See "Fig.2 RIAA preemphasis (record) curve (blue), with the +6dB/octave equalization required for a true constant-amplitude cut (red). " Note carefully: the blue (RIAA) curve is above the red (CA) curve at low frequencies, and below it at high frequencies. It follows inescapably that

RIAA-cut grooves are bigger at low frequencies than at high frequencies.

Nowhere in that article do they say that RIAA "attenuates bass to save space" or "boosts highs to overcome noise". In fact they don't say it attenuates bass or boosts highs at all. They like straight CA. The only reason they give for RIAA not being straight CA is a parenthetical aside to the effect that the latter requires "too much gain" (about which I'm skeptical), and they do not explain the complex nature of the curve.

Vintage Dave (talk) 19:07, 29 April 2014 (UTC)

"Audio IC Users' Handbook", by R.M. Marston, Newnes, page 41, "RIAA Equalization".

"To help enhance the effective dynamic range and signal-to-noise ratio performance of discs,
frequencies below 50Hz and those in the 500Hz to 2.12kHz midband range are recorded in a
nonlinear fashion that [...] causes a midband drop of 12dB when played through linear-response [...] pickups."

It's clear from the lead-up to this they acknowledge that most of the range is CA, with "nonlinear" CV regions. From 50 to 500 it's CA, then a 12dB drop, then CA again. Therefore bass is wider than treble. By 12dB. (Except below 50Hz, where it's even louder.)

Vintage Dave (talk) 19:27, 29 April 2014 (UTC)

Copeland, Peter (Sep 2008). MANUAL OF ANALOGUE SOUND RESTORATION TECHNIQUES p.105. London: The British Library.

"[T]he constant-velocity section ran only from 500Hz to 2120Hz, so more of the frequency range approached constant-amplitude,
and surface noise was further diminished. [...] But it was found this type of bass cut caused too little amplitude at extremely
low frequencies, and turntable rumble was apt to intrude. (Emphasis mine.)
So the 1955 standard specified constant-velocity for all frequencies below 50Hz."

Reason for CV shelf below 50Hz: To mask rumble.

Vintage Dave (talk) 19:38, 29 April 2014 (UTC)

None of the other WP article sources that I have easy access to explain why the boost/cut was done. If you have sources that disagree with mine (and agree with the article's current assertions), please list them here, with the relevant sections quoted.

Vintage Dave (talk) 22:08, 29 April 2014 (UTC)

Dude. I don't know what switch you have in the wrong position, but believe me, you are dead wrong here. See the Alexandrovich section of the Ballou encyclopedia for sound engineers, which I quoted above. Low frequencies are attenuated going to the cutting head, and high frequencies are increased in amplitude. The reverse curve is applied upon playback. Please stop misinterpreting sources to take the opposite view, which is completely wrong. Binksternet (talk) 22:47, 29 April 2014 (UTC)

I want to read the Alexandrovich article, but I don't have the Ballou Encyclopedia and have no plans to drop the money for a copy. If you can get me a peek at the relevant pages, please do, and I promise you I will read them closely.

...

I found Alexandrovich's Section 27.3 in the Handbook For Sound Engineers, courtesy of Google Books. I had to trick it into giving me page 1017. Table 27-1 is another RIAA chart where you can look up any octave(decade) and see for yourself that it's less than 6(20)dB. Which means that Alexandrovich is now one of MY sources. All your base are belong to me!

I agree with everything you quoted. You and he are assuming a velocity-output cutting system. He even has a graph showing the groove displacement on a CV-cut sweep. Bold emphasis mine below.

"[I]f disk recording is attempted without equalization [i.e., constant-velocity], then the
low-frequency excursions would take all the space. The high frequencies
[whose CV groove excursions were decreasing at 6dB/octave] would be of such a
low amplitude that during the playback, high frequency signals could be very close to the
noise level of the system. ...
[I]n the early days... the cutting head sensitivity was decreased at low frequencies
[yes, the Blumlein Curve et al, which attenuates the lows so they are cut CA instead of CV]
so that the amplitudes in midrange and at high frequencies could be recorded at higher levels...
[without those pesky bass groove excursions]
[T]he playback amplifiers were adjusted to boost the low frequencies to compensate for the losses introduced in recording."

