How does RIAA pre-emphasis work on reducing record noise?

Given that we've not actually *seen* a strain gauge cartridge that was properly equalized, its hard to know how one could discern if that was the right choice. I get the thing about simpler being better and all but I think it was Einstein who said 'Make everything as simple as possible, but not simpler.' A strain gauge is simple but not equalizing it properly is a bit too simple. My opinion of course.

Although IEC has thrown in another inflection point this does not affect the main RIAA which is a accomplished by an expansion of groove size starting at 500 Hz . This EQ is then reversed on playback ( accuratey we hope) with the benefit of lowering the noise starting at 500 HZ. Over this two octave span the EQ boosts the recording cutter's depth putting more signal on the disc. The 3 db rise is called at 500 Hz as one does in any filter. At 2,200 the curve flattens out again and we cut at constant amplitude just as we did below 500 Hz. Over the audio band the total rise is 12 dB. (6 dB x two octaves). Not an insignificant number.

Compare with:

Note that Ralph brought in cutting and other things that have nothing to do with this particular RIAA playback question.

Its obvious that LP mastering has everything to do with this.

I'm of the opinion that I didn't go nearly as far as the situation merits.

In more ways than one Roger, you owe me an apology.

All of that still leaves me uncertain, however, as to how the rising response of the phono cartridge is compensated for. If the velocity response of the cutter is compensated for by an EQ of around 15 db in the associated electronics, and the 40 db curve that is applied by the phono stage compensates for a complementary 40 db curve that is applied in the cutting process, what compensates for the velocity characteristics of the cartridge?

@almarg

This is the fun part. And with considerable ’Yikes!’:
If a cutter doesn’t express 6dB per octave and for that matter isn’t even a linear curve I’m challenged to understand why the same would not also apply to a magnetic pickup. Both exist in the *real* world as opposed to that of theory. Neither are ’perfect’ and fall well short of ’theory’, probably due to mechanical constraints, winding differences, ignored capacitance and the like.

A real world cutter head doesn’t get 3dB out of flat until below about 100Hz or so; between 100Hz and about 15KHz or so depending on the individual character, its flatter than many loudspeakers! Given that such is the case and how adamant some audiophiles are about their favorite cartridge, I suspect that most cartridges simply aren’t true constant velocity devices either. IOW I think they are a lot flatter in response than most people that think about these things think :)
Typically the real world is a lot messier than that of theory and this is probably a pretty good example of that.

^^ Thanks @clio09 for that correction.
I've since tested the tape input as if it was RIAA; even there its off by less than 1/2dB between 3KHz and 20KHz.

At this point I don't think Roger tested this at all.

@rauliruegas
Ceramic cartridges have been most of the market since the 1950s.

This is just my opinion of course but I think its a mistake to say that by equalizing the error in a strain gauge its not going to sound as good. You don't have to run it through a phono section- the EQ needed is not profound so as I mentioned earlier it could be done on a line level input that simply included the needed EQ.

@almarg

In a typical cutter system, as the input signal progresses through the amplification pathway, the velocity characteristic of the cutter is equalized first. This is not a constant slope and covers a roughly **15 dB** range from 20Hz to 20KHz. This EQ also deals with the peak that is typical of any cutter head so you can't describe the cutter has having a 6dB slope- its nowhere near and is somewhat curvaceous!
Also worthy of note: the cutter is within about 1/2dB from 1KHz to about 5KHz- not sloped at all.

After that, the 40dB RIAA pre-emphasis is added. The reason things are done this way is that no cutter has exactly the same characteristic and so the electronics have to be matched to the cutter by hand-selecting components in the cutter EQ (independently of the RIAA) to get it right.

*************************
In other news, clio09 has been a customer of ours for some years. It was his MP-3 that was **allegedly** tested by Roger in a recent phono section shootout. Roger made some pretty outrageous claims about that shootout! My first impression which I expressed earlier is that Roger tested the wrong input. Clio09's preamp is optionally equipped with a tape head input and the phono section thus has switchable equalization.


