RIAA EQ curve software emultation?

Archived from groups: rec.audio.pro (More info?)

play on <playonAT@comcast.net> wrote:
>The guy who posted asking about how to get LPs recorded into a PC got
>me thinking... shouldn't there a way to use software EQ emulate the EQ
>curves for old records, such as the RIAA curve, and also the older
>non-standardized 78 RPM EQ curves? Are there plots of these curves
>anywhere? If so it would seem a simple matter to copy the curve with
>digital EQ if you knew what it looked like.

The manual for the Millennia Media phono preamp has a list of the standard
LP curves and some suggested beginning points for various 78 labels and eras
and for broadcast transcriptions. It's very incomplete but it gets better
with every new edition. So much of the stuff on 78s needs to be done by
ear, even electrics. Acoustics have to be done entirely by ear because they
are all over the place.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Archived from groups: rec.audio.pro (More info?)

play on wrote:
> The guy who posted asking about how to get LPs recorded into a PC got
> me thinking... shouldn't there a way to use software EQ emulate the EQ
> curves for old records, such as the RIAA curve, and also the older
> non-standardized 78 RPM EQ curves? Are there plots of these curves
> anywhere? If so it would seem a simple matter to copy the curve with
> digital EQ if you knew what it looked like.

The EQ in audacity.sourceforge.net has a dozen curves that plot nicely
on the screen when selected. I don't know enough to have an opinion on
whether they are right or not, but they look like someone put some time
and effort into getting them right.

--
Phil Nelson

Archived from groups: rec.audio.pro (More info?)

If you are already using a preamp with the standard RIAA EQ, then you
would want the digital EQ to EQ the error (difference) between the
curve you have (standard RIAA) and the curve you want. Should work
pretty well if you know the curve you want.

Mark

Archived from groups: rec.audio.pro (More info?)

On 2 Mar 2005 17:49:01 -0800, "Mark" <makolber@yahoo.com> wrote:

>If you are already using a preamp with the standard RIAA EQ, then you
>would want the digital EQ to EQ the error (difference) between the
>curve you have (standard RIAA) and the curve you want. Should work
>pretty well if you know the curve you want.

Right. I was originally thinking that if you knew these curves that
perhaps you would not have to use a phono premamp at all to get the
audio in, just use a conventional mic pre and/or DI. Although perhaps
that isn't a good idea.

Al

Archived from groups: rec.audio.pro (More info?)

On Wed, 02 Mar 2005 21:09:44 -0800, play on <playonAT@comcast.net>
wrote:

>Right. I was originally thinking that if you knew these curves that
>perhaps you would not have to use a phono premamp at all to get the
>audio in, just use a conventional mic pre and/or DI. Although perhaps
>that isn't a good idea.

This issue comes up here pretty frequently lately, and (fortunately)
nobody thinks it's a good idea.

Let's start with a coupla problems:

1. Conventional phono cartridges only give predictable response into
a unique loading (47K ohms and small capacitance) and definitely *not*
into a microphone input.

2. The EQ is a beotch. The defined part of the curve is 40 dB. And
there's (well, maybe we should wait for Scott's comments on current
practice, but still...) more stuff above 21.22 KHz that nobody wants
to talk about.

3. No, that's enough for starters. Phono equalizers are roughly equal
to mic preamps in the small-signal-welter-weight-class tough guys.
Right behind D/A summing junctions, by small margins.

Chris Hornbeck
"If what the player is giving you is good, you can hardly miss. If it's
not, there's nothing you can do with a mic that will make it right."
-Mike Rivers

Archived from groups: rec.audio.pro (More info?)

On Thu, 03 Mar 2005 05:48:58 GMT, walkinay@thegrid.net (hank alrich)
wrote:

>play on wrote:
>
>> The guy who posted asking about how to get LPs recorded into a PC got
>> me thinking... shouldn't there a way to use software EQ emulate the EQ
>> curves for old records, such as the RIAA curve, and also the older
>> non-standardized 78 RPM EQ curves? Are there plots of these curves
>> anywhere? If so it would seem a simple matter to copy the curve with
>> digital EQ if you knew what it looked like.
>
>That's the easy part, and I think it's been done. The hard part is
>getting software to properly load the output of a phono cartridge.
>There's nothing virtual about that chunk of the signal chain.