So you see that bass frequencies are decreased during the disc cutting operation

I certainly do. They certainly are. In the signal chain on the way to the cutter. Which reduces cutting velocity. If the decrease in signal amplitude - and cutting velocity - is 6dB/octave, this results in a CA groove instead of a CV groove. A CV groove gets bigger and bigger as the frequency goes down. A CA groove stays the same size. But its speed decreases.

What I'm pointing out - and seem to have kicked a hornet's nest - is that an RIAA-cut frequency sweep will have bigger grooves at bass than at midrange. Not faster, just bigger. The RIAA curve slopes at less than 6dB/octave. Which means you don't get straight CA, it's part CV too. (Actually CV/CA/CV/CA.)

You can see it on the chart. You can see it on the graph. What I can't see is how you can disagree with it.

Vintage Dave (talk) 23:53, 29 April 2014 (UTC)

The vinyl does not care what the velocity is. Instead, the amplitude of the wave is the critical factor for cutting the vinyl. Too much amplitude and it throws the playback needle out of the groove. Too small and the signal is down in the noise. Why do you keep harping on velocity? The RIAA curve is all about amplitude. The velocity of the components is what the circuit designer cares about, to achieve the amplitude changes through the filter.

You have not stumbled on a hornet's nest, you are instead caught up in too many details, failing to grasp the larger picture. Give it up, man, put your mental powers on something else. This article is not going to flip a century worth of engineering 180 degrees to please you. Binksternet (talk) 02:02, 30 April 2014 (UTC)

the amplitude of the wave is the critical factor for cutting the vinyl

The lathe people always assume velocity unless otherwise stated. The reference level when cutting is typically 5cm/s. All Shure pickups are specified in terms of velocity. The spec (quoted below) is for velocity, not amplitude. Go back and re-read the Alexandrovich you quoted at me. He describes the "constant velocity" recording curve.

Too much amplitude and it throws the playback needle out of the groove. Too small and the signal is down in the noise.

Have you ever heard me disagree? No, I agree, and have said so.

The RIAA curve is all about amplitude.

No, it's velocity. See actual spec, quoted and linked below.

It's not a question of "flipping" tradition, nor of pleasing me. It's about correcting a local misconception that has the potential to confuse widespread readers.

RIAA Recording Characteristic Specification from the Westrex 1700 manual. Bold emphasis mine. (Also, transcription typo corrected: "DISC" not "DISK".)

APPENDIX C
THE R.I.A.A. FREQUENCY RESPONSE CHARACTERISTIC FOR DISC RECORDING

DEFINITION   The recording characteristic is defined as the radial velocity of the modulated groove as a function of the modulation signal frequency, with constant
signal voltage applied to the recording channel bus.

There is a clear, straightforward relationship between frequency, amplitude, and velocity. It's even simpler for speed (what they meant by "velocity" above). Then, v is just ωA. Dynamics of simple harmonic motion

Vintage Dave (talk) 16:52, 30 April 2014 (UTC)

Here's an excerpt from the technical manual for the

SCULLY AUTOMATIC FEED CONTROL

In view of the RIAA characteristic inserted into this unit, it will be noted when sweeping the
frequency spectrum that a greater expansion will take place for the lower frequency and lesser
expansion for high frequency.

Post by user "EmAtChapterV" of The Secret Society Of Lathe Trolls, consisting of a transcription of the entire manual. Search for "RIAA"; one hit, quoted above.

Vintage Dave (talk) 19:56, 30 April 2014 (UTC)

"Constant velocity", a.k.a. "constant angular velocity" refers to the platter of the turntable turning at the same speed from throughout playback. Constant linear velocity is the other choice, where the turntable spins faster when picking up signal near the center than when picking up signal near the outer perimeter (like a CD). Don't mix up the turntable speed with the cutting head or the magnetic cartridge velocity, nor with the electronic velocity which sets the frequency response of components in filters such as the RIAA filter. Binksternet (talk) 21:03, 30 April 2014 (UTC)

I'm not confusing platter rotation with cutting. (If I wanted to do that, I'd talk about wavelength. <g>) Read your own quote from Alexandrovich. He - and I - use the term "constant velocity" in describing the side-to-side stylus motion in a cutting head. The speed the tip moves at as it vibrates, in response to audio coming into the cutting amp. When the cutting system is constant-velocity, the vibratory motion of the tip maintains the same speed regardless of the frequency, given a constant-amplitude exciting waveform. This has nothing to do with platter rotation. In fact, the platter does not need to be present - let alone spinning at a particular speed - for the cutting-head stylus to vibrate. It's all about the vibration.

