Not another debate thread - Just a question

More particles definitely equals a higher resolution of sound reproduction. That's why higher speed recording and wider track width almost always equates to a more lush sound. But, our human ears are not the best test instruments to tell specifically what those differences are as we can be fooled quite easily by simple things like loudness. Make something a db louder, and we think it sounds better. Not exactly scientific. Same thing goes for distortion. Our ears can spot a difference between 10% THD and 0.5 % THD. But would be hard pressed to tell the difference between 0.5 and 0.005 THD. And when it comes to spotting the differences between analog and digital, its usually only the absence of audible hiss or harder sounding high frequencies that gives our human ears a clue as to which is which.

Cheers! :)
 
ashcat_lt said:
I don't know enough about it to say exactly what physical properties cause it - but typical analog tape machines have only slightly better high frequency response, and quite a bit less dynamic range compared to even CD quality digital audio.
Click to expand...

If by typical, you mean some pitiful, portable, picnic player (10 points for identifying that reference :-) )then yes. Two of the multitracs in my modest studio run out to 45Khz with a dynamic range, using Dolby SR, of about 110dB.
 
mdainsd said:
If by typical, you mean some pitiful, portable, picnic player (10 points for identifying that reference :-) )then yes. Two of the multitracs in my modest studio run out to 45Khz with a dynamic range, using Dolby SR, of about 110dB.
Click to expand...

Ok, ok. Let's try to contain this right now. This is not meant to be another debate thread! I understand some clarification, and that's cool, but let's not let this line of thinking get away from us please. :)
 
I did some quick research, and quite frankly could not find anything specifically talking about the amount of oxide particles acting in some sort of sampling scenario like you have with digital conversion. The particles simply get magnetized (they store nothing), and then that magnetism -- flux-- is what is read by the heads on playback. Less oxide may yeild more easier distortion but that's not the same as less/more sampling frequency.

Yes, wider tape, larger formats = more flux capacity (not to be confused with the flux capacitor used for time travel) or more flux strength/depth, and while that will give you greater audio fidelity (cassette tape track width VS 1/2" 2-track width)...it's still not quite the same as when looking at different digital sampling rates.

Sampling rates are "slices" of the audio over time. Magnetic flux is more about signal strength/depth...which is why some open reel formulations will withstand higher audio signals (stronger electronic signals converted to stronger magnetic fields) than others, and certainly much more so than typical cassette tapes. As the signal strength gets to the limits of what the tape can handle (there's nothing left to magnetize) the tape is "saturated" and the signal starts to distort, first gently, but eventually it will get nasty....unlike with digital where it just goes instant nasty.
Faster speed of tape allows the wavelengths of the flux to magnetize more oxide without going into saturation as easily as slower speeds.
 
famous beagle said:
Ok, ok. Let's try to contain this right now. This is not meant to be another debate thread! I understand some clarification, and that's cool, but let's not let this line of thinking get away from us please. :)
Click to expand...

Wasnt meant to stir debate, just a response. I'll stay out of your threads.
 
miroslav basically nailed it in his first couple posts. Analogue doesn't work at all like digital sampling, but the farther we get from when audio engineers were well versed in basic physics the more difficult it gets for people to see audio recording and reproduction outside of the digital model. When you have a rudimentary understanding of digital sampling its more about how your mind is being oriented than how magnetic particles are getting oriented.

I say rudimentary understanding of digital because when you gain a greater more detailed understanding of digital sampling the similarities to analogous electromagnetic audio/video systems quickly disappear. It's more about what people don't know, like that signal-to-noise ratio is not the only thing impacted by word length. The effectiveness of error correction schemes is another issue. Digital systems all have a certain number of uncorrectable errors and the greater the word length the fewer uncorrectable errors there are.

Finer more densely packed oxide particles make a better recording in analogue terms, but we don't really have a direct comparison in digital terms until the digital data enters the analogue realm, which everything must before we can hear it... of course. The OP has the basic concept down, but its just a bit contaminated by digital concepts that understandably skew the thinking of today's recordist.
 
Veering OT: Not to muddy the waters, but would this be analogous (pun intended) to 12 inch 45 rpm records vs 16rpm records (yes they existed) squeezing grooves together?
 
Not squeezing grooves together, but just being at a slower speed. Have you seen those audiophile-quality reissues on 45rpm discs? I remember seeing Metallica albums presented that way. The 33 1/3rpm disc was for cramming a bunch of songs together, originally. One can hear a noticeable difference between a song on a 33rpm disc and that same song on a 45rpm disc. 16rpm is for voice-grade stuff...audiobooks, learning Spanish; it's the 1 7/8ips (maybe 3.75ips?) of the record world. Groove width is for better bass response and loudness...I guess that does have to do with how much music you want on one side of a disc, say an average of 16mins at +3dB compared to 20 (22?) mins at -3dB.

I like how Miroslav explained saturation; I think that pretty much said it all, but to translate that into a digital perspective, I would say that an audiotape can store infinite information at a certain volume level. Once you're above 0dB however, you run out of oxide particles - all particles have been aligned magnetically (saturation). Play a guitar straight into a tape deck and overload the signal, and you'll get a quite similar sound to distortion through an amp. So yes, in a sense, there IS a limit to how much information oxide particles can store. That's where we have +3, +6, and +9 tapes coming in. They are high-capacity tapes designed to be able to store more information. Where (more = louder) and (information = signal/noise ratio). That's my thought on this.
 
Last edited:
Used to know more about this, long time ago....But here's my 2pen'oth..

They are magnetic "domains" and the more you have passing the head per second the greater the amount of information they can store. So, early tape machines with rubbish tape and few domains (big particles) had to run fast, 30ips, to give even modest fidelity. Pretty soon tapes were good enough for 15ips to be the "pro" standard (except a few Dolby hating diehards!) . As particles got ever smaller and better dispersed speeds could drop and we now have cassette at 1:7/8ips getting near those ancient OR machines, at least for frequency response and noise. Smaller and better behaved particles give lower noise but also doubling tape speed doubles the signal output, 6dB, but noise is "incoherent" and only increases by 3dB (doubling track width does the same)

As for maximum output the situation is as I think someone said, you have used up all the magnetic domain/seconds (TINY bit like using all yer digits?).

Tape is of course "sampled" in a way by the bias frequency and there is a direct parallel with digital here in that frequencies above 1/2 the bias F must not enter the recording amplifiers or they cause in band beats. Some of you will know about FM stereo subcarrier filters?

Dave.
 
ecc83 said:
Used to know more about this, long time ago....But here's my 2pen'oth..

They are magnetic "domains" and the more you have passing the head per second the greater the amount of information they can store. So, early tape machines with rubbish tape and few domains (big particles) had to run fast, 30ips, to give even modest fidelity. Pretty soon tapes were good enough for 15ips to be the "pro" standard (except a few Dolby hating diehards!) . As particles got ever smaller and better dispersed speeds could drop and we now have cassette at 1:7/8ips getting near those ancient OR machines, at least for frequency response and noise. Smaller and better behaved particles give lower noise but also doubling tape speed doubles the signal output, 6dB, but noise is "incoherent" and only increases by 3dB (doubling track width does the same)

As for maximum output the situation is as I think someone said, you have used up all the magnetic domain/seconds (TINY bit like using all yer digits?).

Tape is of course "sampled" in a way by the bias frequency and there is a direct parallel with digital here in that frequencies above 1/2 the bias F must not enter the recording amplifiers or they cause in band beats. Some of you will know about FM stereo subcarrier filters?

Dave.
Click to expand...

So how are does "normal" bias and "high bias" tape come in to play here? I mean was like normal bias "bigger" particles and high bias "smaller" particles?? Ive always gotten much better results with high bias cassettes but with open reel tape it was just called a "higher output" tape....wanna further explain how this works and thanks !!!
 
Okay, as I understand tape, it's not the amount of particle, but the quality of the particle that counts. Type I tape (iron oxide) is low to mid quality (because of the quality of the particles) and will give you a fairly unflat response and low headroom; Type II tape (chromium oxide) is mid to high quality and gives flatter response on the high frequencies and more headroom (because the particles, although not more populous, are finer, or more in tune with the magnetic fluctuations) thereby giving less distortion; and Type IV (ferric or metal) has the highest quality, flattest response and best headroom because the particles again are better at forming the pattern.

It's not so much an analogy between digital and analog bit depth, because it really doesn't compare (not to start a debate, but there are two entirely different processes going on). The digital signal (in a strictly theoretical sense) is perfect until it distorts, at which point it becomes complete garbage (my opinion). The tape is technically never perfect, but a certain amount of analog distortion makes things actually sound better (again, my opinion, based on the fact that many digital systems try to emulate tape noise and distortion...).

So I like beagle's analogy because the quality of the particles pretty much directly correlates to the reduced risk of digital distortion from higher bit depth, both of which allow higher headroom (or S/N).:thumbs up:
 
TASCAM MAN said:
So how are does "normal" bias and "high bias" tape come in to play here? I mean was like normal bias "bigger" particles and high bias "smaller" particles?? Ive always gotten much better results with high bias cassettes but with open reel tape it was just called a "higher output" tape....wanna further explain how this works and thanks !!!
Click to expand...

Me and my big mouth....Oo! My head hurts.
Magnetic materials come with different magnetic properties. The first basic tapes were gamma ferric oxide (Fe2 O3 IIRC) and this has a relatively "soft" magnetic charateristic. Easy to magnetize but tends to lose its magnetic force easily as well (the property of how well a material resists demagnetization is called "coercivity") Cobalt loaded tapes could handle higher levels but you had to "hit them harder". Chromium Dioxide is similar but can have even smaller particles and thus lower noise.

Because of the co-existance of typ 1 and type ll tapes (and later auto switching) cassette deck makers fitted bias/eq switches, even bias pots and lineup oscillators and in the limit, CPU controlled autosetup (I had a LUVERLY Denon HX with that!) . Such niceties never made the move to open reel AFAIK? The whole casstte tape sitution is a morass of interlocking compromises of film thickness, coating thickness, dispersants, adhesives...And everybody made the BEST tape and they all lasted for ever. Some could even break glass!

I have a Hi-Fi News annual somewhere with a big feature called "Cassettes and Coatings". This was a lot of very heavy science, chemistry and industrial nonce that has all but been forgotten.

Dave.
 
Dave made a great point I missed earlier. Tape speed is almost a perfect analog to bit depth. The faster you run the tape, the better the sound, just like going to a deeper bit depth, but going from 7 1/2 to 15 doubles the amount of space needed (tape used). Going from 8 bit to 16 bit doubles the disc space used (not that most record at 8 bit, but most don't record with a 4 track 1 7/8 cassette either.

The main reason for higher bit depth (imo) is to get more natural amplitude decay... same as running higher tape speed. Also there are definite trade-offs to both. I've never heard of a 60ips tape machine (they may exist), getting past 30 wouldn't give you that much of a boost. I've never heard of anyone recording past 24 bit, but going farther than 24bit doesn't really give enough of a boost to make it worthwhile, either.

So sample rate relates to the tape coating and the better or smaller the particles are (allowing higher frequencies to come through correctly).
 
BroKen_H said:
Tape speed is almost a perfect analog to bit depth. The faster you run the tape, the better the sound...
Click to expand...

Not necessarily.

Different speeds alter some of the frequency responses of the tape...so actually, you get more low-end thump with slower speeds on many machines, and lots of times 15 ips was favored over 30 ips for that reason.
Less bit depth is always a limiting thing....less info.

IMO....things like bias, tape speed, tape type, machine choice.....in the heyday of analog tape recording, those things were/are more like color choices than pure fidelity choices like you have with less/more bit depth and sampling frequencies.
 
As a few of the experts here have already explained, it doesn't really make sense to compare because it's really different worlds.

In the analog world, you're 'battling' signal to noise, distortion, etc ... even the junkiest analog is still a 'pure' continuous signal, just a crummy, noisy one. With crappy digital, you wouldn't really have noise problems, it would just be a badly constructed recreation of the signal.

Tape speed, track width, etc. are only relevant in the analog world and have no digital equivalents. Even these things are probably less important than the electronic design of the machine itself (the full audio path). 30 ips is not objectively better than 15 ips, there are pros and cons to both (15 ips has better low end, 30 ips has better high end and lower noise). Tape width, maybe, but once you reach a certain point, it's really just better signal to noise.

I don't think oxide particles can be compared at all. The oxide is really just the canvas ... the digital equivalent is probably a hard-drive. I mean, the tape type does influence the sound, but again it really isn't comparable to anything in the digital realm, because in a sense NOTHING digital actually has a valid sonic 'texture' -- only the analog stages do. In theory, once digitized, it's 'perfect', i.e. it will not change. The signal on a strip of tape will actually change physically over time.

Another reason these comparisons are not relevant is in analog, these things are not subtle -- they're very obvious. The differences in digital once you reach a certain point are very subtle things. Lots of people can't tell the difference between standard 16-bit CD and the fancier digital choices people have these days. Ditto basic, cheap converters and super-$$$ ones.
 
famous beagle said:
Yeah, I'm not saying they store the audio. I was trying to say that they align in a pattern that represents the magnetic field. But ... "they" are not infinite. They're particles, so there has to be a finite number of them, correct? And doesn't that mean that there's a limit as to how accurately they can represent the magnetic field (i.e., signal)?
Click to expand...

but they're MUCH smaller than the magnetic pulses they store ...... MUCH smaller .... sure ... they are finite in number but the amount of info any individual particle stores is negligible and, in fact, probably couldn't even be measured.
Plus ..... they don't really store an audio info anyways. It's only when the tape is moving that the info that millions upon milions of them can be read.
There simply is no correlation whatsoever between bits stored on a computer and individual oxide particles on tape. No way to compare them and no correlation between the two. There is no similarity between how the two work and it's meaningless to compare the two.
 
Terminology issues

You are thinking correctly, but your use of terminology might not be completely accurate. Giving you the answers would rob you of the opportunity to do the research into a question only you have really posed. Here's what I suggest you do. Take your terms and look them up in a Home Recording glossary. Then research their uses online by Googling the words. Now, using your theoretical thinking try to understand the analogue vs digital question with this new information. There are differences that make this an apples and oranges question and your question really has two parts.

One- How does sound get converted and transferred?
Two- What are the similarities and the differences?

Do the research, find the answers and then come back and post your findings. You are thinking way above the level of most recording enthusiasts. Good luck,
Rod Norman
 
"Different speeds alter some of the frequency responses of the tape..."

Err, not really. The tape induces a current in the replay head and the output is proportional to rate of change of flux. Assuming a constant recorded flux (constant current drive) the the head output therefore increases at 6dB per octave. The replay characteristic for high speeds is therefore pretty simple with only a minor tweak needed to account for small losses at HF . Cassette speeds demand considerably more tweaking!

The lumpy bass response is a function of tape speed and head contour. I guess very few heads were optimized for 30ips?

Dave.
 
Analogue as this analogy refers to it, an analogue being made of the sound, is often confused with being an exact recording but it's not; it's an analogue not a replication/clone etc.
In terms of confused understanding it's a little like the situation of simile and metaphor.
In truth no recording is a metaphor (that horse is a jet/that sound is a trumpet) but all recordings are similes (that horse is like a jet/that sound is like a trumpet) the quality of the recording being based on how accurately or emotively the sound is reproduced, (which can be shown to be very subjective or confused by the emotioanl attachement to the poor translation/reporduction of many a mellotron "sample" which is close but not by a mile yet is sought out for it's emotive quality).
Then again, the memory is probably the worst recording device know as it's is highly manipulable on the way in and out let alone the problems of data corruption, drop out and cyclical error.
I hope I muddied the waters.
 
You must log in or register to reply here.
Back
Top