Hi everybody, I have been thinking about how we now have arrived at Sound reproduction being mostly MP3, how about doing it the way Edison did it, would we with help of better materials be able to record and reproduce Sound in a natural sounding quality with a reasonable loud output?
Probably. An analog recording theoretically contains a near-infinite amount of data, making old records, according to some fans, superior. Not really. Most of that data is pops, clicks, and the grinding sound of a needle being dragged through a groove that has felt too many needles.
Centennial Songs - The Antique Phonograph Music Program with Mac Playlists and Archives MAC covers the acoustical era, when a singer’s voice drove the stylus. Even turned into MP3s, I think there are layers separating the recording and the playback that are missing, so you have something closer to an historical document.
Somebody makes a cylinder recorder that uses plastic cups. Still looking.
Well, your problem might be with what you define as a reasonably loud output - a gramophone-style horn will only go so far, and will also not help the sound quality. As a general rule you can say that the louder you want it to be without electrical amplification, the shittier it is going to sound.
That’s gonna be the deal-breaker, I fear. My guess is that even the very best acoustic horn won’t bring the volume up higher than conversational speaking, and certainly not loud enough for a party. But that’s just my guess – hard data, anyone?
Yeah, making it loud without electricity is going to be very difficult.
There is such a thing as a mechanical amplifier, but these types of things usually work at much lower frequencies than audio signals. Getting something like that to work at audio frequencies would be a bit of a difficult engineering challenge, to put it mildly.
How about making one master recording, copying it a bunch of times, and playing all of the copies on a bunch of different record players at once? Care would have to be taken to synchronize them properly, but that shouldn’t be too hard.
Synchronization is easy. You just use gears.
You could also take your original recording use a mechanical amplifier to make really big grooves out of the really small grooves you started with. You wouldn’t be able to do it in anything close to real time, but that would allow you to make several louder recordings. Synchronously play back all of the louder recordings in real time and you’ve got a louder sound.
You will also get a lot of noise and signal distortion introduced at every stage of the process, but hey, that’s what old fashioned analog recording was all about!
if the way Edison did it was so great, we’d still be doing that. unfortunately, the key parts of a gramophone have too much mass to provide full range. Further, acoustic horns have a cutoff frequency below which they’re useless. If you’re happy with scratchy audio with no bass whatsoever, played back from a medium which is gradually destroyed by the implement used to play it back, then knock yourself out.
something isn’t more “natural” just because electronics were kept out of the chain. OK? if the medium can’t capture the sounds to be recorded as-is, then it’s not natural. hell, even vinyl can’t do that; if it could, there would be no need for RIAA equalization.
jz, I don’t think the OP is under the impression that it would be desirable; he’s just curious as to whether it’d be possible.
I need to see one!
I’m still not convinced that Vannevar Bush is not related to our presidents, except he was smarter than Dubya. The family resemblance is too strong.
MAC suggests you zero out your bass and treble because there will be neither in his broadcasts. The recording range of acoustic records was worse than a modern telephone. I love it because it’s old. I don’t love it because it’s accurate.
I heard somewhere (how is that for a cite) that a calliope was one of the loudest pre-electronic instruments - though how to record and then play back sound with one eludes me.
A pipe organ can sythensize many sounds - could you use a steam powered pipe organ (for the compressor) somehow?
Record each frequency separately (I’m thinking a tuned pipe for the recording also - you could record on film if you wanted)
Then the intesity would control the loudness of the particular key (I’m not sure how loudness is controlled in a pipe organ)
Just some brainstorming
Brian
A steam calliope is still one of the loudest musical instruments ever invented, and I’m thinking a carillon would be at second loudest. I believe either can be “programmed” via some variant of punch cards, so either instrument could readily reproduce a performance.
As for acoustic horns, my Victrola can play surprisingly loud. Actually, it can play annoyingly loud to the point of needing to close the doors (a primitive form of volume control) if you want to be in the same room with it. The one thing I don’t know is whether it’s truly putting out a lot of volume, or if the tonal quality just has a lot of “punch.”
I am curious if the old grammophone devices had some kind of volume control?