In other words, when you give it a level frequency sweep, the tip swings from side to side at the same speed, from beginning of sweep to end. But as frequency (repetition rate) goes up, the tip can't swing as far before it has to turn around and come back: amplitude decreases. Alexandrovich shows this in the figure accompanying the source material you quoted.

On the other hand, when the cutting system is constant-amplitude, then the tip displaces by an amount exactly proportional to the voltage applied at the input jack, regardless of frequency. A flat sweep makes the tip trace the same size back-and-forth pattern, but faster and faster, travelling the same path more times per second: velocity increases.

But we're talking about constant-velocity cutters. I typed the preceding paragraph purely to compare and contrast.

Vintage Dave (talk) 21:55, 30 April 2014 (UTC)

Constant-Velocity Recording
4.050  Constant-velocity recording indicates a mechanical recording
characteristic wherein, for a fixed amplitude of a sinusoidal signal,
the resulting recorded amplitude is inversely proportional to the frequency.

Pre-emphasis (Pre-equalization)
4.580  In recording, pre-emphasis is an arbitrary change in the frequency
response of a recording system from its basic response (such as
constant velocity or amplitude) for the purpose of improvement
in signal-to-noise ratio, or the reduction of distortion.

Read, Oliver: "The Recording and Reproduction of Sound", pages 697 and 705, Recording And Reproducing Standards, Glossary.

Also see page 690, graph, "NAB RECORDING and REPRODUCING STANDARD": the vertical scale is designated "RECORDED VELOCITY", and it's a rising curve. It's not RIAA, but it shows that an immediately-preceding curve represented velocity, not amplitude. Also that constant-velocity has nothing to do with the platter.

Note especially that in constant-velocity, the recorded amplitude is inversely proportional to the frequency. In other words, as frequency goes up, amplitude goes down. This agrees with Alexandrovich, and everyone else.

Vintage Dave (talk) 23:27, 30 April 2014 (UTC)

"Radio and Television News", July 1954, article, "The Curve That Conforms", shows a graph, on which the (sole) recording curve is labeled

RIAA-NEW ORTHOPHONIC-NEW AES-NARTB

which indicates that the RIAA and NARTB curves are the same. (They have the same CV/CA/CV/CA structure, but I did not attempt to graphically compare dB levels.) Since the NARTB curve documented in "The Recording and Reproduction of Sound" states unambiguously that the curve represents velocity, the RIAA curve, being the same curve, also represents velocity.

Vintage Dave (talk) 23:49, 30 April 2014 (UTC)

Olson, Harry F.

Modern Sound Reproduction

Van Nostrand Reinhold, 1972

Excerpt from Section 9.2, "DISK PHONOGRAPH RECORDING PROCESS", page 171. Bold emphasis mine.

The Record Industry Association of America has established a standard
recording frequency velocity characteristic as depicted in
Fig. 9.6.

Figure 9.6 depicts two curves, labeled A and B. Curve A is the familiar rising curve. Curve B falls. They cross at 1kHz/0dB.

[...]
The frequency amplitude characteristic corresponding to the velocity is also shown in Fig. 9.6.

Although Olson's Curves A and B are exactly the same as Fig. 6 in the Radiotronics article, the latter includes a third Curve C.

Ceramic cartridges, like all piezoelectric devices, are responsive to displacement,
so that the playback equalization required is the inverse of Curve B and not of
Curve A (the inverse of Curve A is of course the playback equalization required
with electromagnetic devices), and is shown in Curve C.

Vintage Dave (talk) 00:17, 2 May 2014 (UTC)

Excerpt from page 1 of the Neumann VMS-66 manual, courtesy of user "cohearent" at Lathe Trolls, posted Mon Aug 31, 2009 9:29 pm in Inverse RIAA Revisited. Bold emphasis mine.

This curve [RIAA Recording Characteristic] has an irregular slope but is, in any case,
sloped less than the 6 dB per octave which would be ideal.
The difference between its 30 Hz and 15 kHz level is 35.8 dB,
which leaves a disparity of 18.2 dB from constant amplitude recording,
a 30 Hz tone occupying about eight times as much space as 15 kHz.

