Is it possible to say that the distinction between analog and digital comes down to the resolution?
Digital=no. of samples per second
Vs.
Analog=no. of oxide particles on a second's length of tape.
If we can sample at the quantity of particles then we wouldn't need a distinction.
Under the "tick tock" strategy, Intel is scheduled to build atom sized transistors in the next few years (HEXUS Mobile Beta)
The resolution of tape is more complicated than that, there are many particles under the gap at any instant so there is a continuous average taken. The tape needs bias too. Taken to the extreme the cheapest dictation tape has "more" than digital.
If you're defining "resolution" as "noise floor," then analog tape is much inferior. If you're defining it as bandwidth (you're implying Shannon-Nyquist), it's close to a wash at 44.1, and analog tape is far behind at 96k or 192k. Analog tape will also suffer from HF saturation, head bumps, scrape flutter, pitch instability, and other artifacts that are eliminated in a digital system. Not to mention archival stability.
Some of the analog tape problems can be reduced with professional formats (e.g., 35mm tape at 30ips) at a jaw-dropping cost. If you're willing to throw out 6 figures for hardware and perhaps $50 per minute of tape, you can almost equal a portable 24/96 unit at $200. Of course, there's no prerecorded material in that analog format...
In the past, I've given some thought to making my own noise reduction scheme for open reel tape, but something superior to the dbx systems of the day that I usually noticed dynamic artifacts from, which is a significant reason that I designed phase correction circuits for both LF and HF. I was able to get phase accurate r/pb tape phase response from below 50 hz to 21Khz at 7.5 ips, but never actually did anything for NR, except record on bulk erased tape & erase head off whenever possible and employ a peak unlimiter which I estimate added almost 10db to the effective dynamic range when used together. I wondered if a constant compression level across the whole dynamic range would have sounded acceptable with or without playback compensation, so that, for instance, a 70db medium could be restored to 90db effective SNR.
I wonder if it would be possible even today to design a low noise erase head function by possibly cascading a pair of erase heads in such a way that the erase field at the tape surface would decrease more slowly, or perhaps a dual gap erase head with the second gap having a reduced field to achieve a minimum horizontal field 'taper' over the critical erase range.
I think it would be interesting to add peak unlimiting such as I designed to playback of a lot of commercial master tapes. The additional 2-3 db or so added to the top end dynamic range with lowered distortion probably would have worthwhile sonic benefits on many recordings. With my design, I was able to record a triangle wave 3 db into saturation and then restore its triangle shape to the point with this circuit. Of course, its effectiveness will be significantly dependent on the exact tape characteristics and bias point in record, but at least half to two-thirds that improvement should be available across the board for a given tape with specified saturability.
I wonder if it would be possible even today to design a low noise erase head function by possibly cascading a pair of erase heads in such a way that the erase field at the tape surface would decrease more slowly, or perhaps a dual gap erase head with the second gap having a reduced field to achieve a minimum horizontal field 'taper' over the critical erase range.
I think it would be interesting to add peak unlimiting such as I designed to playback of a lot of commercial master tapes. The additional 2-3 db or so added to the top end dynamic range with lowered distortion probably would have worthwhile sonic benefits on many recordings. With my design, I was able to record a triangle wave 3 db into saturation and then restore its triangle shape to the point with this circuit. Of course, its effectiveness will be significantly dependent on the exact tape characteristics and bias point in record, but at least half to two-thirds that improvement should be available across the board for a given tape with specified saturability.
Last edited:
If by better you mean closer to the original recording with less distortion and induced artifacts from duplication, then, considering most vinyl was sourced from tape, tape is closer to the source format and therefore probably more accurate to the original, however in reality what's available to the consumer is not the original master tape, it's probably a 3rd or 4th generation copy so we have to factor in duplication noise and tape compression.
OTOH with vinyl (and I've not measured this myself), it is said that the grooves incur a second motion on the needle along a different vector which translates to a second movement for each event and creates a very soft harmonic exciter. If true this probably makes vinyl a bit hyperreal yet rounded and physical sounding at the same time...
