A Reel Person
It's Too Funky in Here!!!
Dogs...
Dogs can hear ultrasonics!
Dogs can hear ultrasonics!
Why analogue and not digital?
- Thread starter cjacek
- Start date
apl
Stand Up Comity
And how much of it is aurally available during analog reproduction.
evm1024
New member
Thats begging the question....
But to answer "none" in digital (cd's) and quite a bit in tape.
Also, come on! Here is a paper that shows humans do perceive above 20 kHz. Let's put to rest the myth that we do not. We are talking about scientific truth here.
Step up to the plate and say "Humans perceive above 20 kHz and I do not know how significant that perception is. That is a scientific truth."
Last edited:
apl
Stand Up Comity
I don't believe I've ever said otherwise.
mshilarious
Banned
I am with you there. I was impressed by that too.
But it's only a first step. Follow-up work is needed to quantify what is required. For instance, they used two different cutoff points, at 22kHz and 27kHz, for different tests. The results seemed similar (actually I was confused why they had done those differently), but it didn't address what the upper limit should be. Practically, it has to be 40kHz, I don't think there is a commonly available playback technology that will support anything higher.
In the meantime, we ought to be pressing the double speed button on the converters, which I haven't minded anyway.
mshilarious
Banned
Yes.
But let's keep in mind consumer format vs. production format. Now that we know this, we ought to try to include it in production.
It's hard to say what will make the consumer care. Hannah Montana ain't a gamelan, and her fans don't listen in an anechoic chamber. For the specialty type of releases for audiophiles, they already have ultrasonic formats with presumably sympathetic production.
That's not an analog vs. digital issue per se; you gotta do your best engineering and hope someday somebody will care.
apl
Stand Up Comity
I have a question, a thought experiment.
A CD has nothing above 20kHz.
Analog has some information above 20kHz.
An artist records a work. All the tracking, mixing, effects, and mastering are done in analog. It then gets put on a CD.
It has been noted before that such a process results in a better product, and I'd agree. Kind Of Blue on CD sounds wonderful, as does the Love project. And the new Frampton album. They have that analog feel.
But we're listening to a CD! How can the analog grooviness be something that's within the 20-20kHz bandwidth?
A CD has nothing above 20kHz.
Analog has some information above 20kHz.
An artist records a work. All the tracking, mixing, effects, and mastering are done in analog. It then gets put on a CD.
It has been noted before that such a process results in a better product, and I'd agree. Kind Of Blue on CD sounds wonderful, as does the Love project. And the new Frampton album. They have that analog feel.
But we're listening to a CD! How can the analog grooviness be something that's within the 20-20kHz bandwidth?
theblackBay
Mean average best idiot..
ok
yeah zero involving my self in this particular argument.. (Why are you posting then?)…
well, because apl said something there that i said a long time ago to someone:
the Analogue all the way through until CD scenario so my argument was from that point on, generally and for 90% of the population, the analogue argument was essentially an EQ argument.
that is to say digital equipment in the early days was too accurate and transparent and has since “smartened up” to include the "malfunctions" of analogue equipment, "malfunctions” that for “binaural” reasons were nice to a human.
humans adapt however, and in future maybe everyone will be saying:
wow, look Jeff your so Dumb! I don't care what Nano transient polymer membrane you used!, it sill doesn’t have that Nice square old Digital Distortions tone!
I love both, analogue is for the people, but digital technology freed human creativity, so it is the peoples.
btw i own some analouge equpment so don't write me off, it's just market v monopoly.
yeah zero involving my self in this particular argument.. (Why are you posting then?)…
well, because apl said something there that i said a long time ago to someone:
the Analogue all the way through until CD scenario so my argument was from that point on, generally and for 90% of the population, the analogue argument was essentially an EQ argument.
that is to say digital equipment in the early days was too accurate and transparent and has since “smartened up” to include the "malfunctions" of analogue equipment, "malfunctions” that for “binaural” reasons were nice to a human.
humans adapt however, and in future maybe everyone will be saying:
wow, look Jeff your so Dumb! I don't care what Nano transient polymer membrane you used!, it sill doesn’t have that Nice square old Digital Distortions tone!