Sort of. On mine, if you leave the upper set of doors closed, you’ll get minimum volume, and opening the doors yields maximum volume. The “horn” in this design is folded inside the cabinet, rather than being external.
I think with external horns, about all you can do is stuff a sock in it.
That’s a great look at the mechanics of the machine. Wow the needle looks like a nail!
A pneumatic amplifier is a possible answer to getting arbitrary loudness out of a phonograph. In essence they are a very sensitive valve that is modulated by an actuator. Coupled to a horn and you could remove paint. (Apparently a version fed from turbine bleed air was mounted on helicopters. The scene in Apocalypse Now where they play Ride of the Valkyrie from a helicopter was played though such a system.)
One might well be able to design a quite reasonable device with the application of some serious CFD codes to the problem.
On the subject of resolution, and analog having “infinite resolution”. This comes up quite often, but simply isn’t true. Whilst there is a sort of intuitive idea that digital has “steps” and therefore “finite resolution”, and analog doesn’t have steps, and thus has infinite resolution, information doesn’t work this way. All forms of sound transmission are subject to exactly the same laws, and the resolution of all of them is curtailed in an essentially identical form. The key determinant for resolution is the signal to noise ratio. In the normal world everything has noise, in the limit thermal noise is unavoidable. In electronic systems you have Johnson Noise. In audio you are at base limited by the thermal noise of air molecules banging around. The self noise of microphones is so limited. So much so that larger diameter microphones have lower self noise than smaller because they average out the randomness of the noise over a larger area.
Digital systems are limited by their quantisation level. This forms the equivalent of the thermal noise floor in an analog system, and indeed where the boundary between the two exists they are treated as a common and critically interrelated issue.
The other determining metric is the bandwidth of the system. For audio we typically use 20kHz, since humans can’t hear any higher. Once you have the signal to noise ratio and the bandwidth you exactly determine the information capacity. The capacity calculation is identical for both digital and analog systems. This is Shannon’s Theorem.
There are some counter-intuitive tricks in here. You may note that thermal noise does not seem to create a hard limit to one’s ability to hear a signal. It is quite possible to hear a very weak sound that has a level below the noise. What is often not appreciated is that this isn’t a feature of just analog systems - quantised systems (i.e. digital) are also able to resolve signals below the quantisation steps. A first sight this seems ridiculous - the signal was quantised - surely that throws away any information at a level lower. What this misses is that the quantisation occurs (must occur) in the presence of true random noise. This noise modulates the quantisation, and information about processes at a level lower than the quantisation threshold become part of the digital signal. This mechanism exactly obeys Shannon. Indeed this synergy between quantisation level and noise level is so intertwined that a digital system that does not work this way will fail to work properly and will suffer form a number of unfortunate distortion mechanisms. The term that is usually used for this is “dither.” When correctly implemented the resolution of a digital system turns out to be exactly the same as an analog system with the same noise floor, including the ability to hear weak signals below the noise floor (and the same limitations on what weak signals can be heard and under what circumstances.)
One reason that early records had a high rotational speed is that all the sound energy came from the groove. The higher the linear speed of the groove, the more volume you could get out of it. Pathé had public address system that used 20-inch records that spun at 120 RPM. Few of these records survived - I’ve only seen pictures of them. I’ll bet they were loud.
As for sound quality: it would probably be possible to use modern technology to make better-sounding acoustic recording/playback systems than they had in the twenties (when the transition to electrical recording happened), but there’s no way it could be made to sound as good as a modern electrical/digital recording. Achieving a flat frequency response, extended treble and bass, low harmonic distortion and good transient response is very difficult when the signal has to go through horns, diaphragms and cutting needles.
I was imagining a recording of an orchestra where every musician is recorded separate and then replayed using one player for each instrument at it’s original place in the orchestra.
If you would scale up the recorder, would this not also increase the loudness?
This would not be recording as I meant it.
Say we have a very though material as a recording medium, and we reduce friction to be near zero,
this should increase the clarity of the recording, the question is how to amplify it, some of a Pentograph? for sound.