What does 70us and 120us equalization really mean?

[oldvinyldude]

I have been wondering for a while, what these values really mean. I know there are other values associated with open reel recording and playback as well.

I think I understand the general eq idea, in a broader audio sense. For instance the RIAA curve, and how it helps fit more information on a record and lower the noise floor. I understand, I think, that something similar occurs with NAB and IEC equalization curves.

But my question is not about these curves, but perhaps they cannot be excluded in the 70us vs 120us second playback eq discussion....???

What has me confused is that when I think about equalization, I think about + and - values to the frequency in certain areas-ie +12db at 30hz or -6db at 10khz-things like that.

I understand, I think, that 120us is stating a time value. How does this relate to equalization, which is about frequency emphasis/deemphasis? (or is it?)

I fear a smoldering head (mine) will occur when I see the responses, but I will try to hang in there and understand....

Thanks,

Paul

 

[Blue Shadow]

http://stephan.win31.de/tech_en.htm#Entzerr talks about it.

 

[1tumbleweed]

This subject has always made my head hurt, too. Here's some random info I found, may or may not help:

"All tapes have a playback equalization curve that begins a -6dB/octave roll-off at a time constant of 3180uS, or about 50Hz. Ferric oxide (normal/standard) has a flattening out at 120uS or about 1326Hz, while CrO2 (chrome/ferrichrome) and metal have a flattening at 75uS or about 2122Hz."

What I think this means, in plain speaking, is that the rolloff for high bias tapes starts at a higher point than regular tape. Since high bias tape doesn't saturate as quickly as regular tape, it can take an increased recording level - which helps the signal-to-noise ratio. YMMV.

 

[tnmike1]

Blue Shadow: WOW, thanks much for posting that link. That site will answer tons of questions I have on recording and playback. Very helpful

 

[oldvinyldude]

I am still having trouble, but that's to be expected.

Let me see if I have some base information right:

In recording, NAB curve will boost low frequencies, and cut high frequencies. Upon playback, the opposite will occur. Correct?

If so:

I see that at some point along the frequency curve, there will be a threshold where there is greater/less than 3 db modification from flat. This is where the 70us vs 120us difference comes in?

I also note that this 70us vs 120us setting only concerns us with high frequency modifications in the curve?

Low frequency modification seems to have been fixed at 3180us? At least for cassette?

The consensus seems to be that Type II tapes traditionally use the 70us equalization setting (I know that this was not always the case toward the end of the cassette era, but that is a side issue, I think....) The effect of this 70us setting was to move up the frequency band, where boosting of high frequencies would occur?

How am I doing?

 

[Blue Shadow]

In recording, NAB curve will boost low frequencies, and cut high frequencies. Upon playback, the opposite will occur. Correct?

This is the opposite of what happens.

Tape has a high frequency noise, tape hiss, so the signal is recorded at higher levels in the high frequency and reduced on playback.

 

[oldvinyldude]

Well, that makes sense, as a major challenge in tape has been to reduce hiss. I looked up NAB equalization, and got what seemed to be a knowledgeable discussion, that included the attached diagram. Either I am interpreting it incorrectly, or it is mislabeled.

Is this perhaps, showing NAB playback eq?

 

[jancumps]

I see that at some point along the frequency curve, there will be a threshold where there is greater/less than 3 db modification from flat. This is where the 70us vs 120us difference comes in?

I also note that this 70us vs 120us setting only concerns us with high frequency modifications in the curve?

Attempt to give a simple explanation:


If you record the tape with normal frequency (20Hz-20KHz range), the behaviour is very non-linear. It sounds distorted at playback.

The tape only gets linear way up the frequency range (where the wavelength is 70 or 120 us).

So they have built a generator that creates a signal with that high frequency (the bias signal).
On that bias signal, they add your audio signal.

That is what they send to the tape when recording.

During playback, they filter out that bias signal. The remainder is your original signal - but now linear and undistorted.

The 70 - or 120 us bias signal pushes your normal audible frequencies from the non linear range of the tape into the linear range of the tape.

Nah I mixed eq with Bias

 

[Wilhem]

I'm afraid that all of the previous answers confuse a number of issues. I'll post a copy of Terence O'Kelly's BASF Inventor's Notebook No. 3 on equalization to explain everything in detail. In summary, changes in magnetic flux across two poles create electrical current just as changes in current across two poles of an electro-magnet create magnetic flux. When a tape plays back, the flux changes for low frequencies change at a slow rate and produce little current or output. Higher frequencies produce more current--up to a point. To balance things out so that sound of equal loudness produce equal current, they have to be "equalized." Low frequency output is boosted and high frequency output is reduced. (The concept is the same for LP playback, but the physics are different.)

The reduction in high frequency output quits at a point where the current from tape begins to drop off for several reasons. For regular ferric cassette tape, that point is around 1,326 Hz. The circuitry that deemphasizes output at lower frequencies than that now allows the output to naturally rise again. For tapes that have higher coercivity and greater short wavelength capacity, the changing point is 2,274 Hz. Because the turnover point of 2,274 Hz is higher in frequency than 1,326 Hz, it reduces accompanying noise by 4.7 dB--the chief reason for the higher turnover point.