I love both, analogue is for the people, but digital technology freed human creativity, so it is the peoples.
btw i own some analouge equpment so don't write me off, it's just market v monopoly.
mshilarious
Banned
I think we have to accept that the basic sonic difference between tape and digital is audible spectrum. Or at least we should postulate it, because the difference seems to be apparent quickly to listeners who express a preference, whereas that was not necessarily true in the ultrasound study--it seemed to take more time for that perception to reach the level of cognition.
Anyway, ultrasound is really a separate topic, valid for consideration in recording media, but at this point rarely realized in a consumer format/playback combination.
I was struck by the quality of the "Kind of Blue" recording listening on CD on a boombox with a "full-range" driver, so I know that wasn't ultrasound. I don't know if I'd attribute it to tape, either; it's possible, but I have plenty of AAD or ADD CDs (most of my collection, probably, acquired in the late '80s when everybody was quick to reissue older classical recordings on CD), yet that CD really stood out. It's a much more recent remaster than most of my collection, and sympathetic and knowledgeable use of improved digital converters could be part of that.
But I suspect it has more to do with microphone selection and placement, and live recording in a good room. Of course, microphone selection was simpler in the days of deciding which U47 to use . . .
And no, it isn't just the fact that it's Miles and Coltrane, because I never liked any of Miles' later period stuff.
Did you see this post:
https://homerecording.com/bbs/showpost.php?p=2930409&postcount=22
and follow the link to the PSW thread? I hadn't seen the PSW thread before, but independently three years later, I post here (yes, on the Analog board!) that I thought "Kind of Blue" was a benchmark, and the multitrack was the beginning of the end. And there it is . . .
I also liked that one guy that got his head handed to him, in a nice way, after he said it was all tube gear that made the difference, and the transistor ruined audio . . . and then somebody tells him all the records he listed were done on solid state consoles
I like tubes too, but life is a little more complex than nostalgia . . .
Tim Gillett
Banned
Exactly. So I wonder why it gets dragged into this thread about analog vs digital.
If the criterion is super wideband response, digital audio recording (even using magnetic tape) creams analog magnetic because of one crucial factor ; inductance. A little detail called a magnetic tape head. Direct analog type recording stretches the magnetic head to its electrical limits already. Add a few more octaves and we have problems.
Regardless of the Redbook CD, these days, even home recordists can buy a modest PCI card that is basically flat to 80khz (I have one).
If ultra wide band is people's bag (personally I think they are misguided) then I'd say definitely go digital recording.
Cheers, Tim
Last edited:
famous beagle
Well-known member
The reason I brought it up is that you made a blank statement about the limits of human hearing, and I don't quite think that holds water.
famous beagle
Well-known member
evm1024
New member
Actually
Actually we are not agreeing this.....
Before anyone gets their feathers up let me continue.
"Ultrasound" was not the cause in the paper per se. It appeared to me that the authors noted that there might be a detection mode in humans but also that the interaction of the ultrasound and souunds in the "normal" hearing range of humans.
Further, I'm noting that we are sliding back to mis-information: Ultrasounds don't matter, it's the artifacts of tape that make it sound better etc. We cannot state those things with any degree of certainty which makes them mis-information if stated at a fact. (not picking on anybody here just trying to not slide back)
We have an open conversation at this point in which we are looking honestly at the question "Why Analogue and not digital". We have bashed through some older thoughts as to how digital worked and tossed the false aside (do we need to rehash that?)
So let's move forward and not lose the gains we have made.
Actually we are not agreeing this.....
Before anyone gets their feathers up let me continue.
"Ultrasound" was not the cause in the paper per se. It appeared to me that the authors noted that there might be a detection mode in humans but also that the interaction of the ultrasound and souunds in the "normal" hearing range of humans.
