Measurements of Nakamichi Dragon Cassette Deck

[L5730]

It does roll off, but it's much more subtle than anything like lossy formats and their cliff at progressively low frequencies. I don't think tape acts as a low pass filter, which could make things sound warmer. It's more of just a gradual roll off, not extreme, more like a wide high shelf. I suppose lower quality tapes have more roll off.

My results may have also been offset by incorrect bias, despite using "auto-tune". This deck doesn't have 3 heads, so I don't have the luxury of fiddling with bias and seeing instant results.

I'll have to try different things, and post results.

This is strange though, as the typical associated sound of cassettes is that woolly and muffled P.O.S. sound in your car stereo or on the school system.

 

[MakeMineVinyl]

Its important to remember that '20kHz' response is only practical at levels -20dB down from the normal reference level. Recording material which relies on a lot of energy in the high frequency region is going to yield a mushy/compressed sound quality since those frequencies are saturating the tape.

 

[JoachimStrobel]

The worst point is the bad wow and flutter measurements of even the best cassette decks. There is a significant difference between the German Gleichlaufschwankung and the wow and flutter values, so this is difficult to compare. The very best values a German Magazin measured ever for a cassette desk were around 0.1% even if the Japanese spec sheet said 0.05%. Values of 0.15% can be recognized easily, turntables have values around 0.05%. It is hardly audible for rock and pop, but try to listen to the Cologne concert with 0.2% and you run, with 0.15 % it is not enjoyable and with 0.1% there is still a certain roughness to it. And Manfred Eicher was counting on Keith Jarret not holding single notes for too long as his recording tape deck was not the best.
So all these expensive cassette decks with their Dolby xyz, full Metall tapes and auto adjusting heads (auto demag too?) where such a waste. Nothing could be done about the plastic enclosure not allowing for a proper steady tape speed. A third priced reel tape beat them by length. I remember those days, and this insanely expansive cassette decks that were engineering marvels but so senseless - except if one did not care for piano music. And then of course, in a moving care it did not matter, nothing did run steady there, so an unsteady tape modulated by the acceleration was just fine.

 

[L5730]

The worst point is the bad wow and flutter measurements of even the best cassette decks. There is a significant difference between the German Gleichlaufschwankung and the wow and flutter values, so this is difficult to compare. The very best values a German Magazin measured ever for a cassette desk were around 0.1% even if the Japanese spec sheet said 0.05%. Values of 0.15% can be recognized easily, turntables have values around 0.05%. It is hardly audible for rock and pop, but try to listen to the Cologne concert with 0.2% and you run, with 0.15 % it is not enjoyable and with 0.1% there is still a certain roughness to it. And Manfred Eicher was counting on Keith Jarret not holding single notes for too long as his recording tape deck was not the best.
So all these expensive cassette decks with their Dolby xyz, full Metall tapes and auto adjusting heads (auto demag too?) where such a waste. Nothing could be done about the plastic enclosure not allowing for a proper steady tape speed. A third priced reel tape beat them by length. I remember those days, and this insanely expansive cassette decks that were engineering marvels but so senseless - except if one did not care for piano music. And then of course, in a moving care it did not matter, nothing did run steady there, so an unsteady tape modulated by the acceleration was just fine.

Well, I'll try and find something like a sustained piano chord or note and record that (a MIDI synth will do if I can't find a suitable free to use sample, I'd imagine). I do not have high hopes for the format in that regard. With some regular rock/pop even Donald Fagen it sounded OK, but definitely different to the original.

edit. I don't have a top end deck, just a reasonably respectable Yamaha deck.

 

[pma]

Amir seemed to be surprised what he measured on that cassette deck. In fact it shows the only thing - how pointless is all that SINAD story.
Another tape recorder -

 

[MakeMineVinyl]

Amir seemed to be surprised what he measured on that cassette deck. In fact it shows the only thing - how pointless is all that SINAD story.
Another tape recorder -

View attachment 111552

Magnetic recording by its nature cancels out even order harmonic distortion, leaving only the odd order components. The presence of such strong even order distortion in the graph shows the crappy nature of cassettes (I assume the graph is of a cassette).