The developer behind Ableton Live documented this years ago and wrote a plugin to emulate this vinyl playback phenomenon.
Sampling theorem works really well, but until you get above 384khz I don't see it as equal to analog systems when, for instance filtering. If you perform an EQ in the digital domain, the iterative nature of the filter pole (feeding back on itself to accomplish the time domain phase manipulation) recycles quantization error, truncation, src hash, and whatever other bad habits the DSP math might creating... This error (and there will always be error, it's just much less in a 64bit system) will become more and more present due to continued feedback of samples in dsp loops... The more poles used and higher the Q the more obvious it will be. Analog systems have a different problem (when performing EQ for example) the error in the system is mostly due to noise floor (assuming modern opamps are in use) and so iterative processes like filters will take on the character of the noise floor to some extent. This can create an extra softness... Many prefer this sound, but it is an artifact of the noise floor and other imperfections due to trimming, bias, etc.
Having designed music software and analog synthesizers I can say I prefer the all analog signal path from instrument -> speakers -> ears, however comparing synthesizer DSP running at 512k to analog synthesizers, complex harmonic modulation relationships are starting to behave the same. I think it's safe to say when we reach 1mhz with our sampling rates (at 32bit float or higher internal res) that analog systems will start to lose their identity. For now the majority of desired and revered big consoles used in studios are still analog as well as top mastering engineers still using analog EQ and compression... There are plenty of bands who opt to record to tape (both for the workflow of having to actually play something rather than edit work together) and for the acoustic benefits tape provides (the artifact of tape compression can be very helpful for aesthetic or subtle dynamic range compression).
IMO best or better is based on what you're enjoying at any given time and thankfully that can change as your tastes and interests in music and music reproduction shift.
From a random access standpoint I prefer vinyl over tape... I hate fast forwarding and rewinding. I also don't like lapping heads or calibrating my tape machines even though they sound glorious when finely tuned. For a long time I was into recording masters to DSD and found it the closest digital equiv of tape, but anymore I'm enjoying music too much to care what format it is (as long as it's not exhibiting artifacts (different to each medium)...
Do you like dust and crackles or do you like hiss? Do you like compression or do you like soft harmonic excitation? fwiw tapes fall apart.. the binder holding the iron degrades over time... it's a pretty lame medium from that standpoint.
OTOH with vinyl (and I've not measured this myself), it is said that the grooves incur a second motion on the needle along a different vector which translates to a second movement for each event and creates a very soft harmonic exciter. If true this probably makes vinyl a bit hyperreal yet rounded and physical sounding at the same time...
The developer behind Ableton Live documented this years ago and wrote a plugin to emulate this vinyl playback phenomenon.
Sampling theorem works really well, but until you get above 384khz I don't see it as equal to analog systems when, for instance filtering. If you perform an EQ in the digital domain, the iterative nature of the filter pole (feeding back on itself to accomplish the time domain phase manipulation) recycles quantization error, truncation, src hash, and whatever other bad habits the DSP math might creating... This error (and there will always be error, it's just much less in a 64bit system) will become more and more present due to continued feedback of samples in dsp loops... The more poles used and higher the Q the more obvious it will be. Analog systems have a different problem (when performing EQ for example) the error in the system is mostly due to noise floor (assuming modern opamps are in use) and so iterative processes like filters will take on the character of the noise floor to some extent. This can create an extra softness... Many prefer this sound, but it is an artifact of the noise floor and other imperfections due to trimming, bias, etc.
Having designed music software and analog synthesizers I can say I prefer the all analog signal path from instrument -> speakers -> ears, however comparing synthesizer DSP running at 512k to analog synthesizers, complex harmonic modulation relationships are starting to behave the same. I think it's safe to say when we reach 1mhz with our sampling rates (at 32bit float or higher internal res) that analog systems will start to lose their identity. For now the majority of desired and revered big consoles used in studios are still analog as well as top mastering engineers still using analog EQ and compression... There are plenty of bands who opt to record to tape (both for the workflow of having to actually play something rather than edit work together) and for the acoustic benefits tape provides (the artifact of tape compression can be very helpful for aesthetic or subtle dynamic range compression).