Do you-all (Jeh and binksternet) have any more objections? I'll work on condensing all this into an article revision.

Vintage Dave (talk) 20:50, 6 May 2014 (UTC)

I recommend that you do not touch the article. Your misinterpretation of the sources is so severe that I doubt you will be able to improve the article. Binksternet (talk) 22:32, 6 May 2014 (UTC)

Point out how and where I've gone wrong. Use verifiable sources, and quote chapter and verse. Be specific. Show your work.

Vintage Dave (talk) 00:27, 7 May 2014 (UTC)

I cannot interpret "bass amplitude is increased" in any way except the following: in an RIAA-preemph'd recording the grooves take more space (larger bass amplitude) than they would if not preemph'd. And that makes no sense at all. Also, it is obvious that if you play back an LP without RIAA deemphasis it comes out with too little bass and too much treble (and so too much groove noise). Therefore RIAA preemphasis must decrease bass amplitude and increase treble amplitude during recording.

I am not interested in "relative to 1 kHz" or in adding the effects of the cutter's response or velocity vs. amplitude. I am interested in just this: Is the required groove pitch increased or decreased by use of the RIAA preemphasis curve during recording? And is high-frequency groove noise (pops and clicks and swishing) decreased or increased by the RIAA preemphasis and deemphasis curves, relative to what it would have been had they not been used? If you say "decreased" then you are not disagreeing with the fundamental claims the article makes now; you have basically gone off on a tear about terminology. If you say "increased" then I have to ask, why would they do that? Jeh (talk) 03:11, 7 May 2014 (UTC)

You're asking a trick question, Jeh. It depends on the context, that velocity-vs-amplitude you want to ignore. In fact, the whole thing hinges on it!

Statement 1.

The required bass groove pitch is decreased by the use of the RIAA recording characteristic.

Statement 2. Neumann VMS-66 manual, quoted again:

a 30Hz tone occup[ies] about eight times as much space as 15 kHz.

These two statements do not contradict each other. The reason: one is talking about velocity and the other is talking about amplitude. You must not deny the distinction, or ignore it. They are two different things, and if you confuse or conflate them, you will get some nonsensical results. They have a simple relationship:

v=ωA.

I did not invent this; you can find it in the WP article on Simple Harmonic Motion. You can also find it in some of my sources. For example,

NARTB RECORDING AND REPRODUCING STANDARDS
Section 4
Glossary

As reprinted in
Read, Oliver: "The Recording and Reproduction of Sound", on page 697. Bold emphasis mine.

Constant-Velocity Recording
4.050  Constant-velocity recording indicates a mechanical recording
characteristic wherein, for a fixed amplitude of a sinusoidal signal,
the resulting recorded amplitude is inversely proportional to the frequency.

Here are Statements 1 and 2 put together, in a way that explains why they don't contradict:

Bass groove amplitude is decreased by the use of the RIAA recording characteristic -
compared to constant-velocity -
the default -
but decreased less than a constant-amplitude characteristic would decrease it -
therefore -
somewhere between constant-amplitude and constant-velocity -
louder than CA -
softer than CV.

For example, at 30Hz, the RIAA-cut groove swings about 12dB wider than constant-amplitude, but moves about 19dB slower than constant-velocity. Decrease? Or increase? Depends on how you look at it. And that's the point.

I am not engaging in pilpul. I am exposing a fact obscured by the current article.

Now high frequency...

I agree with the article, as long as the context is made clear.

Compared to no EQ (i.e., constant velocity), high-frequency groove noise is decreased when the RIAA recording characteristic is used.

To be symmetric with the bass side, I am tempted to add, "but not as much as when a constant-amplitude characteristic is used", but it's not really true. Limited by the tracking ability of a typical player, RIAA is already as hot as they dared make it. It's a compromise.

I've said all this before, some of it in the form of direct quotes from verifiable, published sources. I don't think you are being deliberately obscurantist, so I must not be presenting the facts well. I'll keep saying them different ways, and something will get through.

