Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
In Reply to: RE: Several problems digital has that analog seems to not have: posted by Ralph on April 17, 2017 at 13:18:04
A subjective proof of a medium's faithfulness to any signal, is given by a generation loss test as was normal to do in the days of audio tape.
The argument being the more faithful the process was to the original signal, the more generations the music could survive without transmogrification.
Since the errors accumulate with each generation one gets a caricature of "what's wrong" and at work, this proved to be a useful tool developing loudspeakers.
My point is, here is a way to prove your argument subjectively.
For instance one could do a low respolution digital copy of an analog tape recording and on the first pass lose an aweful lot of audible information. But...one can also take that low res digital copy and do a thousand generational copies with zero degradation. OTOH in the analog domain one can do a one to one copy on analog tape and have the first generation be virtually indistiguishable from the original but if one were to make a thousand generations of that original recording in analog you'd have an end copy that was far far more degraded than the low res digital copy. This test just doesn't hold up when condiering digital media. different rules
I don't think so.
Here's the problem, which has nothing to do with how media has or has not generation loss:
The human ear/brain system converts all forms of distortion into tonality. It only detects distortion as such when it predominates, such as clipping in an amplifier.
Otherwise what the ear does is assign tonality; in most cases this usually results in the playback being "brighter" than the original (as an example, the 2nd harmonic often associated with tubes causes 'warmth' or 'lush' qualities).
However if one were to measure the frequency response, the 'brightness' would not show up in the frequency response test. That is because its not the result of a frequency response error!
The reason it won't show up is that the additional harmonics (or inharmonics) only need to be in trace amounts. This is because the human ear is tuned to birdsong frequencies and also because higher ordered harmonics are used by the ear/brain system to gauge sound pressure. As a result, in this regard the ear is more sensitive than the best test equipment.
The audio industry in general likes to ignore this fact; this is why most audio systems sound like audio systems rather than real music.
So 'A subjective proof of a medium's faithfulness to any signal' has nothing at all to do with generation loss, and everything to do with not making distortion. It does not matter if you can duplicate endlessly if the master recording is distorted. Right now the LP does that better than digital...
In a nutshell, this is the cutting edge of digital technology. We are pushing of course for greater scan frequencies and more bits, but if the basic problem of inharmonic distortion is not solved, then analog will continue to be around. That's the bringing home the bacon aspect that digital has not solved.
Mind you, its a lot better now than it was! So I have hope for the future, but from what I've seen of the industry so far, it regards this matter as trivial because its pulled the wool over its own eyes.
I am not sure that there would be any kind of alteration to the signal, that would not be captured and then when fed back as the input signal on the next generation, accumulate.
The ONLY thing that doesn't accumulate is the original unaltered signal. The result is a caricature of what's wrong with the original signal.
In an old days analogue tape system, each generation accumulated noise and whatever distortion present was also made more obvious each pass. Digital is not immune to generation loss if even one bit or timing is altered.
Fwiw, back when CD's were first introduced, at least at the studio level, they were intentionally made a bit brighter to drive home that they were "clearer" and "better" than lp's.
I think one can see this in the tests Floyd Toole did which shows what the listener preffered loudspeaker response shape is and guess what it is, about a -1 dB /oct roll off broad band (which over the entire band would imply a -10dB reduction at 20Khz) compared to "Flat response " which used to be the target in analogue days.
Record's have another thing CD's don't have, the cartridge on the record is also a microphone of sorts, it's picks up some amount of room sound while playing the record and that goes back for another pass through the system. It isn't much but you can hear it if you put one speaker in another room, put on an old record but not rotating and set the needle down on the lead in. Then turn the volume up to a normal level and have friends talk or the tv on so you can hear how much is coming out of the speaker in the other room. After doing this in the old days, I ended up putting the Thorens on a granite slab on springs and using a brass weight to keep the record down as much as possible.
Consider too that by any measure you choose, loudspeakers are by far the weak link so far as reaching realism, even when abundant in a recording. For example you can have two sets of loudspeakers with very similar frequency responses and yet with one pair, a mono phantom floats solidly in the center and you are not even aware of a right and left source. With the other, there is a phantom image if you're in exactly the right position and the right and left sources are obvious. What properties could be behind this kind of difference, one that has a profound effect on imaging and not on timbre or frequency response?
Record's [sic] have another thing CD's don't have, the cartridge on the record is also a microphone of sorts, it's picks up some amount of room sound while playing the record and that goes back for another pass through the system. It isn't much but you can hear it if you put one speaker in another room, put on an old record but not rotating and set the needle down on the lead in. Then turn the volume up to a normal level and have friends talk or the tv on so you can hear how much is coming out of the speaker in the other room. After doing this in the old days, I ended up putting the Thorens on a granite slab on springs and using a brass weight to keep the record down as much as possible.
That's not a fault of the media, its a fault of the playback apparatus, which clearly has problems! Its not that hard to solve that one BTW- my 'table has no issues with that at all.
Its the initial encoding of the audio which is an area where the problem exists. Once there, sure you can copy it endlessly (much to the dismay of the music industry) but the damage was done on the initial encoding and does not seem to be something that the consumer can fix with better gear (although that certainly helps).
Don't get me wrong- I'm not against digital. I want it to work. I'm just pointing out that when the industry does not think it has a problem, it won't do anything to fix it. That's the case and that is why no matter how much we talk about it, the LPs are still around decades on. They're old but not obsolete.
It was no contest, LP was better than digital in the 80s. But vinyl is still around due to a hobbyist movement. Digital sources, DSP crossovers, etc. are doing the real work and taking us closer to the elusive goal of a live mike-feed.
Look at the measurements of LP -inc. speed stability and explain how this could be a high-resolution system. How about inner-groove distortion ?
Strange -as LP 'improved', here comes tape ! The reference (now) for some audio writers.
Even more scary was the finding that 45 and 78rpm records are better sounding than LP. Robert Harley and Jacob Heilbrunn (of TAS) wrote blogs on systems that used 45's (as a source), not LP. Herb Reichert recently made comments in line with these demos.
I had no idea analog recording/LP playback were that bad. Maybe I'm doing too much reading. But most reviewers have given up LP, so my findings can't be too far off...
I believe the change is partly because so much of music is mixed to be played on ear buds and low rate mp3. The recording industry's loudness wars are responding to th every low dynamic range one has in the car or limited systems, they raise the average volume without raising the peak level with compression.
In other words, it is the companies who in effect decided and determine what is popular and implemented through mass marketing. The folks here are the 1% who have been known to sit down and listen as opposed to having musical background entertainment.
Anyone who is working in that direction gets my vote.
Re-written, but high end *is* a movement. Just has no name.
J. Gordon Holt thought of it as a movement, wanted accuracy and truth, and never listened to LP as such.
Read his interviews -he used tapes as a reference.
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: