Bias, EQ, HXPro, Dolby, Chrome tapes with 120 us EQ

[Late Boomer]

Hello, a new user here! Got some questions regarding tape recording, would be nice if someone in the know could explain whatever I am not getting

-- Bias

So bias is, basically, a sort of a shakeup to ensure that low-amplitude (or low-frequency? or both) signal is recorded sufficiently well. Bias frequency is pre-set and cannot be changed. Lower and higher bias means lower and higher amplitude, not frequency. Lower bias - reduction in low freqs, higher bias - reduction in high freqs. Ferric vs chrome: chrome requires higher bias, hence marking on tapes "high bias". Is it possible to have chrome tapes with low bias or ferric with high bias? I guess it is an academic question, in any case, it is set for recording only and knowing about it is not needed for playback.

But then again, why some pre-recorded Time-Life tapes with Chrome tape are marked "Normal Bias"? How this information can be used on playback? Did they actually want to say "Normal tape" or "120 us Equalization", so users would play this tape on a ferric setting?

-- HX Pro

Adjusts bias depending on amplitude (or frequency? or both?) From the user's standpoint nothing to bother with: either a deck has it or not. The result is better output levels without overdrive. No special equipment needed for playback, so a win-win feature.

-- Dolby NR (particularly Dolby B)

Compresses higher frequencies. What is the threshold frequency above which they are compressed? Does it have a sort of wedge shape, so the higher the frequency the more compression (at given amplitude)? If played on a non-Dolbyized equipment requires reducing treble, though many prefer extra treble. Even when reduced, the treble will still be compressed, so the tonality in high freq region will not be the same as with Dolby decoder.

I noticed that I can switch my walkman into Chrome mode to bring down the highs, so it kinda-sorta works for ferric Dolby tapes, but nothing can be done for Chrome Dolby tapes. I am trying to figure out what exactly is happening here, when I switch the player to Chrome tape mode, it uses 70 us EQ, which means it continues de-empasizing frequencies in 1.3K-2.3K range, which 120 EQ does not do. But Dolby affects much higher frequencies, so switching the walkman into Chrome mode suppresses midtones but not so much the highs, where Dolby is active? I guess this is why even when switched into Chrome mode, the music sounds dull yet still ringing at the same time. So, a player either needs a tone control or a true Dolby decoder to play Dolby tapes.

-- Equalization

I've read Terence O'Kelly's bulletin #4 on EQ. He writes, "the output increases at a rate of 6dB per octave" but then stops increasing starting at rather low 1.3K. So, from which freq is it increasing? From 100 Hz? In this case it would be about four octaves before it stops increasing, and all this in a "telephone" range (I think it was defined as 300-3000 Hz). Then output flattens out, then I guess it decreases, so he talks about record EQ later to flatten out the output. The Fig.1 shows natural increase in output (1A), which I presume is recorded on tape as is, and then mitigated by playback EQ (1B) to give flat response, at least in the "telephone" range. So, if the increase in output happens naturally and depends on the magnetic properties of the tape, should not it be labelled "Ferric" and "Chrome", not 120 us and 70 us? On the other hand, the playback EQ is what the playback deck is doing, and here I can see why it is labelled with 120 and 70. Now, adding record EQ to this, it starts amplifying the signal for higher freqs starting from which freq? This is just to overcome natural drop-off, so it is not compensated by the player.

If the recording facility uses Dolby, then the recording EQ occurs before adding Dolby?

So, when tape improved over 1970s and 1980s and 1990s, I suppose it got closer to older Chrome tape in terms of MOL and noise, but the playback correction remained frozen in the late 1960s, so newer ferric tape sounds fuller and brighter with the same 120 EQ setting. I remember seeing the frequency graphs on cassette packaging, showing a hump for "new" tape in the above 10K range.

Regarding 120 us and 70 ms, for me as a user it would be easier to understand that cut-off frequencies of 1.3K and 2.3K than some numbers with units in microseconds. I understand that this is the result of some R-C circuit calculation, but for a non-engineer in this area frequency values, I think, are easier to understand.
I saw cassettes with Chrome tape that say "120 us Equalization", this is the setting that must be used for playback, it does not necessarily mean that it was used for recording? On a deck with separate bias and EQ settings, the EQ setting is not used during recording, because EQ is what happens during playback, as explained in the tech bulletin, right? If it works during recording, than this is recording EQ by definition, which affects the top end, and should not therefore be labelled as 70 us or 120 us? But if EQ switch is for playback only, how is it different from Fe-Cr switch, why having both? Confused. Chrome tapes labelled as "120 us EQ" are just regular Chrome recordings then? I could as well listen to any Chrome tape in 120 us mode, that is, in ferric mode to get extra punch? On an auto-sensing player I would have to tape over the chrome hole in the cassette. The extra 4dB that BASF engineers talked about, would be gained mostly in the midrange, around 2K? Wilhem said, "Playing back a tape recorded for 70-microsecond playback at 120-microsecond equalization will boost high frequencies by about 4.5 dB". Um, how high are these "high frequencies"? If EQ happens in the below 3K Hz range, then I would not call it high, I would call it mids.