The changes balancing or equalizing the current are done through a combination of resistors and capacitors in the playback equipment. In engineering terms, multiplying resistance and capacitance results in a figure known as a "time constant" since all other parameters cancel themselves out in the multiplication process. The two time constants for 1,326/2,274 are 120/70 microseconds. This is significant to physicists and electronic engineers but has caused a lot of confusion for consumers. That's why the IEC decided to call the settings for bias and EQ selection either Type I, II, III, or IV.

There is also some equalization applied to the incoming signal before it reaches the record head. This is known as record EQ or record pre-emphasis. It is one of the last adjustments a technician makes in setting up a tape recorder's electrical values. Mechanical adjustments come first (azimuth, height, zenith, and so forth); then proper settings for accurate playback EQ; bias settings for a balance between output and distortion; meter settings; and then record EQ for flat frequency response at the already select bias point.

The Inventor's Notebook will explain everything better, with graphs that make sense.

 

[oldvinyldude]

I love this "Last edited by jancumps; Today at 02:42 PM. Reason: I was wrong"
:lmao::lmao:

May I ever be as humble.....?

 

[1tumbleweed]

I love this "Last edited by jancumps; Today at 02:42 PM. Reason: I was wrong"
:lmao::lmao:

May I ever be as humble.....?

Well, better him telling himself than us.

Before things get any more out of control:

BIAS is a signal applied ONLY during recording. It does nothing on playback. It varies from tape formulation to tape formulation.

EQUALIZATION is applied during recording and playback; playback equalization is the inverse of recording EQ. It's constant across tape TYPES, but not formulation.

 

[Wilhem]

Recording EQ is not the inverse of playback. It is merely a boost in the higher frequencies to assist high frequency output for flat, extended response. Playback EQ is compensating for increasing output for flux changes as the wavelengths grow smaller, not for compensation of recording alterations.

 

[oldvinyldude]

Recording EQ is not the inverse of playback. It is merely a boost in the higher frequencies to assist high frequency output for flat, extended response. Playback EQ is compensating for increasing output for flux changes as the wavelengths grow smaller, not for compensation of recording alterations.

Another assumption (that record eq and playback eq were inversions of each other) dispelled-Thanks Wilhem

 

[Wilhem]

I hope these come out clearly enough. What is not covered completely is the reason for the roll off of high frequency output, but that is a function of a number of things, including thickness losses and other factors. These bulletins were written for BASF customers and consumers in a relatively simple language for non-engineers to understand, so they do not cover material as deeply as engineers or afficianados would like. (Some of O'Kelly's later writerings on optical discs and flash memory had sections demarked for much deeper technical information that casual readers could skip if they preferred. Maybe he can be persuaded to rewrite these in the same way to add more detailed information.)

 

[oldvinyldude]

Thanks again Wilhem

I will re read these pages several times. Even with their simple language, I get to a point where what I thought I understood a paragraph or two earlier becomes cloudy again.....

The main initial point that I was totally unaware of was the inherent higher energy contained in higher frequencies as they pass by the tape head during play, necessitating compensation.

There's a reason that gifted minds brought to bear on a technical challenge will find ingenious solutions to problems that lesser minds did not even know existed.

The good news is that we all enjoy the fruits of their talents :yes:

 

[eddisc]

searching: I found THIS and probably won't be back for a day or two...
http://archive.org/details/gd1975-06-17.aud.unknown.87560.flac16

 

[1tumbleweed]

Recording EQ is not the inverse of playback. It is merely a boost in the higher frequencies to assist high frequency output for flat, extended response. Playback EQ is compensating for increasing output for flux changes as the wavelengths grow smaller, not for compensation of recording alterations.

So, all those years reading High Fidelity and Stereo Review and Audio, and I've still got it wrong?

I'll be in the corner, curled up in a ball...:tears:

 

[Wilhem]

Don't be so hard on yourself. Audio, High Fidelity, and Stereo Review never covered the topic of EQ very deeply, although Ralph Hodges of Stereo Review did cite Inventor's Notebook Number 3 that discussed bias. It's easy to get lost in all the technical information, especially if one does not have a background in some applicable subject. Once the technical director at Fuji wrote a small article about tape noise for one of the magazines and got the explanation of Dolby NR all mixed up. The editor could not follow his written discussion and sent the draft to me to check. I rewrote it and sent it back to the editor. The Fuji guy was a friend of mine--as was the editor--and to this day I don't believe he ever knew I rewrote his article.

The technical people were generally like that in the tape industry. The sales and marketing people were the ones who disliked each other.

 

[jan_stevns]

.......The technical people were generally like that in the tape industry. The sales and marketing people were the ones who disliked each other.

And quite often the technicians wondered why the engineers made things the way they did.

We used to tell the engineers there are 3 kinds of people , those who can count, and those who can't

 

[perryinva]

That playback is the inverse of record EQ, is often cited, I believe, because of course, that is the way Dolby B works! Easy to get mixed up. In the 70s as a teen, I always wondered why designers didn't just incorporate NR and EQ in to the same circuit...funny now...