Recently I had discovered a tape EQ error and so I informed clio09 of this fact in a phone conversation a few weeks ago. I had him send the preamp back to us for this reason; of course I was very curious to find out what was up with this 'measurement' Roger had made earlier!

Right out of the box the preamp went on a test bench, and I set up the input with a iRIAA box (inverse RIAA curve, IOW RIAA pre-emphasis as used in recording) and observed the output on the bench oscilloscope. The preamp spec'ced out just fine- the RIAA characteristic was well within 1/4dB over the entire range, falling off about 1/2 dB or so starting at about 25KHz-27KHz.


My iRIAA box is single-ended, so I used an adapter to run it into our phono section which is fully balanced (the CMRR of the input section is plenty high so this will not affect performance). I observed the output off of one of the tape record outputs which are single-ended.


In email with clio09, Roger had told him that he used essentially the same technique. Now Roger has gone to some length to castigate me in this thread and if the preamp was actually in error he might have had a point. But as I've pointed out twice (now three times), I used Dr. Stanley Lipshitz's formulae to calculate the RIAA values and the preamp has independently been measured and found to be quite a bit more accurate than Roger had claimed.


I find this aspect of this conversation a bit disappointing. Roger has repeatedly attacked me, making claims to the effect that I don't know what I'm talking about, inferred I can't design a preamp and so on.


I can't say if there is a motivation on Roger's part to make me look bad or it he simply made a gross error. The inaccuracies I've encountered on this thread don't help me shed any light on the situation.


Anyway, I thought people reading this should know what's up.

I went to hear a strain gauge at CES about 15 years ago- I brought an LP that I had recorded. It didn't sound right but I kept my mouth shut and just listened to the sales pitch, which was that the cartridge needed no EQ correction. At this time I'm still uncertain if anyone has done anything about this issue.

This is from the Soundsmith website (https://www.sound-smith.com/faq/what-so-unique-about-strain-gauge-cartridge-how-does-it-work-why-no-...):

The Strain Gauge cartridge is a displacement device, producing an output that is dependent and directly proportional to the amount of displacement of the stylus, NOT the velocity, like magnetic cartridges. **It therefore inherently plays the RIAA encoded groove correctly.** If one inspects the RIAA EQ, one will discover that there are two areas that occur where a displacement type cartridge will deviate a small amount from a perfectly flat playback of the recorded RIAA groove, and therefore **will not produce a perfectly flat response.**

Note the double asterisks I placed in the above quote. The sentence in which they first appear is immediately contradicted by the sentence following (in particular at the end where I added the asterisks for emphasis).

@rauliruegas

what do you think at this precise moment with all those statements that says we are wrong?

The matter has always been one of clarity. I mentioned this earlier, in response to a post by Al:

I can’t help but think that somehow what is going on here is that Roger’s text is so poorly written (including numerous spelling errors) that I’ve taken the wrong interpretation of what he’s trying to say. I’ve looked at his posts several times though and can’t come away with anything other than the interpretation that you presented in your last paragraph.

As best I can make out, Roger has had troubles expressing himself- and this has resulted in the corrections you’ve seen from Al and me.
Here’s a recent example of what I’m talking about:

For a strain gauge cartridge, how many dB is the step when taken a few octaves above and below the corners?

There’s no good way to interpret this question! The ’step’ is most likely the relatively ’flat’ portion of the RIAA curve, but that’s about all I can get from this question that makes any sense. I’ve often been accused of being too literal, and I can confirm that’s gotten me into trouble in the past. But when confronted by something like this I just have to do the best I can. In this case, I can’t make sense of it.


However, I have always completely agreed with the text you quoted in your rather lengthy post above. Those quotes are much more clear; when you compare them to Roger’s texts and sort of metaphorically take a step back to view the larger picture, it appears that what Roger is getting at is the nature of cartridges and cutter heads being part of the overall choices made in the creation of the EQ curve.


The curve covers a roughly 40dB range and no matter what cartridge is used for playback you have to find a way to make that work. With magnetic devices this is easy, with amplitude devices this poses some problems. Unfortunately Roger was unclear at various points including his opening post; but I’m thinking now he’s on the right track although not expressing anything new. I guessing that he was more interested in getting amplitude devices (strain gauge and ceramic cartridges) to work properly, and what that would entail as far as the additional EQ you need to add to them to make them work. You’ll note his comments about the Win cartridge (a friend of mine had one of those years ago- it was quite interesting).

As I’ve been writing this I think Roger might have been asking in that question I quoted above what I thought that EQ curve might look like. So I’ll say this about it:
Amplitude devices have a 6dB slope to them that ’sort of’ counters the RIAA curve. The problem is the RIAA curve has a step between 500Hz and about 2.2KHz that throws such cartridges off (If you line up their response to the 1KHz center point of the RIAA step, they will have too much bass and not enough treble but consistently the same across either spectrum, which is to say that as frequency decreases or increases the dB error is constant since they just express a simple 6dB slope). I think one way to deal with that is assume that bass portion of the curve will be correct, then introduce a 3db/octave boost at 500Hz which ends at 2.2KHz; at that point returning to the 6dB slope of the cartridge. This is assuming that you are putting that cartridge output into an auxiliary level input modified by this curve.


In A nutshell, I’m now thinking we’ve all been dancing around the same thing- expressing it in different ways.

Who wrote this?? "  

Uh, I did- and it was in reference to this statement, which is incorrect:

a strain gauge cartridge is a displacement device not a velocity one and can read perfectly/direct any LP through a phono stage WITH OUT that velocity/magnetic eq. de-emphasis curve named RIAA

Although this quote confirms exactly what I was saying:

Precisely, and that proper EQ is indeed the RIAA with corners at 500 and 2200 HZ. Problem is that many ceramic and Strain Gauge makers try to EQ mechanically in the cartridge. This is difficut. I worked with Sao Win for years trying.

Obviously its not both ways. The latter quote is correct.

The recording eq. happens only at two corners between 500hz and 2,212hz  . """

I know you don't agree with but just suppose it happens that way. I'm asking you if the next statement is true because of that or what is happening with a straing gauge device that's not magnetic but a displacement device:

" That’s why a straing-gauge cartridge ( as the SS. ) can’t be connected to any phono stage with velocity frequency eq. ( that are the ones we all own. ), a strain gauge cartridge is a displacement device not a velocity one and can read perfectly/direct any LP through a phono stage WITH OUT that velocity/magnetic eq. de-emphasis curve named RIAA. ""

@rauliruegas
A strain gauge or ceramic cartridge won't have flat output if not properly equalized. So the final paragraph above is false on account of the second half of the sentence, starting at "a strain gauge...". They have high output so you don't need the gain of the phono preamp, but you do need to equalize them nevertheless because of that step in the RIAA curve between 500Hz and 2.2KHz.

How do you most experienced persons feel about introducing the 3.18 usec time constant into the de-emphasis RIAA filter? Yea or nay?

Nay. It was pointless and withdrawn- don't worry about it.

Everyone thinks the 40 dB curve is RIAA, however it is RIAA plus the EQ of the velocity cartridge. I wanted people to think about what is actually on the record and how the cartridge reproduces it. Will those who now understand that please step forward.

As far as atmasphere is concerned it is Ralph who has hit on me not only technically but personally concerning writing and spelling. I am clear that Ralph trolls my posts and of course disagrees with me on everything from damping of an amplifier to the color of the sky. Many times I has asked him to depart yet he goes on promoting his power paradigm and other ideas.

Besides being impolite starting statements with "Roger is wrong" or "Roger doesnt get it" he will take the opposite of anything I say. The problem is that many of the readers here do not know enough to be able to decide.

Look at who was the first to respond to the OP. Look at what he said. He even misinterprets my first sentence to mean two octaves when I say "over the audio band". Then he says  

 So- what is the answer you've been given, and what are you suggesting is the correct answer? Neither is depicted in your OP.

However this is in my OP    Over this two octave span the EQ boosts the recording cutter's depth putting more signal on the disc. The 3 db rise is called at 500 Hz as one does in any filter. At 2,200 the curve flattens out again and we cut at constant amplitude just as we did below 500 Hz. Over the audio band the total rise is 12 dB. (6 dB x two octaves). Not an insignificant number.

If Ralph knows so much about RIAA, why does his preamp measure down 8 db at 7 KHZ where all others I measure are flat?

I come here to educate, nothing more. I believe I have done my research carefully and related in simple terms that others can understand. My fondest wish is that Ralph would make an effort to learn these things rather than jump on anyting I have to say... including this.

Roger- you have it wrong about me and you have it wrong here regarding the facts of the RIAA. I don't troll you- I just look at this forum for posts that look interesting. I found your threads on two occasions and in both found mis-information. As I recall, you didn't use to be so adversarial... I've yet to argue the color of the sky :)  OTOH you've contradicted yourself in this thread; what's a person to think??


Regarding the preamp, Its EQ network values come directly from Dr. Lipshitz's formulae; when you drop the values into an RIAA calculator they come out right and the preamp measured correctly independently. The particular preamp you measured is also equipped with tape EQ; it appears you measured the wrong curve. 


Regarding actual technical matters, the following statement is false:

 Over this two octave span the EQ boosts the recording cutter's depth putting more signal on the disc.

I modified the above quote so it is more correct:

Over the ten octave span the EQ boosts the recording cutter's output putting more signal on the disc as frequency increases.

'two' is replaced by 'ten', 'output' was substituted for 'depth' and 'as frequency increases' added for clarity.


As I mentioned before, groove depth is not a function of the EQ curve. Its an adjustment. Here is the text about it from the Westerex manual:

The advance ball mounting is shown in Figure IV-2. Two thumbscrews provide vertical and lateral adjustment of the position of the ball, thus controlling *depth of cut* and *tracking*

(emphasis added as the original text is in italics which is not available on this website)

@almarg

Agreed. I can't help but think that somehow what is going on here is that Roger's text is so poorly written (including numerous spelling errors) that I've taken the wrong interpretation of what he's trying to say. I've looked at his posts several times though and can't come away with anything other than the interpretation that you presented in your last paragraph. Plus some contradictions.

Hello Raul, If you mean this one, seems alright:
http://hifisonix.com/inverse-riaa-network/

Hello Raul, IMO the pspatial article should be ignored as it is poorly written; Al's interpretation is mine as well.

I think Roger is just having a little fun getting all the usual suspects to perform on this thread. With Ralph on first violin? LOL

Well I’ve known Roger to pose questions when he wants to find out what someone really knows about a topic. However, in this case I think it’s more of a teaching moment.

If you mean teaching about trolling, you might be right. So far he’s not imparted any knowledge about the RIAA curve and why it exists.


Again, its not like the RIAA curve is a mystery- its easy enough to read about on the Wiki page:https://en.wikipedia.org/wiki/RIAA_equalization
We can easily see that the curve exists independently of what transducer was used to cut the LP or play it back. Now compare the actual RIAA curve with the quote below:

They simply said "lets boost the highs starting at 500 hz, do it for two octaves, stop the boost and continue flat".

Essentially, the quote is simply false.

Now compare the two quotes below:

I am also not talking about the familiar curve which INCLUDES the compensation for a velocity cartridge.

the RIAA engineers cared not if the playback cartridge was velocity or amplitude sensistive.

These two quotes contradict each other. The latter post is correct; the curve exists independently of the transducers involved as I was harping in my two prior posts. Roger has contradicted himself and posted enough misinformation that (since he has made preamps that work), I have to assume something is up that has little to do with this thread.

Its OK. Just ignore that bit, and think about an inverse RIAA box connected to the input of a phono preamp. When you put a flat sweep tone signal in, you get a flat sweep tone out.


For that reason alone, you don't have to know anything about constant velocity to know that the RIAA curve covers about 40 dB over the audio band.

I think all this talk about constant velocity has caused confusion.

You really don't have to know anything about that! The simple fact that an inverse RIAA network at the input of the phono preamp yields flat response is all you need to know. Those inverse RIAA boxes have a slope that covers about 40 db inside- which means that the phono preamp does too.


Put another way:
A microphone does not need to be equalized. If you were to put a stylus on a microphone suspension you would have a mono cartridge. IOW the output of a magnetic cartridge is flat within its mechanical limitations. The RIAA curve is there to reduce surface noise at higher frequencies and to reduce the amount of cutting excursion at low frequencies. It does not take into account the character of the playback device.


You can think about constant velocity if you want to, but as we have seen that led to confusion on this thread. So my advice is don't worry about it, since its irrelevant.

Exist the possibility that we all are rigth and ramtubes too?, no way in the way he posted. Or he is rigth or we are rigth but noth both sides on this dialogue.

Look:

"""
I wish i could draw here but take some paper draw a flat line from 20 to 500 hz. Then go down at 6 dB per octave for two octaves (12 db) then flatten out again. That is RIAA playback without the magnetic (velocity) compensation. For recording just go up 12 dB instead of down. """

Note there: " without the magnetic ( velocity ) compensation ".

@rauliruegas
Maybe this will help to explain: you can connect an inverse RIAA box to the input of a phono equalizer and obtain a flat output if a flat input is applied.  All the inverse RIAA networks I've seen have a range of slightly less than 40dB from 20Hz to 20KHz. So the RIAA curve in the phono preamp must therefore cover the same range plain and simple.

Here is a manual for an inverse RIAA box:

http://www.hagtech.com/pdf/iriaa2.pdf
If you read the manual and look at the curve (page 3), you will see that it covers a range of about 40dB across the audio spectrum. No '...boost the highs starting at 500 hz, do it for two octaves, stop the boost and continue flat...' or any such nonsense. Just the inverse of the RIAA curve. If it was anything else these boxes wouldn't work  :)






 

I respectfully disagree.

In that regard you will find (if you look) that Al is well-respected and is apparently in the habit of knowing that about which he writes. You are the lone dissent in this regard.

Being right or wrong on the internet is one thing- everyone posts their opinions and its hard to tell who's right. Worse, often people do what ever they can to make someone else wrong, as its more important to them to look good more than anything else, and making someone wrong (regardless of the facts) is a cheap and common way to do that.

So its rare when someone comes along that seems to vet his comments for accuracy before posting them. Al is such an individual. 

Lets stop for a moment and imagine, and I think it is valid to do so, that the RIAA engineers cared not if the playback cartridge was velocity or amplitude sensistive. They simply said "lets boost the highs starting at 500 hz, do it for two octaves, stop the boost and continue flat".

Can Ralph and Raul please agree with that simple satement so we can move along?

Apparently not, since the statement in the first paragraph above is false.


To anyone who might doubt what I am saying here, please take a look at the RIAA curve. You can find it on the Wikipedia page if you search on ’RIAA curve’ on google where is the first hit. Here is the link so you can follow along:
https://en.wikipedia.org/wiki/RIAA_equalization

Now look at what Roger states above in the quote at the top of this post. He states ’Let’s boost the highs starting at 500 hz, do it for two octaves, stop the boost and continue flat".’ At the middle of the graph on the page I linked (if you click on it, a larger graphic is displayed), you can see 1KHz as the zero dB point. 500Hz is an octave below that, where the curve initially flattens a bit. But look at how many dB are on the vertical scale on the left. We see that from zero dB, the rise is about 20db to 20KHz. 20KHz is slightly over 5 octaves above 500Hz (an octave is a doubling of frequency). Nowhere in this range does it go flat.