Vintage Dave (talk) 17:52, 7 May 2014 (UTC)

More references that say that common practice - and the RIAA Recording Characteristic - treats the cutting system in terms of velocity, not amplitude. Some also explain why. Some also say that professional cutters since the 1940's employ feedback, which is important for achieving the constant-velocity characteristic.

I have scans of all the pages in this group, and I will be happy to upload them if you want to see the figures they refer to.

[My comments are in brackets.]

Benchmark Papers in Acoustics/12 Disc Recording and Reproduction

Edited by H.E. Roys

(c)1978 Dowden, Hutchinson & Ross Inc.

LCCN 77-17927 ISBN 0-87933-309-X TK7881.4.D57 621.389<prime>32

Distributed world wide by Academic Press, a subsidiary of Harcourt Brace Jovanovich, Publishers.

Page 13 (c)1962 Institute of Electrical and Electronics Engineers, Inc. Reprinted from IRE Proceedings, 50:738-744 (May 1962)

Page 742

Disk Recording and Reproduction W.S. Bachman (Fellow, IRE), B.B. Bauer (Fellow, IRE), and P.C. Goldmark (Fellow, IRE)

Standardization of Response Characteristics

"A more convenient method of measuring groove modulation velocity by observing the width of a light patter reflected from a test record was described by Buchmann and Meyer."

Page 17 (c)1926 Institute of Electrical and Electronics Engineers, Inc. Reprinted from Am. Inst. Electr. Eng. Trans. 45:334-346 (1926)

Page 335

Methods of High Quality Recording and Reproducing of Music and Speech based on Telephone Research J.P. Maxfield (Member, A.I.E.E.) and H.C. Harrison (Associate Member, A.I.E.E.)

"Sincwe, for mechanical reproduction, the sound at a given pitch is radiated by means of a fixed radiation resistance, it is necessary that the record must be cut with a device the square of whose velocity is proportional to the sound power. Under these conditions, it is seen that for a given intensity of sound the amplitude is inversely proportional to the frequency of the tone, and that a point will be reached somewhere at the low end of the sound spectrum where this amplitude will be great enough to cut from one groove into the adjacent groove[...]"

"It has been suggested that if electric reproduction were used, it would be possible to cut the record with a characteristic other than uniform velocity sensitiveness and correct for the error by an electrical system whose characteristic is the inverse of the characteristic of record.[sic] If the change which is made in the recording characteristic tends toward cutting at uniform acceleration sensitiveness, the amplitude varies inversely as the square of the frequency and hence the difficulties at the low end of the scale are greatly enhanced. Similarly, if the records are cut more nearly at constant amplitude, the radius of curvature of the sine waves decreases as the square of the frequency, hence the difficulties are placed at the upper end. In the process which is being described in this paper, these limitations have been met commercially by having a frequency characteristic of the uniform velocity type between the frequencies of 200 and approximately 4000 cycles per second. Below 200 it has been necessary to operate at approximately constant amplitude with a resulting loss in intensity which loss increases as the frequency decreases. Above 4000 it has been necessary to operate at approximately constant acceleration with its consequent slight loss in intensity at the very high overtones."

Fig. 9 - Velocity Response for Various Values of Mechanical Constants

Curve D "a band pass type of circuit whose resistance impedance is the

same as that of the system shown in Curve A." [Approximately flat, i.e. constant velocity, between about 300Hz and 3kHz.]

Page 30 (c) 1927 Institute of Electrical and Electronics Engineers, Inc. Reprinted from Am. Inst. Electr. Eng. Trans. 46:903-911 (1927)

Electrical Reproduction from Phonograph Records Edward W. Kellogg (Associate, A.I.E.E.)

Page 904

"In terms of sine waves, the deflection devices [...] give a voltage proportional to amplitude independent of frequency, while the velocity devices [...] give a voltage proportional to amplitude multiplied by frequency.

" Which of these types of device should we choose for a phonograph reproducer? If the system by which the record is cut is so designed that the deflection of the cutting point is proportional to the original sound wave pressure, then a deflection type reproducer is required. If the record is cut by a tool whose instantaneous velocity is proportional to the sound wave pressure, we shall require a velocity type reproducer. Both systems are equally correct from the standpoint that if their conditions are complied with, distortionless reproduction will result. The choice can, therefore, rest on such considerations as scratch noise ratio, wearing qualities of record, interchangeability of records, and designing a practical device which performs in accordance with the theoretical requirements. By interchangeability of records it is meant that it is desirable that the electrical reproducing system shall not only give good results with records that are especially cut for it, but so far as possible give pleasing results with records cut by the old process, and the records which are correctly cut for electrical reproduction should sound well when played on a horn type machine. This requirement is most nearly met by the velocity system of cutting and reproducing. [...] It is, therefore, clear that if the electrical reproducing system is to be such that records may be satisfactorily interchanged, it must work on the velocity principle."