 

[JoachimStrobel]

Well, I'll try and find something like a sustained piano chord or note and record that (a MIDI synth will do if I can't find a suitable free to use sample, I'd imagine). I do not have high hopes for the format in that regard. With some regular rock/pop even Donald Fagen it sounded OK, but definitely different to the original.

edit. I don't have a top end deck, just a reasonably respectable Yamaha deck.

That is my point: It sounds different. And that is not the frequency flatness, or the noise level or what else. It is the wow and flutter. Even a Revox was not immune to that at anything but top speed. And that is the true wonderfullness of digital sound.
You may have to look for a true stereophonic piano note sample with three strings vibrating in harmony.

 

[sergeauckland]

Magnetic recording by its nature cancels out even order harmonic distortion, leaving only the odd order components. The presence of such strong even order distortion in the graph shows the crappy nature of cassettes (I assume the graph is of a cassette).

Actually, that shows the 'crappy' nature of tape. 1% distortion on tape is pretty good. Even the best studio machines has 3% distortion at peak level, as that was how peak level was defined, the 3% point. And that was at 700Hz or 1kHz!


And yet, and yet, some superb recordings were done in the 60s and 70s with machines like those.

Nevertheless, as one who spent his formative years working with studio and Nakamichi cassettes tape machines, I was very happy to go digital. If for no better reason than to get that sort of performance, tape machines had to be immaculately aligned.
S

 

[MakeMineVinyl]

Actually, that shows the 'crappy' nature of tape. 1% distortion on tape is pretty good. Even the best studio machines has 3% distortion at peak level, as that was how peak level was defined, the 3% point. And that was at 700Hz or 1kHz!


And yet, and yet, some superb recordings were done in the 60s and 70s with machines like those.

Nevertheless, as one who spent his formative years working with studio and Nakamichi cassettes tape machines, I was very happy to go digital. If for no better reason than to get that sort of performance, tape machines had to be immaculately aligned.
S

For studio machines running at 15 IPS, the 3% distortion reference is the one used for determination of signal to noise ratio. This level is generally +6dB above normal operating level. At normal operating level, the typical distortion is around 0.6% and this is predominately odd order distortion. Even order distortion indicates either asymmetrical bias waveform or magnetized heads.

This is a typical distortion spectra of tape distortion at reference level (and also vinyl at reference level):

 

[Digital Mastering System]

The worst point is the bad wow and flutter measurements of even the best cassette decks. There is a significant difference between the German Gleichlaufschwankung and the wow and flutter values, so this is difficult to compare. The very best values a German Magazin measured ever for a cassette desk were around 0.1% even if the Japanese spec sheet said 0.05%. Values of 0.15% can be recognized easily, turntables have values around 0.05%. It is hardly audible for rock and pop, but try to listen to the Cologne concert with 0.2% and you run, with 0.15 % it is not enjoyable and with 0.1% there is still a certain roughness to it. And Manfred Eicher was counting on Keith Jarret not holding single notes for too long as his recording tape deck was not the best.
So all these expensive cassette decks with their Dolby xyz, full Metall tapes and auto adjusting heads (auto demag too?) where such a waste. Nothing could be done about the plastic enclosure not allowing for a proper steady tape speed. A third priced reel tape beat them by length. I remember those days, and this insanely expansive cassette decks that were engineering marvels but so senseless - except if one did not care for piano music. And then of course, in a moving care it did not matter, nothing did run steady there, so an unsteady tape modulated by the acceleration was just fine.

I worked on/co-designed the first 3M digital tape machines after having used many analog tape machines when I worked in radio. The greater S/N was nice, but hearing a Bösendorfer recording on our earliest prototype in '78 made me realize the Wow & Flutter in analog tape was truly horrible. A piano sounded like a piano for the first time with digital, because it could be time base corrected! (It also helped we had horn loaded Altec Model 19 speakers which could, in a small room, play seriously loud.)

 

[JoachimStrobel]

I worked on/co-designed the first 3M digital tape machines after having used many analog tape machines when I worked in radio. The greater S/N was nice, but hearing a Bösendorfer recording on our earliest prototype in '78 made me realize the Wow & Flutter in analog tape was truly horrible. A piano sounded like a piano for the first time with digital, because it could be time base corrected! (It also helped we had horn loaded Altec Model 19 speakers which could, in a small room, play seriously loud.)

So unbelievably true!!!!

 

[pma]

Just for fun, two 1kHz sine waves, one is "pure" and the other recorded to a reel-to-reel tape recorder at about -3dB of its max. level

 

[Blumlein 88]

Just for fun, two 1kHz sine waves, one is "pure" and the other recorded to a reel-to-reel tape recorder at about -3dB of its max. level

The wow and flutter seem more noticeable than the distortion to me on this one.

 

[MakeMineVinyl]

Just for fun, two 1kHz sine waves, one is "pure" and the other recorded to a reel-to-reel tape recorder at about -3dB of its max. level

What is the specifics of the recorder used - machine model, track width, speed, tape used, reference level? I ask this because the R-R version sounds a lot worse than any of my 15 IPS studio recorders. What you posted sounds like 7 1/2 IPS 1/4 track with the bias not adjusted correctly.

 

[MakeMineVinyl]

Just for fun, two 1kHz sine waves, one is "pure" and the other recorded to a reel-to-reel tape recorder at about -3dB of its max. level

As a reality check, I recorded a 1kHz tone as in the above post. This tape is recorded at 15 IPS on RTM SM-468 tape (a professional mastering tape) 1/2 track at reference level on a professional vacuum tube recorder. The other file is direct from the oscillator. The only artifact I can really hear with the tape is a bit of speed instability, although I've tested this machine for wow and flutter and it is performing considerably better than spec.

In any event, these results are far better than the files posted above, and more in line with what I would expect. Professional tape machines of more recent vintage would have better speed stability, and other artifacts such as scrape flutter, modulation noise etc would be better.

Either the tape machine used in the original post was out of calibration or the machine/tape was not professional quality.

Oscillator Direct

Tape

 

[guenthi_r]

Just for fun, tested an PIONEER CT-737 @ 0dB with an old Sony HF TYPE I (from Grandma ):

Edit: This Pioneer is in original shape, never serviced, never cleaned,

Tape-In / Tape-Out:

Dolby OFF:

Dolby C:

Frequency response @ -20dB:

 

[restorer-john]

So your initial loopback in done in rec-pause. What input was required for 0dB on the VU and what was the output?

We need some reference voltages.

The -20dB Type 1 response is pretty good. Did you tweak the bias to obtain a flat response and was the THD measured with the flat response or with a different bias setting? THD will skyrocket when bias is reduced, but the response will flatten.

 

[restorer-john]

It's statements like this:

with an old Sony HF TYPE I (from Grandma ):

that make me feel old.

 

[guenthi_r]

So your initial loopback in done in rec-pause. What input was required for 0dB on the VU and what was the output?

Yes. I do not measured the voltage, it was a simple test.

We need some reference voltages.

Ok, i will re-measure and reply with some data.

The -20dB Type 1 response is pretty good. Did you tweak the bias to obtain a flat response and was the THD measured with the flat response or with a different bias setting? THD will skyrocket when bias is reduced, but the response will flatten.

No tweaking at all. Bias was not adjusted, just at +/- 0.
I will do some measurements with better equipment.

 

[guenthi_r]

It's statements like this: that make me feel old.

The heritage from her includes ca. 300 different cassettes (from TDK D to some METAL tapes, some prerecorded, some factory sealed etc..)
Enough for more then one lifetime.

I have some different Cassette-Recorders, mostly Nakamichi´s and some others.