-- good ferric tape sounds better than chrome tape.

Many component deck owners admitted that they preferred using ferric tape because the recording sounds brighter than with chrome tape. Is this caused by 120 us EQ preset? I thought that chrome tape at 70 us EQ and ferric tape at 120 us EQ should sound about the same, except chrome tape has less noise. Or maybe in the late 1960s, when 70 us was standardized for chrome tape, ferric tape was much worse in terms of MOL and noise compared to how it improved by 1990s, while chrome tape did not improve that drastically. Ferric tape improvement was so significant that when used with old 120 us EQ it results in markedly better MOL and better frequency response, kind of skewing the 120 ferric vs 70 chrome equation towards ferric.

How good is 1990s ferric tape compared to early 1970s chrome tape and comparing to early ferric tape with Dolby? Has this improvement affected labels like Capitol to drop their XDR process? I know that it was mostly quality control initiative spun up by their marketing department, and their tape was not very good to begin with, but it included Dolby NR. I have their 1988 Meddle tape, the third Canadian release. The previous release was marked SDR, the one I have shows neither SDR/XDR nor the Dolby mark, but sounds great not only in terms of the sound range and hiss, but also in terms of wow and flutter.

I was surprised to find out that a basic 1990s Sony HF tape recorded on an early 1980s Technics/Panasonic deck without Dolby NR turned out very well, with both meaty low end and bright high end, and the noise at the reasonable sound level (one that I can tolerate while sitting in front of my speakers, not being in another room) was almost unnoticeable.

 

[CSCHOEN]

Just my 2 cents here. regarding a few of your questions: From my experience, if you use high quality tape (Maxell XLII, TDK SA, for example,) there is not much need for any "noise reduction" (Dolby, etc.)
Dolby was developed before tape quality became good enough to not have much "noise" to reduce. Also, the times that I had used Dolby, I found it cut out some of the clarity of the recording.
The best results with cassette tapes can be achieved by using high quality tape, using a deck which is clean and properly aligned, and in good working order. Record levels can be set a little higher than
one might expect when using good quality tape, as there is a fair amount of head-room. Having a manual bias adjustment is a good feature also, and is something that can be experimented with a little
to achieve better recording/playback quality. I happen to have a decent deck and a few cases of blank tape, and do enjoy recording analog sourced vinyl to tape occasionally. But were it not for that,
I would probably not have muck interest in the medium, as it is somewhat clunky (in my opinion). I'm surprised there is any interest in it actually.

 

[CSCHOEN]

Just my 2 cents here. regarding a few of your questions: From my experience, if you use high quality tape (Maxell XLII, TDK SA, for example,) there is not much need for any "noise reduction" (Dolby, etc.)
Dolby was developed before tape quality became good enough to not have much "noise" to reduce. Also, the times that I had used Dolby, I found it cut out some of the clarity of the recording.
The best results with cassette tapes can be achieved by using high quality tape, using a deck which is clean and properly aligned, and in good working order. Record levels can be set a little higher than
one might expect when using good quality tape, as there is a fair amount of head-room. Having a manual bias adjustment is a good feature also, and is something that can be experimented with a little
to achieve better recording/playback quality. I happen to have a decent deck and a few cases of blank tape, and do enjoy recording analog sourced vinyl to tape occasionally. But were it not for that,
I would probably not have muck interest in the medium, as it is somewhat clunky (in my opinion). I'm surprised there is any interest in it actually.

 

[Late Boomer]

Thanks for replying. I am just interested in some things purely from theoretical/informational standpoint. Let me distill my previous questions into a bulleted list.

• Is it a mistake that pre-recorded Time-Life cassettes loaded with Chrome tape were marked "Normal Bias"? Even if the tapes were recorded with normal bias, how this information could be used during playback?
• Terence O'Kelly's bulletin #4 on EQ says, "the output increases at a rate of 6dB per octave" but then stops increasing at about 1.3 KHz. From which frequency is the output increasing?
• Terence O'Kelly's bulletin #4 on EQ shows the graphs of 6dB output increase (Fig. 1A) labelled as "120" and "70". Should it say something like "ferric" and "chrome" instead, considering that this curve is a result of natural properties of tape, not of applied equalization?
• Has the difference in noise and MOL between ferric and chrome tape narrowed from the late 1960s by the late 1980s?
• The inlay for pre-recorded Time-Life cassettes says: "This digitally remastered cassette has been duplicated at 32:1 on BASF chrome tape using normal (low) bias and 120-microsecond equalization, and the Dolby HX Professional headroom extension (for improved high-frequency response) and Type B noise-reduction systems. For best sound quality, use normal (IEC Type I) bias and equalization settings on your tape player and switch Dolby B circuit on." Did it make sense to record on chrome tape with normal bias? I thought the whole point of high bias is to "shake up" more coercive chrome tape, so what is the point of recording chrome tape with normal bias? How I can use "normal bias" on playback, if bias is strictly a record-time setting? 120 us equalization used during recording follows BASF engineers' idea to get extra 4dB from chrome tape, with noise quashed by Dolby B, so this I can understand.