Further, I'm noting that we are sliding back to mis-information: Ultrasounds don't matter, it's the artifacts of tape that make it sound better etc. We cannot state those things with any degree of certainty which makes them mis-information if stated at a fact. (not picking on anybody here just trying to not slide back)
We have an open conversation at this point in which we are looking honestly at the question "Why Analogue and not digital". We have bashed through some older thoughts as to how digital worked and tossed the false aside (do we need to rehash that?)
So let's move forward and not lose the gains we have made.
Tim Gillett
Banned
I dont think anybody is saying we should not move forward, or lose the gains made. On the contrary!
Cheers Tim
cjacek
Analogue Enthusiast
Ethan, if anything, the above gets closer to it. I wasn't sure how to phrase it initially but this makes it easier.
All along I had speculated, from what I knew of sampling and such, that stuff, which many digital folk deem unimportant, gets 'thrown out', in the process of D / A conversion. This 'stuff' may be the 'glue' which holds together or brings coherence to the sounds "in the 'normal' hearing range of humans".
I mean, lets looks at a reasonably encoded MP3 and the CD, again.... To me, at least and on paper [on the screen, to be more exact], all of the major frequencies get represented and most important, when listening back to the MP3 and CD [of the same track], one can't really spot an absence of the immediately heard frequencies. The highs are there, the lows and mids too etc..... BUT there is a significant loss in the dimensions of the track, when you go to MP3. Again, all the frequencies seem to be there, as high / low the human ear can hear BUT you get a sense of how 'small' the MP3 is as compared to the CD. Interestingly, the lower the bitrate [going from CD to 320 bit to 'variable' to 256 to 225 to 192 to 160 to 128 etc...], the smaller the stage of sound, like it goes from multi dimensional, down to a single dimension, 'flat', tiny sound, despite many of the bitrates preserving the readily heard 'frequencies'.
An interesting thing happens to tape..... Remember when I said my cassette portastudio [no, not a cheap version] sounds better than CD, despite not necessarily being able to capture the highest of audible frequencies? Well, the reason that it sounds better, IMO, is because in the range that it does well in, it seems to capture what digital sampling omits ... and that is [to give a simplified analogy] the entire ocean under the surface [if you know my drift]. It sounds 3 dimensional with a coherence absent from digital.
The cassette but lets look at tape machines in general, capture all of the audible and inaudible mixture as provided by nature, the core and the glue which holds it together. Digital seems to only care about the 'core', the surface, while everything below is thrown aside, as it's deemed 'unimportant'.
Kinda reminds me how the traditional / conventional medical system is like. It is popular thought to isolate one body part / function, from the rest of the body. Ya can't do that. I think this stems from the lack of understanding. Everything functions coherently, in harmony, together and is meant to, even if we can't measure or don't understand why. That's the beauty of analogue.. We don't have to understand, we just 'know'. Anyway, just my thoughts.....
----
Last edited:
_brian_
Member
I hate to distrupt the flow of posts but I thought I may add a... thought. If someone else has already mentioned it I'm sorry, but in the end it means I'm in complete agreement with you.
I've always thought analogue was better due to imperfections as well. If you record the same track on analogue & a DAW, then you usually find that the DAW has more clarity & high end, while the analogue recording sounds warmer. Now choose what you prefer. But for me that's not it. When you record a series of tracks in analogue they all have the same inperfections as each other. In essence each track has lost some of it's original characteristics and in it's place has recieved a new generic characteristic (freq responce or tape imperfection or whatever). Now track one now shares a more similar frequency character as track two etc. Digital is so much more accurate at duplicating the original sounds that each track really does sound different. This makes mixing the project much harder in my eyes as you will find it more difficult creating a single entity from all of those differing tracks. Analogue (to me) sounds much more organic & free flowing because each track is naturally welded to the next. To be honest I don't like hearing a pixalated song. I don't want to use my brain power up listening to something complex and unfixed. When I listen to analogue stuff I use my brain power trying to decifer lyrics and arrangment which I think is what it's all about. I really do find it less enjoyable and more difficult listening to digital stuff.