IMO best or better is based on what you're enjoying at any given time and thankfully that can change as your tastes and interests in music and music reproduction shift.
From a random access standpoint I prefer vinyl over tape... I hate fast forwarding and rewinding. I also don't like lapping heads or calibrating my tape machines even though they sound glorious when finely tuned. For a long time I was into recording masters to DSD and found it the closest digital equiv of tape, but anymore I'm enjoying music too much to care what format it is (as long as it's not exhibiting artifacts (different to each medium)...
Do you like dust and crackles or do you like hiss? Do you like compression or do you like soft harmonic excitation? fwiw tapes fall apart.. the binder holding the iron degrades over time... it's a pretty lame medium from that standpoint.
Last edited:
Is there a technical translation for the above? mis-tracking, warp or out of round FM, the list goes on?
As best I remember it was described as a second motion occurring at 90deg to the first and producing either transient exaggeration or added harmonics at higher intervals... It wasn't described in detail, probably because he didn't want to give other plugin developers a formula. Knowing his history I don't doubt he measured a real distortion or modulation that is present with vinyl.
Appreciated reading a detailed post from a pro who is not invested in the SQ of his CD collection.
Wrt the secondary modulation scheme with vinyl that was described, it would seem to be primarily resultant from bass information where the absolute groove modulations are higher, particularly at higher absolute levels. Possiby an odd-order harmonic effect?
I actually started out to make a version of a vinyl tick and pop eliminator roughly based on the architecture of the burwen tne7000 but with a tube audio chain. I should really resume that project - I have the audio paths working in place (the unit already operates as a tubed version of the Carver Sonic Hologram generator - getting sufficient interaural delays with tube circuitry was challenging but was able to do it with 4 cascaded 2nd order all - pass circuits per channel.)
Wrt the secondary modulation scheme with vinyl that was described, it would seem to be primarily resultant from bass information where the absolute groove modulations are higher, particularly at higher absolute levels. Possiby an odd-order harmonic effect?
I actually started out to make a version of a vinyl tick and pop eliminator roughly based on the architecture of the burwen tne7000 but with a tube audio chain. I should really resume that project - I have the audio paths working in place (the unit already operates as a tubed version of the Carver Sonic Hologram generator - getting sufficient interaural delays with tube circuitry was challenging but was able to do it with 4 cascaded 2nd order all - pass circuits per channel.)
Last edited:
Found this from another forum:
I've always assumed LP's were worthless as they were old and analog- digital is better, right?
As I got older I realized that while CD's don't suffer from dirt, scratches, etc, you lose a certain amount of frequency information when digitally sampling the music. Essentially, you sample at 44,100 KHz and everything above that gets thrown away. (Technically the Nyquist Theorem states that a sampling rate of 2x yields a frequency range of x, but you get the idea).
With vinyl, there is no sampling rate. The grooves are a one-for-one copy of the original recording. I'm 26, and had no previous experience with vinyl. I got a turntable for free from a friend who was throwing it away a month ago and I took it out of curiousity. I went to a local book store and bought some LP's for 33 cents. I figured if these sounded like junk I was only out $1.
My first taste of vinyl was "Tragedy" by the Bee Gees. Even with all of the dust and popping from the uncleaned record, it sounded lovely. It's almost like listening to it live, and you can make out instruments and vocals more clearly. I bought some Eric Clapton albums with songs that I've listened to many times, and it sounded so much better. The CD is cleaner sounding, but the vocals and the instruments run together more (less distinct sounding)my comment: Uh, whuh, you can't say that about perfect sound forever! and guitars, violins, and vocals don't have the fidelity or clarity that the uncompressed vinyl can offermy comment: Who needs fidelity or clarity if it thumps?.