 

[vwestlife]

Is it a mistake that pre-recorded Time-Life cassettes loaded with Chrome tape were marked "Normal Bias"? Even if the tapes were recorded with normal bias, how this information could be used during playback?

Yes, that appears to be a misworded direction for the user to play the tape at the normal 120 µS Type I equalization setting, which most decks gang together with the bias adjustment for recording. So what they're really saying is that if your deck has a combined bias/EQ switch, make sure it is set to the normal bias/EQ position. (If your deck has separate bias and EQ switches, you only need to worry about the EQ switch, because the bias switch is irrelevant for playback.)

If they actually did try recording onto chrome tape with normal bias, the result would sound weak, shrill, and distorted, even with the correct EQ setting. Cobalt-doped ferric Type II high bias tape (usually marketed as "for chrome position" even though it is not chrome tape) is more tolerant of incorrect recording bias, but it would still sound excessively bright.

 

[oldvinyldude]

I think the marketing\merchandizing folks often write these things, lacking a true understanding of the subject. The primary "miss" is their confusion about Bias and equalization. As these are prerecorded tapes, just ignore anything about high or low Bias during their duplication.

Has the difference in noise and MOL between ferric and chrome tape narrowed from the late 1960s by the late 1980s?

I think it's safe to say that this is true-but this might not be true of lower quality 80's tape being compared with top quality 60's era ferric tapes. I would expect better results from Maxell's early Ferric tapes, than late 80's generic tape sold by mass merchants sourcing from lowest cost suppliers. (Three "chrome" cassettes in a poly bag for $1.00 was a common sight in the 80's)

 

[vwestlife]

Generally speaking, from the '80s onward, you can get a reasonable idea of how good a tape is by how dark the tape is. Standard, budget-grade Type I ferric tape is rust brown, premium "super ferric" Type I tape is dark brown, and chrome or cobalt-doped Type II high-bias and Type IV metal tape is jet-black. I've seen many pre-recorded cassettes which have black tape even if the label or J-card makes no mention of it being chrome or cobalt.

 

[Wilhem]

1. Time-Life cassettes stated "Normal Bias" because so many inexpensive recorder/players had few settings, and the typical setting for playback of most common tapes was labeled "Normal Bias." This is technically incorrect because bias is not used at all for playback, only for recording. The duplicators who produced Time-Life cassettes used the higher bias required for chrome to record these cassettes.
2. The output begins to increase as soon as the magnetic prints run across the gap in the playback head. The more the flux changes from a positive print to a negative print, the greater the output; so low frequencies change less frequently and, therefore, produce less output. The long wavelengths of the low frequency prints vary significantly from one head design to the next, so it is not possible to state a precise frequency for all decks; but most decks can produce 25-30 hertz reasonably well.
3. The increases in output shown in Figure 1A in "The Inventor's Notebook, Bulletin 4" might better be explained as "ferric" or "normal" and "chrome" as long as the reader associated the increases with the decreases in Fig. 1B. These Fig 1A curves are not a natural property of tape. They are a natural part of electro-motive forces--the more flux changes, the more output. These Figure 1A curves are not recorded onto tape, contrary to what a lot of people think. The rollover points vary by the coervcivity of the tape--cobalt-doped Type I tapes will roll over at a higher point than that shown in the graph, and that means that the record pre-emphasis can be reduced a bit to gain more SOL. The playback curve, however, is a fixed value to assure compatibility from tape to tape.
4. Constant improvements in materials and processes reduced the noise in ferric tapes and increased the MOL for chrome tapes from the late 1960s to the late 1980s. Double-coating processes and mixtures of oxides made even greater improvements from the late 1980s to the mid-1990s. By that time the best Type 1, II, and IV tapes were measurably different but not much different in audible terms. The most apparent differences were in modulation noise and print-through effects.
5. This Time-Life claim is definitely incorrect. The recording was done on the correct bias level for chrome tape. This bias point would be higher than that used for ferric tape. The record pre-emphasis would be lower than that for ferric tape. I suspect that Julius Konins of Cassette Productions in New Jersey recorded these cassettes on his own custom-built equipment that ran at 32:1 speeds. Mr. Konins was an electronic and recording genius who produced many of the best quality recordings throughout the 1980s.