Page 52 Editor's Comments

"Vieth and Wiebusch, well aware of the benefits derived from the use of feedback in amplifiers, applied the same principle to the development of a vertical recorder. A small coil, located in a magnetic field, was tightly coupled mechanically to the cutting stylus. The induced voltage during operation was thus proportional to the stylus velocity. This voltage, when returned to the input of the driving amplifier after passing through a suitable control circuit, provided the desired feedback control. The result was a controllable recording characteristic, freedom from distortion, and a reduction in noise arising from amplifier and recorder. " Probably the first lateral feedback recorder was the one developed under the direction of Arthur Haddy of the Decca Record Company of London. [...] Later the system was used to record the "ffrr" (full

frequency response reproduction) music records.

Page 57 (c)1938 by the Society of Motion Picture and Television Engineers, Inc. Reprinted with permission from Soc. Motion Pict. Eng. J. 30:96-104 (Jan.

1938)

Recent Development in Hill and Dale Recorders L. Vieth and C.F. Wiebusch (Bell Telephone Laboratories, New York, N.Y.)

"Fig. 1 shows diagramattically a feedback disk recorder-amplifier system. [Amp, drive coil, feedback coil]

"The purpose of the system is to move a cutting stylus in a recording medium with a vibrational velocity whose wave-shape is an exact replica of the wave-shape of the signal voltage. [Discussion of feedback theory, including Nyquist stability criterion.] If B is now increased until, over some chosen frequency range, the magnitude of AB is large compared to unity, equation 5 becomes

V = (1/-B)*E (eqn. 6)

which indicates that over the frequency range considered the velocity of the stylus is independent of the amplifier gain or the efficiency of the recorder."

Fig. 5 Curves showing the stylus velocity for a constant signal input to the recorder amplifier system. (A) Without feedback. (B) With properly controlled feedback. [Curve B is flat. This figure is virtually identical to Fig. 9.3 in Olson, "Modern Sound Reproduction", page 169, quoted previously.]

Page 65 "ffrr" Lateral Feedback recorder A.C.W. Haddy, O.B.E.

"The moving coil driving element and its associated feed-back coil[...]" "Once the feed-back amplifier had been calibrated it rarely needed readjustment[...]"

Page 72 (c)1949 North American Publishing Co. Reprinted from Audio Engineering 33:22-27 (1949)

Lateral Feedback Disc Recorder G.R. Yenzer (Bell Telephone Laboratories, Inc.)

Page 22

"This paper describes the W.E. 2A Lateral recorder, developed around the

same general feedback principles as the corresponding vertical feedback cutter." [Cites Vieth/Wiebusch paper.]

"The recorder is connected to an amplifier system as shown, the object

being to move the cutting stylus with a vibrational velocity V whose wave shape is an exact replica of the wave shape of a signal voltage E. " A general expression for the relation between stylus velocity and signal voltage in an electromechanical feedback system has been previously developed." [Cites Vieth/Wiebusch again.]

Page 74 (Page 24) "Another method for determining stylus velocity is applicable only to the feedback cutter. The voltage E3 generagted in the feecback coil is proportional to the velocity of the coil. [...], voltage E3 is also a true indication of the stylus velocity, V.

"In Fig. 5 curves A and B, the stylus velocity was measured by the

feedback coil method, and in curve C the stylus velocity was determined by the Buchman-Meyer optical method."

Fig. 5 Frequency response of feedback cutter. [Curves B and C, "Optimum feedback" are flat.]

"For the feedback condition a signal voltage E of 0.5 volts applied to the input terminals produces a stylus velocity of approximately 1 inch per second maximum velocity [...]"

