This continues my evaluation of old cassette tapes, this time a lot of BASF (true) chromes. The tapes were all used, from my own collection and well cared for, and from unknown origin, often less well cared for. All tapes were erased in a Weircliffe BKE200 prior to testing.
The tests were done on a modified Nakamichi BX-300 with known playback response. The test procedures are explained in detail here.
The deterioration of chrome dioxide
It became clear in recent times that true chrome dioxide tape does not age well. While existing recordings can still be played without any problems, making new recordings on old tape now is subject to significantly degraded magneto-acoustical properties.It appears that this ageing process drastically reduces the tape's sensitivity, and more so in the treble than in the midrange. This can be compensated with harder driving during recording, which increases distortion. The lacking treble sensitivity has to be redressed with reducing bias, which again increases distortion, reduces Maximum Output Level in the bass and midrange, but increases Saturation Output Level in the treble region.
As a result chrome today, be it new old stock or used, requires a lower bias, and returns a MOL significantly worse than when the tape was freshly-manufactured: often no more than 0dB (ref. Dolby level). Chrome's distortion now exceeds 1% starting from 6-7dB below the MOL point, and rises smoothly with level. This differs from ferro-cobalt, which has a lower distortion until 3-4dB below MOL, then rising quickly (also see here). Also chrome tape now exhibits a lot of compression when approaching MOL. So the sound character between chrome type II and ferro-cobalt type II is bound to differ markedly, more so than ever before.
While I only have a handful of tapes to test, I know from others that this ageing is not uniform: the same tape type and age can end up with different properties due to differing storage conditions, even when the amount of mechanical wear and tear is similar. But they all go down in quality: I have not seen a single chrome tape that is still close to its original performance.
As expected the same phenomenon can be seen with CrO2-based 'EE' reel-to-reel tape. An investigation can be found in this thread at the tapeheads.net forum.
All of this seems bad news, but at least the breed's low and smooth-sounding bias noise has not suffered. In fact there are indications that bias noise drops thanks to ageing. I still have to investigate this in some detail, but if true this would be good news: the reducing noise would help compensate somewhat for the age-reduced MOL. This way chrome tape can still offer an acceptable dynamic range, around 60dB and of the same order as late-70s and early-80s ferro-cobalt tapes like SA and XLII.
How to handle chrome tapes?
What this means for the tape enthusiast is that it is probably better to record on chrome tape today than tomorrow: we don't know, yet, if the ageing goes on, or if it flattens out after a while.It goes without saying that due to the aberrant sensitivity and low bias needs a calibrating deck is mandatory. Chrome's low bias noise can only be exploited on a deck with a low self-noise, which is a function of the head material, head quality, and the replay amplifier: not something to be taken for granted.
Since chrome tape now has a lot of compression, record level calibration should happen near 0dB and not at -20dB. This is especially critical if Dolby is to be used with anything resembling success. The old cognoscenti trick of using chrome with 120us equalisation, and not 70us, no longer works. Back in its prime chrome tape was SOL-limited, not MOL-limited, and using 120us improved treble saturation at the cost of noise. Nowadays chrome is MOL-limited, and using 120us will only increase noise.
The best-suited music programme is low in level most of the time, with only isolated loud peaks. These peaks should be kept at, or even below, 0dB (Dolby level or DIN).
How do we know this, actually?
Now, how do we know that this really is an effect of ageing, and not the way it always has been? After all, there are numerous stories from the past of how chrome tapes behave differently from the regular Japan-style cobalt-doped ferrics.The answer is simple: magazines in the 70s, 80s, and 90s regularly tested cassettes, and some of these magazines did this to a very high standard, also disclosing their measurement methods. Comparing the historical published results of e.g. BASF tapes to wide-spread type IIs like SA and XLII shows that the true chrome tapes generally had a lower bias noise, a wider span between MOL400 and SOL10k, and often a sensitivity and a MOL quite close to SA and XLII, and thus a dynamic range of the same order. These days these relationships no longer hold: we find lower sensitivity, lower MOL, a smaller span between MOL400 and SOL10k, a lower dynamic range, and a reduced bias point. This is how we know that at least one of these tape classes has drifted significantly under the pressure of time. And since measurements of SA and XLII today still are broadly consistent with past magazine articles, we know that it is the chrome tape that has degraded. There is an elaborate example of this further down this article.
Another apparent source of information is the original Nakamichi 1000's promo folder. It contains frequency curves recorded on chrome tape at -20dB, -10dB, and 0dB relative to Dolby level. These curves are directly comparable to my curves below. You can easily see the fundamental differences at -10dB and 0dB, both in the nature of the treble rolloff and in the amount of compression (i.e. the vertical distance between the curves). And while Nakamichi might have doctored these curves to look as good as possible, it would be a far stretch indeed to assume that they modified a deck in order to perform as good as shown above with tapes with a natural performance as shown here below: this would have turned the deck into something entirely non-standard. And it still would not have cured the compression ...
You can find more details on chrome instability in this Tapeheads thread and reviews of tapes in this thread. And finally there are these June 2020 and February 2021 statements from T.O'Kelly, erstwhile BASF USA duplication tape sales manager and a staunch advocate of chrome...
Now to the results ...
Maxell XLII 90 (1994) (my calibration reference)
Relative sensitivity: (reference)
THD @ Dolby level: 0.62%
MOL400(THD=1%) : +1.8dB
MOL400(THD=3%): +5.3dB
MOL1k(THD=3%): +5.5dB
SOL10k: -3.4dB
Bias noise: -54.8dB, -59.2dB(A)
Dynamic range: 64.5dB
I repeat here the results for the tape my Nak BX-300 is calibrated for in type II position, the 1994 XLII. All following results take this as reference.
BASF Chromdioxid (1979)
Relative sensitivity: -3.5dB
THD at Dolby level: 6.5%
MOL400(1%): -10.6dB
MOL400(3%): -3.3dB
MOL1k(3%): -3.7dB
SOL10k: -7.8dB
Bias noise: -55.2dB, -59.2dB(A)
Dynamic range: 55.9dB
This is BASF's 1977 single layer model, but my sample had a 1979 date code. Today its performance is awful, dynamic range worse than many a budget ferric. But the flatness and consistency of the frequency curves, and the lack of drop-outs, are testament to BASF's manufacturing prowess.
BASF Chromdioxid II (1981)
Relative sensitivity: -2.9dB
THD at Dolby level: 3.9%
MOL400(1%): -8.1dB
MOL400(3%): -1.2dB
MOL1k(3%): -1.8dB
SOL10k: -7.3dB
Bias noise: -55.6dB, -60.0dB(A)
Dynamic range: 58.8dB
I was lucky to find a few of these in very excellent condition, almost as if they popped out of a time capsule. As often with tusk-equipped shells winding was chattery, but apart from this, and apart from the inevitable chrome degradation, performance was impressive for something 40 years old.
BASF CR-E II (1985)
Relative sensitivity: -3.0dB
THD @ Dolby level: 3.5%
MOL400(1%): -8.1dB
MOL400(3%): -0.9dB
MOL1k(3%): -1.9dB
SOL10k: -8.0dB
Bias noise: -56.0dB, -60.8dB(A)
Dynamic range: 59.9dB
BASF CR-E II (1986)
Relative sensitivity: -3.2dB
THD @ Dolby level: 5.0%
MOL400(1%): -10.1dB
MOL400(3%): -2.3dB
MOL1k(3%): -3.1dB
SOL10k: -7.4dB
Bias noise: -56.0dB, -61.0dB(A)
Dynamic range: 58.7dB
This tape should be identical to the 1985 issues, but despite looking very good time has done its worst here. MOL is very low, compression very high, as can be seen clearly in the varying distance between the frequency plots in the 100-400Hz range.
BASF Chrome Extra II (1991)
Relative sensitivity: -2.2dB
THD @ Dolby level: 2%
MOL400(1%): -4.9dB
MOL400(3%): +2.0dB
MOL1k(3%): +0.8dB
SOL10k: -6.5dB
Bias noise: -55.7dB, -60.6dB(A)
Dynamic range: 62.6dB
BASF Chrome Standard II (1991)
Relative sensitivity: -1.7dB
THD @ Dolby level: 2%
MOL400(1%): -4.1dB
MOL400(3%): +1.8dB
MOL1k(3%): +1.0dB
SOL10k: -5.5dB
Bias noise: -54.5dB, -59.0dB(A)
Dynamic range: 60.8dB
A budget product, this is supposed to be the single-layer Chrome Extra II tape in a cheaper shell. There is ageing, but it is not that bad. The assembly date code puts this cassette just after mid-Summer 1993.
BASF Chrome Extra II (1992)
Relative sensitivity: -1.5dB
THD @ Dolby level: 1.7%
MOL400(1%): -3.4dB
MOL400(3%): +2.5dB
MOL1k(3%): +1.6dB
SOL10k: -6.0dB
Bias noise: -54.6dB, -59.2dB(A)
Dynamic range: 61.7dB
This sample is from 1993, and measures the same as the model here below, so it is likely cobalt-enriched.
BASF Chrome Extra II (1993)
Relative sensitivity: -1.1dB
THD @ Dolby level: 1.5%
MOL400(1%): -3.3dB
MOL400(3%): +2.6dB
MOL1k(3%): +1.5dB
SOL10k: -6.6dB
Bias noise: -54.6dB, -59.0dB(A)
Dynamic range: 61.6dB
BASF developed an 85%/15% chrome/ferro-cobalt tape for duplicators, Chrome Plus, aiming at increased MOL while keeping bias noise still low (see further below). From the second half of 1993 this tape was also introduced into the single-layer consumer models. Due to a transition period and the many different versions of BASF basic type II tapes that were then circulating around the world's different markets it is hard to know whether any given sample is this fabled Chrome Plus or not.
MOL and sensitivity of the one tested here are slightly better than every other chrome tested. At first I did not find this convincing proof for the presence of cobalt, especially as the j-card proclaims 'linear directed chrome'. I had to revise this later, because we know that CP got into the wild by end of 1993, while my sample's date code puts it in the first half of 1995. Moreover, the measured parameters are very close to these of the next two tapes in this list, which were added much later, both confirmed CP types.
BASF Chrome Super Quality II (1995)
Relative sensitivity: -1.1dB
THD @ Dolby level: 1.5%
MOL400(1%): -2.4dB
MOL400(3%): +3.2dB
MOL1k(3%): +2.7dB
SOL10k: -5.6dB
Bias noise: -54.6dB, -59.3dB(A)
Dynamic range: 62.5dB
This is a budget product sold by the LIDL chain stores. I expected a ferro-cobalt like 1995 Chrome Super II or like the next-generation Super Quality II, but this one confounded me with a set of measurements pretty close to Chrome Extra II. So this CSQ II is also one of the elusive chrome-ferro-cobalt hybrids.
Emtec CP Extra (1998?)
Relative sensitivity: -0.9dB
THD @ Dolby level: 1.4%
MOL400(1%): -1.9dB
MOL400(3%): +3.5dB
MOL1k(3%): +2.8dB
SOL10k: -5.1dB
Bias noise: -54.4dB, -58.8dB(A)
Dynamic range: 62.3dB
This tape was added to the collection in February 2021. I purchased it as part of a lot of newly-made alignment tapes. The documentation identified it as CP Extra duplication tape, a tape that in recent years also was loaded by NAC (771) and Delta Media. As such we know for sure that this is chrome-ferro-cobalt.
BASF Chromdioxid Super (1980)
Relative sensitivity: -1.9dB
MOL400(1%): -7.4dB
MOL400(3%): -0.5dB
MOL1k(3%): -3.1dB
SOL10k: -7.2dB
Bias noise: -56.4dB, -61.3dB(A)
Dynamic range: 60.8dB
This is BASF's 1977 dual layer model, but my sample had a 1980 date code. The imperfect crossover between the low-frequency layer and the high-frequency layer causes the typically dished response. This cassette pre-dates IEC standardisation, coming from an era when Europe aligned to DIN C401R.
BASF Chromdioxid Super II (1980)
Relative bias: -3
Relative sensitivity: -3.0dB
THD @ Dolby level: 4.7%
MOL400(1%): -9.2dB
MOL400(3%): -2.4dB
MOL1k(3%): -4.5dB
SOL10k: -7.5dB
Bias noise: -56.5dB, -61.7dB(A)
Dynamic range: 59.3dB
One would expect this to be the same or similar than the other 1980 model seen above, now under nominal IEC alignment. However, time seems to have been very unkind to this sample.
BASF Chromdioxid Super II (1982)
Relative sensitivity: -1.9dB
MOL400(1%): -6.9dB
MOL400(3%): 0dB
SOL10k: -5.8dB
Bias noise: -55.5dB, -60.2dB(A)
Dynamic range: 60.8dB
BASF CR-S II (1985)
Relative sensitivity: -2.3dB
THD @ Dolby level: 3.0%
MOL400(1%): -6.5dB
MOL400(3%): 0dB
MOL1k(3%): -1.8dB
SOL10k: -7.0dB
Bias noise: -56.2dB, -61.4dB(A)
Dynamic range: 61.4dB
Again a dual-layer tape. In its time it was tauted to be of absolute top class.
BASF Chrome Super II (1988)
Relative sensitivity: -2.5dB
THD @ Dolby level: 2.5%
MOL400(1%): -5.7dB
MOL400(3%): +0.9dB
MOL1k(3%): -1.0dB
SOL10k: -7.3dB
Bias noise: -56.0dB, -61.1dB(A)
Dynamic range: 62.0dB
BASF Chrome Super II (1989)
Relative sensitivity: -2.2dB
THD @ Dolby level: 3.1%
MOL400(1%): -7.3dB
MOL400(3%): -0.1dB
MOL1k(3%): -2.0dB
SOL10k: -7.6dB
Bias noise: -56.5dB, -61.9dB(A)
Dynamic range: 61.8dB
Impressively flat for a dual-layer tape, but otherwise nothing to write home about. But let's make a detour ...
Here is a review from Hi-Fi Choice January 1991. It discusses exactly this tape. In this period HFC performed the reviews with Nakamichi Dragon, CR-7, and CR-4 aligned to the three IEC reference tapes. All performance measurements were done at the IEC operating points, i.e. non-optimal for deviant tapes.
The review uses the DIN/IEC standard as signal reference, i.e. 0dB here is 250nWb/m, 1.2dB above the 0 level of my own measurements. Observe how the review puts MOL at +3.5dB DIN (=4.7dB Dolby Level). Sensitivity relative to the then-valid IEC II reference tape is -0.3dB, in other words almost on-standard. HFC find noise at -59.4dB(A) ref DIN zero, or -58.2dB(A) ref Dolby Level, resulting in a dynamic range of 62.9dB. (I measure a lower noise, but let's not forget that a recent insight suggest that chrome's noise drops also due to ageing.)
Now have a look at the 1988 TDK SA. My own measurements are here, and here is the HFC review from April 1989:
And here is the 1991 review of 1990 SA:
At that time SA was nearly-equivalent to the IEC II standard U564W. Hi-Fi Choice found SA to be only 0.6dB more sensitive than CSII, while the tapes' MOL and dynamic range were almost the same.
Today, however, my measurements show that SA has 3.1dB more sensitivity, 6.1dB more MOL, and 2.6dB more dynamic range. These are large differences, and incontrovertible proof that something has happened to the chrome tape.
Let's take another approach. In all of my measurements series I set bias to obtain a flat frequency response at -20dB. This is the usual approach in cassette recorders, but that does not make it the correct approach. Ideally bias is set for an optimal balance between distortion at low frequencies and saturation at high frequencies. The resulting frequency reponse, whatever it looks like, then should be flattened with record equalisation. But most cassette recorders, including the BX-300, have only variable bias and fixed equalisation.
Some people say that with a different equalisation chrome would perform better, i.e. show a higher sensitivity and a higher MOL. This can be tried with a fixed-eq deck too, simply by setting bias for the criterion of choice, and then measuring how much equalisation would have to be added at the recording amplifier to flatten the response.
Well, let's try this. Increasing bias resulted in this:
Relative sensitivity: -2.3dB
THD @ Dolby level: 1.8%
MOL400(1%): -4.1dB
MOL400(3%): +2.5dB
MOL1k(3%): +0.5dB
SOL10k: -11.8dB
Bias noise: -56.8dB, -62.0dB(A)
So yes, MOL improved significantly, although it still did not reach the levels reported in the 1989 test. However, at the same time sensitivity did not move, and SOL dropped dramatically, to -12dB. You can imagine the resulting frequency responses:
For flattening these responses a matching record equalisation would need +5dB at 10kHz, and over +10dB at 20kHz. That is a lot. But remember that due to the reduced SOL at this high bias point the tape is already saturating in the treble. The added equalisation would make this even worse. Clearly, this operating point is unworkable.
Even then there will be doubters, people insisting that chrome tape always was something special, fundamentally different in its care and feeding from ferro-cobalt type IIs, and unsuited to be exploited properly by Japanese decks and/or Philistines.
It is true that when chromium dioxide appeared on the scene in the early 1970s it suddenly posed new problems to the recorder manufacturers. Compared to ferric tape, bias current had to double and record current had to be raised by 3dB. Heads had to be found that could pass this all without saturating. Initially there appeared many recorders on the market that were not quite optimal for chrome: these decks underbiased the tape to avoid head saturation, resulting in flat frequency responses (thanks to matching record equalisation) but low headroom/MOL and high distortion. This gave rise to the notion that budget decks were not compatible with chrome.
Confronted with generally unsatisfactory results when using CrO2 tape, the industry responded with two solutions.
1) Ferrichrome tape saw the light of day, operating at a bias only slightly higher than ferrics, while offering the treble sensitivity of chrome and the reduced noise coming with the 70us equalisation. In other words: all gain and no pain. Except for cost: ferrichrome was very expensive to make.
2) TDK engineered cobalt-absorbed ferric, 'Super Avilyn', creating a tape specifically matched to the typically-underbiased CrO2 setting of many decks, but offering a higher sensitivity and higher MOL than chrome (and higher noise, of course). This was truly all gain and no pain, as decks now did not even need yet another tape selector, associated circuitry, and associated consumer confusion. The only issue was that once set up for SA, the decks' levels were 2-3dB off for true chrome, and vice versa. This then must have spawn the myth that Japanese decks intrinsically were not compatible with chrome.
But from the late 70s or early 80s on these issues were settled too: metal tape had arrived, demanding even more bias current. This necessarily pushed up head performance. A head capable of recording metal surely was capable of recording chrome! Further, decks adopted more and better internal or external alignment capabilitie, making it easier to match them to any tape.
Ignoring recorder-induced issues, early chrome indeed had a lower sensitivity, a lower MOL, and a much lower SOL, than pseudo-chrome. But this too evolved through time. In 1981 IEC standardisation was initiated and then adopted, aiming to bring the bias needs of all tapes closer to a certain reference. For type II this reference initially, and for a long time, was single-layer BASF chrome. The Japanese ferro-cobalt tape manufacturers simply had to make IEC II-aligned tapes (which they did). And while IEC did not attempt to standardise tape sensitivity with as much vigour as it did for bias, Japanese marketing had instilled the notion that high sensitivity was a good thing, and this forced chrome to catch up (as many review articles testify). So in short, after the early 80s both sides had to grow closer to each other. Failing to do so would have amounted to commercial suicide.
BASF Chrome Super II (1991)
Relative sensitivity: -1.9dB
THD @ Dolby level: 2.9%
MOL400(1%): -7.0dB
MOL400(3%): +0.1dB
MOL1k(3%): -1.5dB
SOL10k: -7.5dB
Bias noise: -56.6dB, -62.0dB(A)
Dynamic range: 62.1dB
For all intents and purposes identical to the 1989 version.
BASF Chrome Super II (1993)
Relative sensitivity: -1.8dB
THD @ Dolby level: 2.8%
MOL400(1%): -6.3dB
MOL400(3%): +0.5dB
MOL1k(3%): -1.3dB
SOL10k: -7.2dB
Bias noise: -56.2dB, -61.6dB(A)
Dynamic range: 62.1dB
BASF Chrome Maxima II (1989)
Relative sensitivity: -2.2dB
THD @ Dolby level: 2.7%
MOL400(1%): -6.3dB
MOL400(3%): +0.5dB
MOL1k(3%): -1.1dB
SOL10k: -7.3dB
Bias noise: -56.6dB, -61.8dB(A)
Dynamic range: 62.3dB
This issue carries the infamous claim of being '99.9% Noise Free', an imaginative way of saying that the signal to noise ratio is 60dB. Well ... it is even better than that!
Performance is identical to the 1989 and 1991 Chrome Super II. But then it is well known that Maxima used selected batches from Super production.
BASF Chrome Maxima II (1991)
Relative sensitivity: -1.9dB
THD @ Dolby level: 1.5%
MOL400(1%): -2.3dB
MOL400(3%): +1.1dB
MOL1k(3%): -0.5dB
SOL10k: -7.0dB
Bias noise: -56.4dB, -61.5dB(A)
Dynamic range: 62.6dB
There are marked differences with the previous version, clearly seen in the reduced span between the 1% and 3% distortion points, and in a MOL that conclusively exceeds Dolby flux (also see 1993). It is not clear if this due to a difference in tape formulation, or to a different degree of ageing.
BASF Maxima Design Edition (1991)
Relative sensitivity: -1.8dB
THD @ Dolby level: 2.3%
MOL400(1%): -5.4dB
MOL400(3%): +1.1dB
MOL1k(3%): +0.8dB
SOL10k: -6.0dB
Bias noise: -55.5dB, -60.7dB(A)
Dynamic range: 61.8dB
BASF Chrome Maxima II (1993)
Relative sensitivity: -1.4dB
THD @ Dolby level: 1.25%
MOL400(1%): -1.6dB
MOL400(3%): +2.2dB
MOL1k(3%): +0.5dB
SOL10k: -7.0dB
Bias noise: -56dB, -61.3dB(A)
Dynamic range: 63.5dB
This is still the chrome version of this CMII. The last production versions allegedly were loaded with ferro-cobalt tape. Performance is still more than reasonable, as shown by the very useful dynamic range.
BASF Chrome Maxima II (1996)
Relative sensitivity: -1.8dB
THD @ Dolby level: 2.7%
MOL400(1%): -5.8dB
MOL400(3%): +0.4dB
MOL1k(3%): -1.15dB
SOL10k: -6.8dB
Bias noise: -56.2dB, -61.5dB(A)
Dynamic range: 61.9dB
I expected this to be a ferro-cobalt tape, but it was chrome, and in a surprisingly bad shape.
BASF Ferrochrom (1977)
Relative sensitivity: -2.3dB
THD @ Dolby level: 2.2%
MOL400(1%): -5.0dB
MOL400(3%): +1.8dB
MOL1k(3%): -3.9dB
SOL10k: -8.8dB
Bias noise: -54.8dB, -59.1dB(A)
Dynamic range: 60.9dB
Operated with Type I bias and 70us replay equalisation. For my first experiences with ferrochrome see here.
Initial attempts to bias for maximum flatness led nowhere: there was always a large difference in level between the channels. Then I set bias for peak output at 400Hz, and accepted the resulting frequency response. Increasing bias did not reduce the high treble peak.
Despite the BX-300 not having a dedicated Type III position the result is not that different from what A.N.T.Audio (Alex Nikitin) got from a Sony TC-177SD (!) with Sony FeCr tape in 2015 (note that '0dB' here corresponds to 160nWb/m DIN):
BASF Ferrochrom III (1980)
Relative sensitivity: -1.9dB
THD @ Dolby level: 1.9%
MOL400(1%): -4.4dB
MOL400(3%): +2.4dB
MOL1k(3%): -3.6dB
SOL10k: -7.0dB
Bias noise: -55.0dB, -59.5dB(A)
Dynamic range: 61.9dB
INDEX OF ALL CASSETTES