 

[Late Boomer]

These Figure 1A curves are not recorded onto tape, contrary to what a lot of people think.

Could you elaborate on this? For example, this book shows remanent magnetic flux on Fig 1.8, which is flat from about 50Hz to say 1300Hz (a familiar value!) then falls off. How come it is flat - when recorded on tape? They say that "-6dB/octave fall in output, with increasing frequency, is needed to compensate for the increasing output during replay of a constant recorded signal"?

Ah, I maybe start getting how it works: it is constant amplitude ON TAPE, but when tape moves along the head, the flux generates electric current in the head, and the higher the frequency (at the same recorded amplitude), the higher the current. It is THIS CURRENT which increases 6dB per octave. "Output" is the current that comes from the head, not something coming directly from tape.

On a second thought, is it possible to measure what is recorded on tape at all? I guess, you have fresh tape with randomly charged particles. Then you magnetize it and count what share of the particles have oriented themselves - the more particles have lined up, the higher the "amplitude" on tape. But as this cannot be done directly, this is done indirectly by moving tape along a head, and then what is measured is the head current... I probably should go back to MP3s

 

[Taketheflame]

I've seen many pre-recorded cassettes which have black tape even if the label or J-card makes no mention of it being chrome or cobalt.

I've come across quite a few of those myself. It's always a pleasant surprise when a pre-recorded unexpectedly turns out to be cobalt/chrome tape

To the OP - while I can't answer everything from a technical perspective, I'll give input where I can.

Regarding tape bias: Yes, the tape "bias" is effectively fixed for normal/type 1, and high bias/type 2 as they are formulated to be used with specific microsecond EQ settings (70/120). Many pre-recorded tapes that used type 2 tape were duplicated at 120MS EQ in order to have the highest playback compatibility among consumer playback decks (as not all machines can adjust between 70 and 120 MS, even if it's a pretty standard feature on any most component/deck type machines). So yes, any "normal bias" or "120MS EQ" labeled chrome tape should be played on the normal bias/120 MS EQ setting.

Regarding HX Pro: IIRC, it basically adjusts the bias according to the frequency response being recorded. Type 1 tapes generally don't handle high frequencies as well as Type 2/4 tapes, so it's probably most beneficial for Type 1 (although it benefits all tape types).

Regarding Dolby: IIRC, Dolby B (and C) mainly effects frequencies from 5Khz and up (well above where most of the music "information" is, but where hiss may be more prevalent).

Regarding Type 1 vs Type 2 tape: It's very hard, if not impossible to truly compare 1990s (and newer) ferric tape against early chrome formulations. Not only is it going to be much harder to find a sealed/unused chrome tape from the 1970s than from later years, word on the street seems to suggest that early true chrome (i.e - not cobalt based type 2) may suffer from age-related problems with recording sensitivity (again, I haven't put this to the test - it's what I've read over on the Tapeheads forum). I've found that the quality of the tape deck you're using may matter more than the tape type itself. A really good deck can make fantastic recordings on Type 1 tape IMO.

 

[Wilhem]

Ah, I maybe start getting how it works: i

Correct.

... the tape "bias" is effectively fixed for normal/type 1, and high bias/type 2 as they are formulated to be used with specific microsecond EQ settings (70/120). Many pre-recorded tapes that used type 2 tape were duplicated at 120MS EQ in order to have the highest playback compatibility among consumer playback decks (as not all machines can adjust between 70 and 120 MS, even if it's a pretty standard feature on any most component/deck type machines).

I'm not sure what you mean. The bias level is "fixed" only if the user selects a switch on the face of the recorder and does not open the machine to adjust the levels internally. The more sophisticated equipment allowed users to adjust bias by making adjustments on the face of their equipment. There is no machine I know that would allow anyone to "adjust between 70 and 120 microseconds." That makes no sense. Type II tape recorded for 120-microsecond EQ playback is what we suggested, not for compatibility, but for improved SOL values that could closely match Type IV. The only reason that chrome tapes appeared with 70-microsecond EQ playback was to reduce tape hiss because no one at Philips or BASF suspected that Dolby B NR would catch on as it did. Once it was clear that Dolby B would be a success, limited SOL was a greater problem than tape hiss. But Advent 201 incorporated BOTH Dolby B NR AND 70-microsecond EQ, and there was no going back until I convinced A&M Records to try the 120-microsecond playback standard on the Police's Synchronicity. The popularity of that release and the inability to duplicate Type IV tape at high speed led convinced the music industry to release chrome cassettes the way we wanted in 1970.

 

[Late Boomer]

Wilhem, thanks for your invaluable insight and explanation. I have a tricky question for you regarding chrome tapes with 120 EQ. I think it was you who responded to those saying they prefer recording with Dolby and listening without it, that what they like may differ from what an engineer would consider correct. Which is, what goes in must come out. If flat signal comes in, flat must come out, and for this reason listening to Dolby-recorded tapes without Dolby is incorrect.

Now to 120 EQ chrome tape. If I understand correctly, these EQ values depend on tape properties - signal recorded to chrome tape continues to increase 6dB per octave farther than 1.3kHz all the way to 2.3kHz. It is this increase that 70 EQ compensates on playback. In this case, would not listening to a correctly recorded chrome tape produce brighter, and therefore incorrect, sound?

Another option I see is de-emphasizing 6dB per octave increase in 1.3-2.3 kHz range DURING RECORDING. This recording EQ would be different from other recording EQ that lifts the very high frequencies, and this sort of setup cannot be achieved on standard consumer equipment, unless machines that have separate tape/bias and EQ controls do just that. (Still cannot understand why having separate type/bias controls, considering that bias depends on tape type).

In this case, when listening with 120 EQ, the play deck will take care of the output increase in 50Hz - 1.3 KHz, and the 1.3-2.3 kHz would not sound emphasized because it has been corrected during recording. But in this case I don't see where the purported 4dB or increased dynamic range come from: if you record duller signal on tape, the relative share of noise in the overall signal during playback will higher, not lower?

On the other hand... If I wanted to record louder signal than I would normally do, this louder signal would be "compressed" to the regular SOL signal (figuring the difference between SOL and MOL is my other issue), and then it would be "expanded" because 120 EQ will not touch frequencies higher than 1.3 KHz. So, the 4 dB increase can be achieved only to frequencies above 1.3 kHz? Which means, that if one wanted to achieve higher dynamic range using this approach, the record will sound brighter?

So, one way or another, chrome tape recorded for 120 EQ will sound brighter, which is, um, incorrect?

Now, all these chrome 120 EQ tapes came with Dolby NR? What did the recording companies think about people playing these tapes? The owners of hi-fi component decks that had Dolby and chrome switches preferred to record their own cassettes and did not bother with pre-recorded tapes. On the other hand, those who did buy prerecorded tapes listened to them on cheap boomboxes and later on a walkman, most of which did not have Dolby. Oh, yes, the inlay did recommend to reduce treble response on non-Dolby machines, but most walkmans did not have bass/treble adjustment! So your refrain that 70 EQ for chrome tape was a mistake, and 120 EQ + Dolby was the way to go seems to me unsubstantiated by the market. Even if/when cheap machines had Dolby, so many things could go wrong that the effectiveness and correctness of Dolby could not be achieved. No wonder that whenever a question about Dolby comes up, the first reply is usually "make sure your deck is serviced and adjusted to spec".

 

[Wilhem]

...signal recorded to chrome tape continues to increase 6dB per octave farther than 1.3kHz all the way to 2.3kHz. It is this increase that 70 EQ compensates on playback. In this case, would not listening to a correctly recorded chrome tape produce brighter, and therefore incorrect, sound?

The signal increases only on playback because the flux changes across the head gap create current. At about 2.3 kHz the increase begins to fall off, so the playback EQ--which has decreased the output up to this point--is slowly removed to allow the output to "level off" the response. The end result is a flat frequency response that extends slightly farther than that for ferric tapes played back at 120 microseconds.

increase in 1.3-2.3 kHz range DURING RECORDING

What increase? I think you don't fully understand what is happening. You refer to this band as being "emphasized." It is not. Chrome recorded for 120-microsecond playback will have a frequency response just as flat as a ferric tape 20 dB down from the operating level, but the response will be slightly extended to frequencies not likely to be audible. At operating level, the frequency response of chrome will be 4.7 dB higher than that for ferric tape at 10-kHz because SOL increased when 4.7 dB of noise is sacrificed.

Still cannot understand why having separate type/bias controls, considering that bias depends on tape type.

If tape manufacturers had complied with IEC standards to set a bias point for the official reference tapes, a fixed setting would have worked. They did not, so different formulations using different particles and different tape thicknesses required tweaks to bias settings to get the lowest distortion for the respective tape.

So your refrain that 70 EQ for chrome tape was a mistake, and 120 EQ + Dolby was the way to go seems to me unsubstantiated by the market.

I did not say it was a mistake. It was a choice of preferring low noise over SOL. Advent added two features: new EQ plus Dolby B, and consumers preferred two features to a single feature. Most consumers recorded popular music at excessively high levels, and greater SOL would have been an advantage there. Music cassettes were a popular medium for those with good to excellent quality equipment at home, and the boomboxes and Walkman players were mobile devices.

 

[Late Boomer]

So there are at least three signals involved: (1) input, (2) whatever is "stored" on tape according to its "response", (3) output from the head. You are saying that if the same stuff is "stored" on tape, then there will be no difference when it plays back at the same EQ, no matter ferric or chrome. I get it. Both tapes will induce 6 dB increase per octave in the play head, and the major difference will be in very high frequences, say 10K and above, due to higher coercivity. Higher coercivity of chrome tape means (1') it needs to have stronger bias when recording, (2') the tape can "store" higher level of signal, (3') and it better retains charges close to each other, which means it extends into higher frequencies when playing the tape back. Nothing about noise at this point?

O'Kelly's document says "the first ferric oxide tapes made for cassette recorders began to lose their 6dB/octave increase at a relatively low frequency" with transition point at 1.3KHz. The he says, "The invention of chromium dioxide tapes significantly improved cassette high frequency response ... and its higher coercivity allowed it to accept higher levels of high frequency energy." You say, "At about 2.3 kHz the increase begins to fall off." — Is is for chrome tape or for both kinds of tapes?

If both ferric and chrome tape sound the same with the same playback EQ, then using 70 EQ for chrome tape means reduced middle frequencies from 1.3 KHz and above, is this correct? In this case chrome tapes will sound duller? But you are saying that "chrome recorded for 120-microsecond playback will have a frequency response just as flat as a ferric tape 20 dB down from the operating level." Huh? Is "operating level" important here? I don't know what it means, all I see is "as flat as a ferric tape." Anyway, you say that "at operating level, the frequency response of chrome will be 4.7 dB higher than that for ferric tape at 10-kHz." Um, so with the same 120 EQ chome tapes will sound brighter? First, because of extra 4.7 dB at higher frequencies, and second because 120 EQ does not correct for chrome's 6 dB per octave increasing going farther. But you said it does not...

You can see, I am lost ) Sadly, O'Kelly's doc does not help much.

EDIT: My other issue with O'Kelly's doc: it says, "...the playback head loses its ability to resolve short wavelengths due to spacing losses by 3 dB at 1,326.3 Hz, and that is why the downward equalization in FIG. 1B is removed at that point." I suppose he meant that EQ stopped increasing not "removed", and remains at the same level until some point when high frequencies (give or take 10K I guess) start to drop off naturally, this is when EQ should be gracefully reduced in accordance with high frequencies drop off.

EDIT2: Found a relevant discussion on tapeheads (is it ok on this forum to post links to other forums? Wilhem, you seem to be a member of that forum too): http://www.tapeheads.net/archive/index.php/t-23741.html

A.N.T. says, "switching from the standard 70us to 120us EQ gains about 8dB in the maximum output at 16kHz," which is understandable to me. He says, "there is a considerable gain in HF headroom", which to me is the same as the recording sounds brighter, thus not flat, thus not correct.

You said there, "if the tape was recorded for 70-microsecond playback, then flipping the playback switch would boost the high frequency content by 4.5 dB--including the noise." This corresponds to what I understand, but it also means that there IS a difference when recording whether for 70 or 120 EQ playback. This is what I am trying to understand - what this difference entails? Does it mean that input signal is different, but what is "stored" on tape is the same as ferric, or does it mean that input is the same, but what is stored is different? But it should not be different, because it should be played with the same 120 EQ.

Aha, you wrote, "To use chrome tape for 120-microsecond playback you must use the higher bias level used for chrome and reduce the pre-emphasis in the record stage." - here! Reduce the pre-emphasis in the record stage. This is what I alluded to when I was talking about 1.3K to 2.3K range and its "de-emphasis", because pre-emphasis means amplifying, and it should not be, because it already amplifies naturally. I thought it should be attenuated. Ah, "reduction of pre-emphasis" effectively means attenuation, I guess I am getting somewhere.

A.N.T. said, "To simplify the concept, can I state I can record type II tapes at 120us by building a box containing an amplified (gain 4.5dB) high shelving filter with a zero at 120us and a pole at 70us and connecting it on the LINE IN jacks of the tape deck?" which you - or someone very much like you - did not contest, so I suppose it is correct. Still, as a a noob in many things electronic, I may not understand what this phrase exactly means, I guess it means that the incoming signal should be amplified by 4.5 dB starting from 120us (that is, 1.3 kHz) to 70us (2.3 kHz) and then stayed at this level ("shelved"). This corresponds to what I was talking about, although I thought that it should be attenuated (reduced), not amplified, because it is sort of amplified naturally because of how chrome tape works (what O'Kelly said by "The invention of chromium dioxide tapes significantly improved cassette high frequency response").

 

[Taketheflame]

Correct.


I'm not sure what you mean. The bias level is "fixed" only if the user selects a switch on the face of the recorder and does not open the machine to adjust the levels internally. The more sophisticated equipment allowed users to adjust bias by making adjustments on the face of their equipment. There is no machine I know that would allow anyone to "adjust between 70 and 120 microseconds." That makes no sense. Type II tape recorded for 120-microsecond EQ playback is what we suggested, not for compatibility, but for improved SOL values that could closely match Type IV. The only reason that chrome tapes appeared with 70-microsecond EQ playback was to reduce tape hiss because no one at Philips or BASF suspected that Dolby B NR would catch on as it did. Once it was clear that Dolby B would be a success, limited SOL was a greater problem than tape hiss. But Advent 201 incorporated BOTH Dolby B NR AND 70-microsecond EQ, and there was no going back until I convinced A&M Records to try the 120-microsecond playback standard on the Police's Synchronicity. The popularity of that release and the inability to duplicate Type IV tape at high speed led convinced the music industry to release chrome cassettes the way we wanted in 1970.

You've explained it far better than I could. I was basically just trying to tell the OP why his pre-recorded chrome cassettes said "normal bias" or "120ms" on them (i.e - they're duplicated @ 120 microseconds for the reasons you state, and meant to be played as such).

I was under the impression (prior to your post) that the main reasons for doing so were to avoid any playback issues on early machines that didn't allow for switching between 120 and 70 microseconds EQ (i.e - "Normal" and "Chrome" settings), not other reasons like the SOL.

 

[Wilhem]

"there is a considerable gain in HF headroom", which to me is the same as the recording sounds brighter, thus not flat, thus not correct.

Most of the discussion in the "Inventor's Notebook," my explanation, and Alex's posts discuss capacity, not the influence on an incoming signal. Boosting MOL or HF headroom gives the tape more capacity to store, not influence, signals. You are confusing the two. The signal goes into the amplifier, gets adjusted on the low end and on the high end to compensate for artifacts induced by the head and for some tape deficiencies, and the adjusted signal is sent to the head (along with a bias signal.) The magnetic particles passing under the gap in the record head react to the changing signal, and a record of those changes is made as the tape passes under the head.

On playback, the magnetic "prints" on the tape induce a current in the head as they pass back under it. Double the speed, and the current increases. The more the prints change, the more current they produce--up to a point. A recorded 1kHz signal creates more output than a 400 Hz signal. To compensate for this imbalance on playback, equalization is added to reduce the increase in output so that it remains level for all frequencies--up to that point I mentioned. (If you chose to eat soup while the Titanic were sinking, you would have to tilt--equalize--the bowl so that the soup would remain level.) Those points I mentioned are when the tilting down playback equalization is removed. For ferric Type I tapes the compromise turn over point at which the downward tilt was removed was 1,326.3 Hz. (I say "compromise" because a point at a higher frequency would mean more pre-emphasis/less SOL.) For Type II tapes the turn over or transition point is 2,273.6 Hz. This choice meant less tape hiss but required more pre-emphasis--which Type II tapes could take better than Type I tapes at the time. BASF engineers would have preferred to keep the same playback EQ as that for Type I tapes so that pre-emphasis could be reduced, SOL increased, and noise levels handled by Dolby B NR.

I was under the impression (prior to your post) that the main reasons for doing so were to avoid any playback issues on early machines that didn't allow for switching between 120 and 70 microseconds EQ

Marvin Bornstein, the head of quality at A&M, understood the argument for 120-microsecond playback because he was aware of all the half-speed mastered vinyl LPs that audiophiles were buying at the time. Those LPs had much greater dynamic range than before--at least at the beginning of the records. I wrote a technical paper explaining the advantages that he presented to upper management. They didn't get it. I rewrote a much shorter version that touted the greatest benefit as lower cost from duplicators because they didn't have to stock a second type of C-zero. The same housing could be used for chrome tapes as for ferric tapes. (We gave a rebate directly to A&M to offset the extra cost of chrome tape. This was an industry secret) So it wasn't compatibility for consumers or even greater SOL that sold management--it was lower inventory cost from the duplicators.

 

[Late Boomer]

I still don't get the pre-emphasis thing. The 6dB per octave increase stop on ferric tape at 1.3 kHz, but continues on chrome tape up to 2.3 kHz. Does it come "naturally" because of tape composition, because chrome particles magnetize differently than ferric? If yes, why pre-emphasis is needed? Or does this mean that chrome tape can withstand extra boosting of input signal in 1.3-2.3 range - pre-emphasis - which will turn into this extra 6 dB / octave boost, and then it is compensated during playback when EQ is set to 70? But you said that this 6 dB/octave boost depends on frequency, not amplitude of input signal?

So, when recording to chrome tape for 120 EQ, the input signal is not pre-boosted (pre-emphasized) and the 6 dB/octave increase stops at the same 1.3 kHz point as for ferric, but because chrome tape can take more boost, this offers extra 4 dB or so for ultra-loud notes?

 

[Wilhem]

Does it come "naturally" because of tape composition, because chrome particles magnetize differently than ferric?

