| General > General Technical Chat |
| Interesting bit of kit, cassette tape duplicator |
| << < (4/4) |
| Simon:
I would expect that using a tape for digital storage would have been expensive at the time. I suspect you would need more tape as now the resolving power of the medium that would determine the "sample rate" would need to be higher. Although analogue is not dealt with in terms of resolution or datarate because the medium mechanisms are too fuzzy to put a number on there is a physical limitation. For example we used to record loads of films on VHS long play to save money, the quality was visibly less as we were at the limit of the medium but we accepted it and the reason the standard speed of tape was twice long play was because that was what was required to achieve the effective data rate. If you want to digitise sound then what on the tape used to represent 1 "sample" now would be one bit in 32? so you would need 32 times the tape not to mention more expensive equipment. I suspect that discrete "0" and "1" can be store with slightly less tape than an analogue sample point due to the increased signal to noise ratio but it would still require much more tape for something no one understood and the mechanical issues of tapes would remain. It's the same with film. Although there is no resolution definition people chose their films carefully as they knew that an ISO100 ws finer than an ISO800 but ISO800 would allow photos in low light that would be pointless with ISO100, they never put a number on it but it was the same concept of resolution we have today because it was about the size of the crystals that defined the smallest detail that could be recorded so it was very akin to resolution but because the crystals were not an exact number per unit length nor were they in a defined grid like pixels they were just thought about as how grainy they were but it was the same concept. I started shooting on film as i could not afford a DSLR but i could afford a scanner that i still have that can scan 35mm film as a 50Mp image, obviously that image may or may not be usable as a 50Mp image it depends on the "speed" of the film and how grainy the image is. |
| Cerebus:
--- Quote from: Ed.Kloonk on December 18, 2020, 09:24:46 am ---The other misunderstanding with CD's sounding better is in the compression and equalization applied the signal. It wasn't as easy to apply the same techniques on to Vinyl. LP's had a RIAA eq curve already applied so the bass didn't make the groove too phat. When 90's doof doof came along they had to moove to 12" singles so they could be played loudly at discos, and other reasons. --- End quote --- I'd flip that around the other way. With vinyl it was necessary to very carefully apply compression to actually succeed in cutting a record. The guy with the fanciest compressor/expander/limiter was always the guy running the cutting rig. As well as the rather obvious aspect that if you have too large a waveform you can literally cut into the next groove there were also slew rate limits as there's a limit to the velocity that a stylus can track. Also don't forget that we're talking stereo here so you're cutting different signals into both of the side walls of the groove, so there's a depth of cut aspect as well as a width of cut aspect. That in turn means there are differential limits on how far the two tracks can deviate from each other. Then there's trying to trade off available track width with the length of music that you can record on the side of a disc. If you compare a vinyl record of a full orchestra piece with a wide dynamic range and the CD of the same you'll find that the vinyl version has been hugely compressed. |
| Cerebus:
--- Quote from: Simon on December 18, 2020, 10:24:57 am ---I would expect that using a tape for digital storage would have been expensive at the time. I suspect you would need more tape as now the resolving power of the medium that would determine the "sample rate" would need to be higher. Although analogue is not dealt with in terms of resolution or datarate because the medium mechanisms are too fuzzy to put a number on there is a physical limitation. For example we used to record loads of films on VHS long play to save money, the quality was visibly less as we were at the limit of the medium but we accepted it and the reason the standard speed of tape was twice long play was because that was what was required to achieve the effective data rate. If you want to digitise sound then what on the tape used to represent 1 "sample" now would be one bit in 32? so you would need 32 times the tape not to mention more expensive equipment. I suspect that discrete "0" and "1" can be store with slightly less tape than an analogue sample point due to the increased signal to noise ratio but it would still require much more tape for something no one understood and the mechanical issues of tapes would remain. --- End quote --- Ah, you need to go and talk to Claude Shannon. In terms of Shannon Information Theory you can record the same amount of information in the same available space whether you're going digital or analogue. --- Quote ---It's the same with film. Although there is no resolution definition people chose their films carefully as they knew that an ISO100 ws finer than an ISO800 but ISO800 would allow photos in low light that would be pointless with ISO100, they never put a number on it but it was the same concept of resolution we have today because it was about the size of the crystals that defined the smallest detail that could be recorded so it was very akin to resolution but because the crystals were not an exact number per unit length nor were they in a defined grid like pixels they were just thought about as how grainy they were but it was the same concept. --- End quote --- Actually the imaging professionals did think in resolution terms. If you go and look at old film datasheets you'll see resolving power quoted in "line pairs per millimetre" or some equivalent. Here'a a modulation transfer function curve from the datasheet for Kodak PLUS-X Pan film. Full datasheet here. So with that having an MTF of 100% at ~20 cycles per mm, you could say that was roughly the equivalent of 40 pixels/mm. (i.e. Usual Shannon limit rules, sample rate twice the maximum spacial frequency.) |
| Ed.Kloonk:
--- Quote from: Cerebus on December 18, 2020, 01:24:04 pm --- --- Quote from: Ed.Kloonk on December 18, 2020, 09:24:46 am ---The other misunderstanding with CD's sounding better is in the compression and equalization applied the signal. It wasn't as easy to apply the same techniques on to Vinyl. LP's had a RIAA eq curve already applied so the bass didn't make the groove too phat. When 90's doof doof came along they had to moove to 12" singles so they could be played loudly at discos, and other reasons. --- End quote --- I'd flip that around the other way. With vinyl it was necessary to very carefully apply compression to actually succeed in cutting a record. The guy with the fanciest compressor/expander/limiter was always the guy running the cutting rig. As well as the rather obvious aspect that if you have too large a waveform you can literally cut into the next groove there were also slew rate limits as there's a limit to the velocity that a stylus can track. Also don't forget that we're talking stereo here so you're cutting different signals into both of the side walls of the groove, so there's a depth of cut aspect as well as a width of cut aspect. That in turn means there are differential limits on how far the two tracks can deviate from each other. Then there's trying to trade off available track width with the length of music that you can record on the side of a disc. If you compare a vinyl record of a full orchestra piece with a wide dynamic range and the CD of the same you'll find that the vinyl version has been hugely compressed. --- End quote --- Indeed. |
| Navigation |
| Message Index |
| Previous page |