Starting at 500 Hz, about 5.5 dB rise is applied over about 2 octaves and at about 2KHz then 6dB/octave going to 20KHz. At 20KHz the pre-emphasis continues to the limit of the cutter system (although the RIAA did not spec the curve past 20KHz, its assumed to be on a 6dB slope). My cutter system is bandwidth limited to 42KHz; if you play a flat sweep tone into the system from 2KHz to 40KHz and play it back on a conventional turntable with an ordinary magnetic phono preamp, the playback is flat, so apparently other manufacturers of phono preamps, not just me (in the latter case, Tascam) seem to believe that the slope of 6dB/octave above 20KHz should be supported.

I think you all know I am Roger A. Modjeski of Music Reference and was awarded accolades for getting the EQ on my RM-1 preamp to the highest standard of phono EQ at the time in 1978. I am not new to this in any way. I had no idea that this simple concept would be so hard for audiophiles to understand let alone someone else in the industry whose preamp I have measured and found to be off by 7 db above 8 KHZ.

Its perplexing to me that one was able to build a preamp that met the curve but now seems to not know what it is.


Now if you’ve read through this thread, you may have noticed that what Roger stated in the OP of this thread is not what he is stating in the quote above. Both statements were incorrect.


Now Roger also saw fit to attack me with a comment about how just because one has a cutter and uses it does not mean he knows anything about it. Generally, that statement could be true, but ’just for the record’ (if you’ll pardon the expression) when I obtained my cutter system, the lathe was in bits and not functional. I had to build some of the parts that were broken or missing. The electronics were in desperate condition needing complete refurbishment, including repairs due to failing transistors and the like and therefore needed complete calibration once the refurbishment and repairs were completed. You can’t go through a process like that and know nothing about how the device operates (at least one project we mastered was a re-issue and the owner of the label felt our mastering sounded better than the original, so we know it works...). The manual for the system contains the actual RIAA pre-emphasis curve in the appendices- and no surprise it is the very same as shown on the Wiki page and anywhere the RIAA curve appears.

So how Roger is a mile off in his comments and why he started this thread in the first place is a real mystery.


BTW, @clio09 Your comments about Al are un-called for.

Raul, this is easy to sort out.

The RIAA pre-emphasis curve is well-known. You can see it at this Wikipedia page, take a look:
https://en.wikipedia.org/wiki/RIAA_equalization

You will see that the graph is marked in decibels on the left side (vertical axis). The blue line is the pre-emphasis curve and the red line is the complementary playback curve. At 20Hz you can see that the blue line is at minus 20dB. At 20KHz on the other end of the curve its at +20dB. Added together 20+20 is 40; to cover the entire audio range there is a 40dB difference from bottom to top in either record or playback. 

Do you think I'm missing something?

I hope we all agree that the purpose of RIAA EQ is to reduce noise while managing groove size and that the HF boost in recording is 12 db from the bass region to the treble region. Thus the complementary cut is 12 dB for an amplitude sensing cartridge. It is interesting to note that RCA in its 1950s phonograhps had amplitude sensing cartridges and the 12 db RIAA network.

The 40 dB curve that 99.9% of us call RIAA EQ is the combination of RIAA plus the EQ required for a velocity (magnetic) cartridge. Let us not be thinking that the groove itself has 40 dB (100 to 1) of EQ but indeed just 4 to 1.

There was some mention about the EQ of the silent grooves and one best consider they benefit from the same 12 dB of noise reduction. There is no encoding or decoding, just simple 6dB/octave filtering that has its first corner at 500 Hz and second corner at 2,200 Hz.

These statements are false.
The RIAA pre-emphasis curve employs a nearly 40dB rise from 20Hz to 20KHz.

There is about 5 1/2dB rise from 500Hz to 2200Hz. Its not a 6dB slope in that region of the curve as stated in the quote.

the only way the RIAA EQ values that are recorded/generated by the cutter into the record can be duplicated by the phono stage preamp's inverse RIAA EQ - is if the record was played back in the same way it was cut. This is impossible for many reasons.

Anyone care to disagree ?

***************************************

Analog Jack.... is running a customized vinyl setup .