Page 75 (Page 25) "As the recording art is practiced today, the recorded level of the low frequencies will be limited by the amplitude [emphasis his] of stylus motion in order to prevent cutting into adjacent grooves, the level of medium frequencies will be limited by the velocity of the stylus so as to prevent the heel of the recording stylus from touching the groove wall, and the level of the high frequencies will be limited by the acceleration of the stylus to prevent the radius of curvature of the groove from becoming less than that of the reproducing stylus."

Page 125 (c)1940 Acoustical Society of America Reprinted from Acoust. Soc. Am. J. 12(2):309-306 (1940)

Page 303

A new optical Method of Measurement for Phonograph Recordings G. Buchmann and E. Meyer Translated by J.M. Cowan

"The most important magnitude in phonographic recording and reproduction is not the amplitude itself, but the stylus-point velocity; i.e., the product of amplitude and frequency. An ideal electrical cutter should record constant stylus-point velocity at all frequencies when activated by an alternating current of constant voltage; amplitude, on the other hand, varies inversely with frequency for constant voltage input."

Page 137 (c)1942 IEEE Reprinted from IRE Proc. 30:355-356 (Aug. 1942)

Page 356

Recording and Reproducing Standards Lynne C. Smeby (Member, IRE) Chairman, Recording and Reproducing Standards Committee; formerly, Director of Engineering, National Association of Broadcasting.

"14. Frequency Characteristic for Lateral Recording It shall be standard that the recorded frequency characteristic on laterally recorded records be as shown in Fig. 2."

Fig. 2 - Recording characteristic for lateral transcriptions. Stylus velocity-frequency. [Rising curve, same shape as RIAA. Vertical scale is "RELATIVE STYLUS VELOCITY (dB)".]

Page 319 (c)1949 IEEE Reprinted from IRE Proc. 37(8):923-927 (1949)

Page 926

The Columbia Long-Playing Microgroove Recording System Peter C. Goldmark (Fellow, IRE), Rene Snepvangers (CBS, INC), and William S. Bachman (Member, IRE)

"The LP recording characteristic (velocity versus frequency) is shown by curve 1 in Fig. 3. Curve 2 in Fig. 3 shows the NAB velocity recording characteristic, which deviates from the LP characteristic in the low bass portion only. The purpose for the LP bass lift is to reduce rumble and hum pickup, which would otherwise be more pronounced with the smaller deviations."

[Fig. 3 - LP and NAB recording characteristics. Rising curve similar to RIAA. Vertical scale is dB. Curve 1 labeled "LP RECORDING VELOCITY CHARACTERISTIC", cURVE 2 LABELED "NAB RECORDING VELOCITY CHARACTERISTIC".]

Page 339 (c)1958 IEEE Reprinted from IRE Proc. 46:1686-1693 (Oct. 1958)

Page 1686

The Westrex StereoDisk System C.C. David (Eng. Div., Westrex Corp.) and J.G. Frayne (same)

"It was decided that the new [45-45] recorder should retain the feedback principle which has been an outstanding feature of both the Western Electric and Westrex recorders for the past 20 years and has been described previously.* [...] This arrangement of feedback maintained good stability over the significant range of audio frequencies, and [...] essentially constant-velocity recording was provided over a frequency range from 30 to 15,000 cps[...] This characteristic is modified in commercial practice by inserting the RIAA or equivalent recording preequalization characteristic. [...] Since the characteristic of the network used to provide constant-velocity recording is to some extent complementary to the RIAA preequalization recording characteristic, the two characteristics can be combined in a single equalizer with a considerable reduction of insertion loss."

[cites Vieth and Wiebusch.]

Vintage Dave (talk) 16:34, 30 May 2014 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified 3 external links on RIAA equalization. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20110409161239/http://www.bl.uk/reshelp/findhelprestype/sound/anaudio/analoguesoundrestoration.pdf to http://www.bl.uk/reshelp/findhelprestype/sound/anaudio/analoguesoundrestoration.pdf
• Added archive https://web.archive.org/web/20120922040914/http://www.platenspeler.com/background/riaa/uk_riaa_background_1.php to http://www.platenspeler.com/background/riaa/uk_riaa_background_1.php
• Added archive https://web.archive.org/web/20120908023505/http://www.platenspeler.com/background/riaa/uk_riaa_background_2.php to http://www.platenspeler.com/background/riaa/uk_riaa_background_2.php

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 05:50, 18 September 2017 (UTC)