How Many People Are Using Cassettes Here?

ecc83 said:
You are quite wrong. I am not a 'digital expert' but I have read several times by such experts that a digital system DOES reconstruct PERFECTLY the signal, sine if you like up to just below the Nyquist frequency, i.e. 20kHz for a 44.1kHz sample rate.
If you don't believe me, go ask the question at soundonsound.com and the ex BBC engineers will soon put you right!
Click to expand...

It is hard work keeping up with this fast thread.
The issues are:
You can completely recreate an analogue signal if you sample it at higher than twice the highest frequency. (Nyquist)
However, that is only if you take perfect analogue samples.
But with A/Ds and D/As, we are degrading those perfect analogue samples, by quantizing them as voltage steps.
The defence here would be oversampling, and smoothing the output with filtering.
There comes a point at which the unavoidable noise floor is bigger than any conversion error.
So, you couldn't tell the difference between a re-created signal and the analogue original.
 
Slouching Raymond said:
It is hard work keeping up with this fast thread.
The issues are:
You can completely recreate an analogue signal if you sample it at higher than twice the highest frequency. (Nyquist)
However, that is only if you take perfect analogue samples.
But with A/Ds and D/As, we are degrading those perfect analogue samples, by quantizing them as voltage steps.
The defence here would be oversampling, and smoothing the output with filtering.
There comes a point at which the unavoidable noise floor is bigger than any conversion error.
So, you couldn't tell the difference between a re-created signal and the analogue original.
Click to expand...
That is my (limited) understanding Ray and 24bit digital has a noise floor of -144dB but of course, no analogue circuit can get close to that. The very best converters do however have a dynamic range better than 120dB and even quite modest, sub $100 for a 2 in 2 out interface, are now better than 100dB. A figure no analogue recording process can get close to. IF there was any 'chopped' residue left we could never hear it because the analogue outputs, even the best ones are too noisy!

Dave.
 
Slouching Raymond said:
It is hard work keeping up with this fast thread.
The issues are:

However, that is only if you take perfect analogue samples.
But with A/Ds and D/As, we are degrading those perfect analogue samples, by quantizing them as voltage steps.
The defence here would be oversampling, and smoothing the output with filtering.
There comes a point at which the unavoidable noise floor is bigger than any conversion error.
So, you couldn't tell the difference between a re-created signal and the analogue original.
Click to expand...
Good points, Raymond.

It's not as if analog equipment is completely linear in their conversion of the source to a recorded signal and vice versa. Record/playback heads are not any more perfect in their responses than a digital/analog converter. Tape advocates talk about being able to push a tape, but the end result is merely to create massive errors in dynamic range and distortion (several %, not thousandths of a %). Phono cartridges are likewise flawed in their ability to convert the variations in a groove to an electrical signal. If anything, current DA converters are orders of magnitude better at accurately converting an electrical signal to a digital value. And, as Dave mentioned, adding dither will minimize any sampling errors further.
 
TalismanRich said:
Good points, Raymond.

It's not as if analog equipment is completely linear in their conversion of the source to a recorded signal and vice versa. Record/playback heads are not any more perfect in their responses than a digital/analog converter. Tape advocates talk about being able to push a tape, but the end result is merely to create massive errors in dynamic range and distortion (several %, not thousandths of a %). Phono cartridges are likewise flawed in their ability to convert the variations in a groove to an electrical signal. If anything, current DA converters are orders of magnitude better at accurately converting an electrical signal to a digital value. And, as Dave mentioned, adding dither will minimize any sampling errors further.
Click to expand...
Ah but! THEN Ray the Annylogophiles will reply with "But digital sounds sterile". I sometime wonder if some of these guys have done any 'live' recording. In my 20s we were recording Gang Shows and AmDrams (for sweet FA! ) using a home made triode mixer and a 15ips Brennel and Ferrograph 7. I would have KILLED for the preamps in my Behringer mixer never mind my MOTU M4! I could have effectively left everything set recording at neg 30/24 bits and hissed off to the pub round the corner!

We needed two recorders running 20mins behind the other because of the p*** poor running time of tape!

Dave.
 
Here is some frequency response videos for different cassettes on this Sony TC-K555esG. This deck is 33 years old so take that into consideration. Also the heads have some wear on them. Not terrible, but not new. Still does a good job

Check the bottom left side of the Oscilloscope screen to verify frequency

TDK D 15 kHz -20 dB


TDK CD Power 16 kHz -20 dB


Sony Metal Master 18 kHz -20 dB
 
Smithers XKR said:
Paul, I have had this discussion with Boulder, it is all very gentlemanly and courteus but my point is that digital can never recreate analogue waveform by its mathematical calculus nature. No matter how high the digital resolution it can never recreate a perfect sine wave, we respectful agreed to disagree 😅😉😉👍
Click to expand...
Digital audio recreates analog waveforms literally all the time. That's pretty much what it was designed to do. Is it perfect? No. Is it objectively more accurate, consistent and stable than analog? Absolutely. Is it subjectively better? That is, by definition, a matter of opinion. In digital, it's possible to emulate analog on demand (to some degree of accuracy), but the reverse is not true. Sonically, digital gives you the best of both worlds.
 
Exactly what are we seeing on the scope? Is that a source signal vs recorded signal on top and bottom, or is the rock solid signal the generator and the bouncing signal the tape? A little explanation of your video might be helpful. It's impossible to tell if you are just freezing the screen, or alternating from source to tape.

Now, it might be interesting to do the exact same test with a digital interface using 24bit 88kHz recording. You could do the same 18kHz signal you used for the metal test.
 
I am doing a recording on the tape and I am monitoring it on the oscilloscope since they are 3 head decks. You can monitor recordings in real time. When you are seeing the solar scope with both left and right channel jumping a little bit that’s the tape. When it’s still that’s the source. It’s a comparative analysis of what’s going on in real time
 
When I extend the frequency out to 21 kHz you can see that the 2 channels on tape playback are slightly out of phase

F5-E72-CE4-8-C8-E-41-FB-B627-71-E1-CE27-BD36.jpg
 
I installed muting transistors on the back of one of my TC-K890ES decks and that has significantly lowered the noise floor without the use of Dolby
 
clamsterdamm said:
I am doing a recording on the tape and I am monitoring it on the oscilloscope since they are 3 head decks. You can monitor recordings in real time. When you are seeing the solar scope with both left and right channel jumping a little bit that’s the tape. When it’s still that’s the source. It’s a comparative analysis of what’s going on in real time
Click to expand...
OK, that's kind of what I was assuming, but with no information, making such assumption can often be wrong.

That said, you don't feel the bouncing of the signal is a problem? How many dB does that represent, as that is obviously a variation in level? Have you measured the distortion levels? What is the measured noise floor(which should be easily determine with the scope)? Just seeing that it can make an 18k wave at -20dB isn't giving a complete picture.

Then, as I said, why not do a digital recording using the same methodology, You can even do it at CD spec (44.1k, 16 bit). How does that compare for stability and noise level to the metal tape, which appears to be the best tape result.
 
The vertical jumping you see between the channels is the amplitude loss. This represents +/-3 dB. No it shouldn’t be this way, but since the deck is over 30 years old it has been used so the heads are not perfect anymore. Also, this deck could use a recap. There are a few contributing factors to this loss

I have limited equipment here But will try your recommendation
 
clamsterdamm said:
Cassette decks aren’t designed to hit 44.1 kHz

The best they’ll really do is 22 kHz without significant loss
Click to expand...
I'm talking about the digital sample rate, not the frequency response. My point was to use the "lowest resolution" common digital format as a comparison.
 
bouldersoundguy said:
I'm talking about the digital sample rate, not the frequency response. My point was to use the "lowest resolution" common digital format as a comparison.
Click to expand...
Ok I am sorry, I had a bit too few many beers last night and perhaps got slightly argumentative, I do not have the technical knowledge of guys like you and Clam and the others. I am a musician and producer but I am qualified to talk about Maths and digital theory. I just think that there are places for both analogue and digital formats.
I have learned a hell of a lot from good guys here and this is a great forum.
Please take this message in the genuine way it is intended, all is cool as far as I am concerned.

Thanks and best regards 😉👍👍
 
clamsterdamm said:
Cassette decks aren’t designed to hit 44.1 kHz

The best they’ll really do is 22 kHz without significant loss
Click to expand...
But surely 44k is above the range of the human ear so it does not matter? That was my query in 94 when I bought a DCC for £250 instead of the original Sony DAT for £700. I was told at the time... no dont buy a DCC as it is inferior quality using Data Compression. When I asked what that meant it was explained to me that it takes out the frequencies that are inaudible to the human ear... Err sorry ? Was I missing something? Should I have asked my Jack Russell Terier dog to choose between DAT and DCC for me?
Cummon man please 😉🤣
 
AFAIK Smithers, DAT is a 'linear' digital system as CD is and .wav recording. I have never used DAT but I still have a Philips DCC machine and some still sealed blank cassettes. DCC does use a form of 'compression' which a useful but technically wrong word, the recorder uses a process similar to MP3 I believe. I have not fired it up for years but IIRC it was bloody close if not AS good as CD? 18bits I think was claimed. My son used it to store takes from the Teac and remix them. Compared to tape there was no loss of quality. You cannot of course copy tape to tape without SOME loss of dynamic range.

I also have a two Minidisc decks, a format I think never really got a chance. It would be far better in cars than cassette or CD. But like DCC MD uses a 'lossy file reduction system and so really should not be compared to the very best analogue (1st gereration 15ips master tapes) nor indeed CD. Now that you can have 32G of storage on something the size of your little fingernail for about $10.00 I see no need of MP3?

Dave.
 
ecc83 said:
AFAIK Smithers, DAT is a 'linear' digital system as CD is and .wav recording. I have never used DAT but I still have a Philips DCC machine and some still sealed blank cassettes. DCC does use a form of 'compression' which a useful but technically wrong word, the recorder uses a process similar to MP3 I believe. I have not fired it up for years but IIRC it was bloody close if not AS good as CD? 18bits I think was claimed. My son used it to store takes from the Teac and remix them. Compared to tape there was no loss of quality. You cannot of course copy tape to tape without SOME loss of dynamic range.

I also have a two Minidisc decks, a format I think never really got a chance. It would be far better in cars than cassette or CD. But like DCC MD uses a 'lossy file reduction system and so really should not be compared to the very best analogue (1st gereration 15ips master tapes) nor indeed CD. Now that you can have 32G of storage on something the size of your little fingernail for about $10.00 I see no need of MP3?

Dave.
Click to expand...
Thanks 😉👍👍
 
You must log in or register to reply here.
Back
Top