I do have an excellent APT Holman preamp that has some switchable
loading features for the phono input so I guess that's definitely the
way to go. I haven't tried playing any 78s thru it yet.

Al

Archived from groups: rec.audio.pro (More info?)

> > The guy who posted asking about how to get LPs recorded into a PC got
> > me thinking... shouldn't there a way to use software EQ emulate the EQ
> > curves for old records, such as the RIAA curve, and also the older
> > non-standardized 78 RPM EQ curves? Are there plots of these curves
> > anywhere? If so it would seem a simple matter to copy the curve with
> > digital EQ if you knew what it looked like.
>
> That's the easy part, and I think it's been done. The hard part is
> getting software to properly load the output of a phono cartridge.
> There's nothing virtual about that chunk of the signal chain.

There are several hard parts. There are some people selling flat-response
phono preamps; they provide proper loading for moving-magnet or moving-iron
cartridges, and flat response; you do the EQ in your computer, adjusting as
needed.

Here's the problem: The pre-emphasis that has been applied to the LP or 78
or 45 has been done with analog equipment. It affected the frequency
response in the desired way, but it also affected the phase response.
De-emphasizing with the same curve (RIAA if that's how it was recorded)
affects the phase response in a complementary way, so you come out with
proper phase response. At least you're supposed to. (You don't really,
because the cutting head probably wasn't linear group delay, and the
cartridge probably isn't either. But at least the possibility is there.)

When you do the EQ in the computer, however, it *doesn't* typically apply
complementary phase shift along with the frequency changes, so your phase
response is now haywire.

Building a competent phono preamp isn't that hard, notwithstanding the fact
that a remarkable number of manufacturers manage to fail at the task. Radio
Shack's little unit at least comes close to getting the RIAA curve right.
You can build something fairly decent out of a couple of 5532s and a pair of
passive EQ networks between them, plus four 18V batteries. Better results
with a better power supply, still better results with better amplifiers, but
the 5532s do a surprisingly good job. [Non-commercial commercial] I wrote an
article in the Dec. 2001 issue of audioXpress giving details of such a
passive network, one that can be used for RIAA or, with switching, non-RIAA
curves. It works between just about any flat amplifiers, with a few caveats.
[/Non-commercial commercial]

Peace,
Paul

Archived from groups: rec.audio.pro (More info?)

play on <playonAT@comcast.net> wrote:
>On Thu, 03 Mar 2005 00:55:20 GMT, Chris Hornbeck
><chrishornbeckremovethis@att.net> wrote:
>
>>On Wed, 02 Mar 2005 15:40:30 -0800, play on <playonAT@comcast.net>
>>wrote:
>>
>>> They use a term on there that I'm
>>>not familiar with for the low end, which is "turnover"... what does
>>>that mean, exactly? A certain type of boost, or a crossover point?
>>
>>It's the frequency that the curves approach asymptotically. For
>>example, the RIAA curve has 50 Hz, 500 Hz and 2122 Hz corners.
>>Because the latter are close enough to interact, the response
>>is not exactly 3 dB up or down.
>
>OK, I'm having a little trouble with this... what determines the shape
>of the EQ at those frequencies?

Any filter is specified by three factors:

1. The turnover frequency
2. The amplitude
3. The slope of the filter (Q).

The RIAA filter is a complex filter with two poles, each of which have
specified turnover, amplitude, and slope. SOME folks add a third pole
in for rumble reduction.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Archived from groups: rec.audio.pro (More info?)

Chris Hornbeck <chrishornbeckremovethis@att.net> wrote:
>Let's start with a coupla problems:
>
>1. Conventional phono cartridges only give predictable response into
>a unique loading (47K ohms and small capacitance) and definitely *not*
>into a microphone input.

And SOME of them want different unique loading than that, which is why
better preamps have adjustable loading. Loading changes the damping of
the stylus to a varying extent; for the most part the more efficient a
cartridge is, the more it will be sensitive to load variations.

>2. The EQ is a beotch. The defined part of the curve is 40 dB. And
>there's (well, maybe we should wait for Scott's comments on current
>practice, but still...) more stuff above 21.22 KHz that nobody wants
>to talk about.