Yes, chrome, ferric-cobalt Type II, and metal tapes have greater coercivity than ferric tape, meaning it is harder to "coerce" the particles to switch their magnetic poles. That means that it is harder for them to change back, too; so the smaller wavelength prints on tape are more likely to survive. Ferric tape does not automatically begin to drop its output at 1.3 kHz, and chrome does not do the same exactly at 2.3 kHz. Those points are merely reasonable points of decrease matched by the simple R-C playback networks described as 120 and 70 microseconds. Those EQ values remove the 6 db/octave tilt downward so that the output levels even out to reasonably flat response. However, there is still a drop off of high frequencies that needs to be made up. and the high frequency boost (pre-emphasis) is the kick needed to get response beyond 8 kHz or so where there is most of the high frequency energy.

The problem with that boost--a good amount for ferric tape and a greater amount for Type II tapes--is that it pushes the high frequency energy beyond the capability of the tapes. That's why frequency response at -20 dB can measure flat to 18 kHz but at the operating level, the "0 mark on a deck's meters in simple terms, response begins to fall off before 10 kHz. The higher coercivity of Type II tapes meant that they could take the greater boost than ferric tapes and still have almost the same response at the operating level as ferric tapes if the playback of Type II tapes used 70-microsecond EQ. The advantage was that allowing the playback EQ to drop response to a higher frequency reduced tape hiss by 4.7 dB. All analogue design for tape is a series of compromises, and in the early days, the biggest problem with cassettes was tape hiss, not limited SOL.

Then came Dolby B NR that reduced noise by 10 dB. That shifted the compromise to the point that noise was now less of a problem than limited SOL, so we chose to sacrifice 4.7 dB more noise to gain 4.7 dB of SOL at a time when "audiophile" recordings were pushing dynamic range and little compression. (That situation is completely different today. Digital technology allows flat response at the maximum level with very little noise for great dynamic range--and music is compressed to almost no dynamic range.)

 

[Late Boomer]

Wilhem, I am re-reading this thread and the other relevant ones. So, chrome tape does not provide lower noise than ferric per se, but it allows to boost the level above the level of ferric tape, thereby increasing S/N ratio, which effectively means lower noise at the same sound level. 70 μs EQ ensured that "high" (that is, above 1.5-2K) frequency response would be about the same as for ferric tapes, but the higher coercivity of chrome tape meant that the top end did not have a dramatic fall-off at, say, 10K or so and would extend further.

So,
* chrome tape recorded with 70 μs EQ sounds "brighter" that ferric at 120 μs EQ thanks to the preservation of the very top end;
* the lower noise can be achieved by increasing recording level. With the same recording level as for ferric tape, there is no noise improvement, correct?

For a lay user, without measuring response of a particular tape, by how much chrome tape can be pushed further compared to ferric tape?

If I want to have as broad frequency range as possible, even at the cost of noise, I should strive to record at -20 dB, not at 0 dB? On another hand, because I will be cranking up volume during playback, I will hear more hiss, and I may want to fight it with Dolby. But for Dolby to work properly, should not I record at Dolby level, which on my deck is marked to the right from 0 dB? Then I would lose my top end. I am in a conundrum here.

 

[Wilhem]

So, chrome tape does not provide lower noise than ferric per se

Chromium dioxide formulations have less noise than ferric tape by their very nature. It's due to the shape of the particle and the ability to align the particles better when coating.

A chrome tape should not sound brighter than a ferric if the equalization and bias are set correctly. The chrome tape should sound quieter and, if the musical content has a lot more high frequency energy, more accurate because of less saturation. Frequency response should be extended a bit more than that for ferric tape, but that extension is not likely to be noticeable unless one's hearing extends well beyond 15 kHz and the recorded content has information in that range.

You can't lower noise by increasing recording level. Noise stays the same. You can increase dynamic range by increasing the recording level as long as you stay within the signal-to-noise ratio of the tape, which is also a fixed value. If the recording level is the same as that for ferric tape, chrome tape has about 4.7 dB less noise just because of 70-microsecond playback EQ and another 1.5-3.0 dB less noise even at the same 120-microsecond playback EQ.

Meters on cassette decks typically display peak values, not average values. Some meters are weighted so that high frequency peaks get more meter boost than low frequency peaks because low frequency MOL is higher than SOL. If one records at 0 on a peak meter where that value represents 200 nWb/m (the Dolby NR mark), the "average" level of musical content with wide dynamic range is about -20 dB. That recording level is OK for most tapes, but too low for a Type IV metal tape that can take a lot more signal. On the other hand, popular music is too often so compressed that there are no musical peaks, so recording at a slightly lower level brings the average energy to about -10 dB with lower distortion than would be heard if that content were recorded at the zero mark on a meter. Open-reel meters are generally measuring average levels because the faster speeds provide more headroom for the tape. There is a big difference in reading cassette meters and open-reel meters.

There is more information in Alex Nikitin's site here: http://www.ant-audio.co.uk/Tape_Recording/Library/Meters.pdf