I got a much better turntable (Technics SL-1300) with a new needle and a nice phono cartridge. I found out that you need to attach a ground wire from the turntable to your receiver to eliminate hum. I bought a cleaning brush and wet cleaned the old albums I had. With a decent turntable and a cleaned record (even the old 33 cent ones that were rather scratched up) you hardly hear any dust or popping, even during silent parts like the blank parts of the LP in between tracks. Even if you can hear pops, you don't hear them during the song. With this setup, I found that I was enjoying the music much more, and it even impacted me emotionally even more as it's so much more clear, you forget you're listening to a recording. It's renewed my interest in music again, no joke.
With a good turntable that has a cuing lever, you can easily lift and lower the needle without scratching the record. You can also choose individual tracks in this manner.
It's not a scam or fad, it's quite real. I'm going to listen to my Thriller album now- despite listening to "Beat It" or "Billie Jean" dozens of times, listening to it on vinyl is a whole new experience. I brought my friend over to listen to these two tracks and he just sat there shaking his head in disbelief after hearing them. my comment: Another couple Kool-Ade refugees.
I've always assumed LP's were worthless as they were old and analog- digital is better, right?
As I got older I realized that while CD's don't suffer from dirt, scratches, etc, you lose a certain amount of frequency information when digitally sampling the music. Essentially, you sample at 44,100 KHz and everything above that gets thrown away. (Technically the Nyquist Theorem states that a sampling rate of 2x yields a frequency range of x, but you get the idea).
With vinyl, there is no sampling rate. The grooves are a one-for-one copy of the original recording. I'm 26, and had no previous experience with vinyl. I got a turntable for free from a friend who was throwing it away a month ago and I took it out of curiousity. I went to a local book store and bought some LP's for 33 cents. I figured if these sounded like junk I was only out $1.
My first taste of vinyl was "Tragedy" by the Bee Gees. Even with all of the dust and popping from the uncleaned record, it sounded lovely. It's almost like listening to it live, and you can make out instruments and vocals more clearly. I bought some Eric Clapton albums with songs that I've listened to many times, and it sounded so much better. The CD is cleaner sounding, but the vocals and the instruments run together more (less distinct sounding)my comment: Uh, whuh, you can't say that about perfect sound forever! and guitars, violins, and vocals don't have the fidelity or clarity that the uncompressed vinyl can offermy comment: Who needs fidelity or clarity if it thumps?.
I got a much better turntable (Technics SL-1300) with a new needle and a nice phono cartridge. I found out that you need to attach a ground wire from the turntable to your receiver to eliminate hum. I bought a cleaning brush and wet cleaned the old albums I had. With a decent turntable and a cleaned record (even the old 33 cent ones that were rather scratched up) you hardly hear any dust or popping, even during silent parts like the blank parts of the LP in between tracks. Even if you can hear pops, you don't hear them during the song. With this setup, I found that I was enjoying the music much more, and it even impacted me emotionally even more as it's so much more clear, you forget you're listening to a recording. It's renewed my interest in music again, no joke.
With a good turntable that has a cuing lever, you can easily lift and lower the needle without scratching the record. You can also choose individual tracks in this manner.
It's not a scam or fad, it's quite real. I'm going to listen to my Thriller album now- despite listening to "Beat It" or "Billie Jean" dozens of times, listening to it on vinyl is a whole new experience. I brought my friend over to listen to these two tracks and he just sat there shaking his head in disbelief after hearing them. my comment: Another couple Kool-Ade refugees.
Other way round. You throw away everything above 22.05kHz and then sample at 44.1kHz. Anything above 22.05kHz you don't throw away gets aliased down to a lower frequency so it doesn't disappear.thoriated said:
Only in the ideal world with inelastic vinyl and perfectly tracking cartridges etc. Vinyl just swaps one set of compromises for another set.
!?- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.