Nothing leaving my shop has anything up there other than surface noise,
but to be honest the way a phono system reproduces surface noise has a lot
to do with how it sounds. If you have an arm that is ringing and a preamp
that causes intermodulation products of ultrasonic trash to appear in the
passband, you will find the noise is a lot more annoying than with a more
accurate system.

There are records out there with DMM or half-speed mastering that can give
you accurate ultrasonic content and there are some cartridges still that can
reproduce it well. But even without it, it's still important to either
accurately reproduce or to harmlessly remove it because there is so much
noise being produced up there.

Check the Shure S15 test record for some pretty amazing ultrasonic content,
and things like square waves that really are square. That was done by
Don Grossinger on the DMM lathe at Europadisk. I wish I could do that.

>3. No, that's enough for starters. Phono equalizers are roughly equal
>to mic preamps in the small-signal-welter-weight-class tough guys.
>Right behind D/A summing junctions, by small margins.

Modern electronics have made this a lot easier, though. With solid state
parts it's easy to make a phono preamp with 60 dB of gain on the front end,
entirely passive EQ networks, and then make-up gain afterward, and still
not have offensive noise from the preamp. The fact that the high frequency
noise of the front end is reduced in the RIAA network helps a lot. The guys
who came up with the whole pre-emphasis idea really hit on something
brilliant.

The thing is that with the RIAA network, the preamp front end needs to have
a tremendous dynamic range, and it needs to be very quiet on the low end.
Because there is so much ultrasonic trash, the preamp has to either reproduce
it accurately or eliminate it harmlessly. Unfortunately this is precisely
the sort of thing that puts a strain on an A/D system....
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Archived from groups: rec.audio.pro (More info?)

"Scott Dorsey" <kludge@panix.com> wrote in message
news076tn$fa3$1@panix2.panix.com...

> The thing is that with the RIAA network, the preamp front end needs to
have
> a tremendous dynamic range, and it needs to be very quiet on the low end.
> Because there is so much ultrasonic trash, the preamp has to either
reproduce
> it accurately or eliminate it harmlessly. Unfortunately this is precisely
> the sort of thing that puts a strain on an A/D system....

It's not really that hard; with quiet amplifiers, you can split the gain
between the amplifier before the RIAA network and the one after it, and not
have to work so hard on dynamic range. ~30dB gain in each amplifier
(depending on the sensitivity of the cartridge) and you're fine.

Peace,
Paul

Archived from groups: rec.audio.pro (More info?)

"joe at salerno dot com" <captaincapstan@yahoo.com> wrote in message
news:bJuWd.2039$oO4.1244@newsread3.news.pas.earthlink.net...
> Search for DCart or DC-Live. I forget the URL, but this program has
> conversion presets.

The new version is DC-SIX. DC-Live is still around too.

www.tracertek.com

Peace,
Paul

Archived from groups: rec.audio.pro (More info?)

"Richard Crowley" <rcrowley7@xprt.net> wrote in message
news:112m79lc66p1384@corp.supernews.com...
> Isn't there a problem with the very great dynamic range of pre-
> equalized RIAA curve signals? Last time this was discussed
> here the consensus seemed to be that there is not practical way to
> make a decent (quiet, linear, etc.) flat preamp to allow the RIAA
> curve to be done in software.

Naah; you can do it with a 5534. Tracertek even sells one for that purpose,
although I don't know what it uses for electronics. It's not all that hard.

The worst signal you ever get on records is not the music, but the
scratches; the biggest banger I ever measured on a disc was 26dB over the
standard 5cm/sec level, or 100 cm/sec. (It was a jumping scratch) That
measured -18dBu from my cartridge, which is about average output
(1mV/cm/sec). So you build a preamp with enough gain that it boosts the
signal just enough that a scratch of that magnitude doesn't clip into your
A/D converter, and run it into the computer, where you descratch it, then
apply whatever curve you need to. A 5534 and a decent A/D will have all the
dynamic range you need to be way quieter than any disc surface ever made.
Heck, so will a decent tube circuit.

The obstacle isn't in the preamp, but in the not-right phase response of a
computer-generated EQ curve.

Peace,
Paul