1) He has a tonearm and a low level MC cartridge that has an incorrect anti skate setup.

2) he is running 8 feet (in total) of phono wiring comprised of shielded and unshielded. It has 3 breaks before it gets to the phono stage inputs. The wiring is going at one point through the back of his racks and running parallel to some equipment power cords.  

Does everyone agree that the RIAA EQ signal values generated by Analog Jack's setup will be skewed and contaminated due to the noise being picked up, and incorrect setup of the tonearm / cartridge ?

The phono stage RIAA inverse EQ in the mean time - will continue to do its inverse RIAA based on ......the skewed RIAA EQ values of input.

Garbage in ......

Anyone disagree ?

I think I do.


A bit of mistracking won't affect the curve, neither will a bit of tracking error caused by a radial tracking arm. Funky cables, adding a bit of capacitance, will reduce the resonant frequency that is caused by the interaction of that capacitance with the inductance of the cartridge.


Depending on the cartridge that might or might not affect the playback. But a lower output MC cartridge was specified, so this resonance will not affect the RIAA curve unless the phono section has a sensitivity to RFI being presented at its input. So- depending on the preamp, this bit may or may not affect the resulting EQ curve.


If noise is picked up along the way, too bad, but that won't affect the RIAA curve. It will just be more annoying.

Our preamp's EQ was measured by Stereophile (the MP-1) and it was not off by anything like that. The MP-3 uses the same circuit as that original MP-1. The EQ components are precision matched to well within 1% of Stanley Lipshitz' RIAA article presented to the AES (https://www.pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Lipschitz_Stanley/Lipshitz_on_RIAA_JA....).

An online calculator based on this math:https://www.kabusa.com/riaa.htm
But that wasn't around when the MP-1 phono section was designed.

Some of our preamps are optioned with tape head EQ and depending on the curve used it could easily be off by that much if the wrong input were used.

The manual for the Westerex 1700 cutting system is pretty cut and dried. I keep one copy in my office, another at the studio where the mastering lathe resides.

It lays out how the three main modules of the cutting system work (including theory of operation, schematics, parts list and calibration procedures): the input module, the feedback module, and the cutter amps themselves.

The equalization curve is applied by the input module. No bones about it- its a curve that ranges about 40 db from 20Hz to 20KHz. The actual curve is published in the appendices and is the same curve as linked on the Wiki page (by Erik) prior in this thread.

The signal is then amplified by the cutter amps and applied to the cutter head. Since the head has a mechanical resonance, a feedback winding is provided as an output from the head. This is sent to the feedback module, where it is used for monitoring the cut. The feedback signal is also amplified by the feedback module and then applied to the input module both to kill the resonance in the cutter head as well as to improve channel separation, since the cutter head employs a spring suspension.

At no point in the manual is any mention made of the importance of the kind of cartridge used during playback. Its not relevant. Thus the RIAA curve is defined as I laid out in my initial post. This requires a nearly 40 db range over the audio band. Anything else simply isn’t the RIAA curve.

That is the curve of RIAA plus a velocity (magnetic) cartridge. For a displacement cartridge the EQ is 12 dB. That is what is cut into the record.

This statement is false. The RIAA pre-emphasis covers about a 40 dB range.

Why would you think the audio band is 2 octaves when it is well known to be 10?

I don't. That was the only way to interpret this comment from your OP:

Over this two octave span the EQ boosts the recording cutter's depth putting more signal on the disc. The 3 db rise is called at 500 Hz as one does in any filter. At 2,200 the curve flattens out again and we cut at constant amplitude just as we did below 500 Hz. Over the audio band the total rise is 12 dB. (6 dB x two octaves). 

That was because you put the '6 dB x two octaves' in parenthesis as if to make it clear that you really meant just those two octaves as the audio band to which you referred.

But now you are clarifying that you meant the entire audio band:

figure 1 is for a velocity cutter and velocity cartridge. It is not what is on the record.

'Figure 1' is for the RIAA pre-emphasis and has nothing to do with the cartridge used. That's just what's on the record.

RIAA is just 12 dB. RIAA plus magnetic is 40 dB... Big difference.

During recording (encoding is not a good term, there is no code) the bass frequencies are not modified at all. Above 2200 Hz they are again not modified. Between 500 HZ and 2200 HZ the highs are boosted 12 db so they can be cut on playback.

The proper term is 'pre-emphasis'. It seems to me that you are laboring under a misconception; this entire quote is false. Bass frequencies are EQed to help prevent over-cutting the groove (and also helps with the power needed to drive the cutter head). So the resulting curve covers a nearly 40 db range from 20Hz to 20KHz, as has been pointed out several times in this thread. What gives??

+1

IEC thrown in another enhanced RIAA?

Not really. Its playback only and since the LP is mastered on the RIAA curve, to avoid phase shift in the bass (which results in a loss of impact) the IEC curve really shouldn’t be used. It was withdrawn in 2009.

@clio09
A lot of what has been stated here is correct. The most erroneous statements are actually found in the OP.

Take this one from Al:

That is not my understanding. I have always understood that the RIAA curve corresponds to what is shown in Figure 1 of the reference Erik provided, and varies over the entire audible frequency range in the manner shown in the Figure. Not just over the approximately two octaves between 500 Hz and 2200 Hz. With 1 kHz being the only audible frequency that is neither boosted nor attenuated, and with a total variation of approximately 40 db occurring between 20 Hz and 20 kHz as Erik stated.

This statement is completely accurate. I can see that there might be some confusion about what Roger was trying to say, as I pointed out in my initial post. But compare that to:

Here are the details for the technical folks. Although IEC has thrown in another inflection point this does not affect the main RIAA which is a accomplished by an expansion of groove size starting at 500 Hz . This EQ is then reversed on playback ( accuratey we hope) with the benefit of lowering the noise starting at 500 HZ. Over this two octave span the EQ boosts the recording cutter’s depth putting more signal on the disc. The 3 db rise is called at 500 Hz as one does in any filter. At 2,200 the curve flattens out again and we cut at constant amplitude just as we did below 500 Hz. Over the audio band the total rise is 12 dB. (6 dB x two octaves). Not an insignificant number.

In this statement taken from the OP, there are a coupling of problems.

The 12 db comment is incorrect (take a look at the RIAA curves posted by Erik and see for yourself)- its slightly less than 6 db over those two octaves. He also incorrectly states that the depth is increased- but groove depth is adjustable by the engineer and is governed more by that setting than the excursion of the cutter. For example, we have an older cutter- the Westerex 3D, which cut a lot of the stereo projects in the Golden Era of hifi and was used by RCA and Mercury. To set groove depth, the cutter has two aspects- the cutting weight, which is similar to playback tracking weight (but a lot more) and the groove depth itself, which is set by something call the ’lead ball’ or ’track ball’ which is a tiny semi-spherical gadget that rides the lacquer surface a couple of groove widths before the actual cutter stylus. The tracking pressure is carried by the track ball, and a little adjustment screw determines how deep the groove is from there.

Newer setups have variable groove depth which can be handy. But overall, the groove depth is a thing you set up as part of the recording project- how much modulation will be in the groove (deeper grooves can handle more but less time will be available on the disk).

I’m not sure what the point is here- the opening paragraph suggests at *something* but what that something is does not seem to be illuminated by the statements that follow. The responses to the OP are mostly that of correction, and they seem spot on.

The **depth** of the cutter is not modified by the RIAA pre-emphasis. The **excursion** of the cutter is.

I suspect that readers of this will be confused by this statement:

Over the audio band the total rise is 12 dB. (6 dB x two octaves)

It appears that by 'the audio band' you mean 'these two octaves'.

If that is the case, the *actual* rise in this part of the audio band (500Hz to 2KHz, due to pre-emphasis) is slightly less than 6db. 


So- what is the answer you've been given, and what are you suggesting is the correct answer? Neither is depicted in your OP.