Now there is an argument for placing certain frequency boosts & overdrive plugins configured to the same settings in each track to give the whole mix a more organic sound. I've had mixed results with such a method, but not so much success with using analogue or ease was encountered.
I've always thought analogue was better due to imperfections as well. If you record the same track on analogue & a DAW, then you usually find that the DAW has more clarity & high end, while the analogue recording sounds warmer. Now choose what you prefer. But for me that's not it. When you record a series of tracks in analogue they all have the same inperfections as each other. In essence each track has lost some of it's original characteristics and in it's place has recieved a new generic characteristic (freq responce or tape imperfection or whatever). Now track one now shares a more similar frequency character as track two etc. Digital is so much more accurate at duplicating the original sounds that each track really does sound different. This makes mixing the project much harder in my eyes as you will find it more difficult creating a single entity from all of those differing tracks. Analogue (to me) sounds much more organic & free flowing because each track is naturally welded to the next. To be honest I don't like hearing a pixalated song. I don't want to use my brain power up listening to something complex and unfixed. When I listen to analogue stuff I use my brain power trying to decifer lyrics and arrangment which I think is what it's all about. I really do find it less enjoyable and more difficult listening to digital stuff.
Now there is an argument for placing certain frequency boosts & overdrive plugins configured to the same settings in each track to give the whole mix a more organic sound. I've had mixed results with such a method, but not so much success with using analogue or ease was encountered.
This discussion is getting a bit hard to follow. In fact, I didn't even read the last page. But here's some responses from previous pages.
Get an 80-8 or a 58. They're tougher and sound better.
Yes, anybody with basic math skills can calculate how much error is there. But nobody, not man or machine can CORRECT those errors. You can only hope to make the least intrusive guesses possible.
I've done tests like these. The outcome is always the same. Most people favor the live sound most and favor the analogue recording second. But people are so heavily bombarded with false information about digital recording, they usually guess that the analogue recording is digital because it sounds better. I want to run the experiment again though using a PET scan to see the difference in brain activity.
Not that I can tell. A well tuned machine is going to produce less noise than your room itself. Now if you don't have a head shield, the tape can become microphonic and pick up vibrations in the air. No kidding, it's possible to record to tape without even using mics or preamps. But a proper shield over the head stack pretty well eliminates the problem. Punches are a slightly different matter. When you punch IN a correction, you will have the old wave form and new waveform overlap eachother for a period of about 100 ms at 15I/S which is actually quite usefull. Punching OUT a correction will cause a 100ms gap between the new and old takes. This is due to the space between the erase and record heads. So if I punch out a correction, I wait till there's a rest in the instrument so the gap is not noticable.
There IS still information missing at lower frequencies, though I think it's an acceptible amount. I'm mostly concerned with 3KHz and higher.
You can reconstruct it, but it won't be the same as the original. The higher the resolution, the more accurate, but it's NEVER an exact reconstruction unless you're using pure digitally generated sine waves. Hardly a real reflection of music.
No, this experiment shows that you can't trust FFT measurements. In both examples, there's obvious displacement in the time AND intensity domains seen as the conical formations that should be single, needle thin lines in this case.
That comment came out of left field. Nobody will argue against that. I personally prefer not to overdrive anything, hence why I record 400nW/m on GP9.
They exist and can be had for $15,000. Actually, it's whole turntable unit with 4 lasers and an LD-style load tray. Which sucks because I came up with the idea in high school.
That's very true which is why I tend to use digital tools for long term mastering jobs. I can recall settings and it'll sound the same as it did 6 months ago. But I still prefer my analogue gear if I'm going to master a whole album in a single day.
Get an 80-8 or a 58. They're tougher and sound better.
Yes, anybody with basic math skills can calculate how much error is there. But nobody, not man or machine can CORRECT those errors. You can only hope to make the least intrusive guesses possible.
I've done tests like these. The outcome is always the same. Most people favor the live sound most and favor the analogue recording second. But people are so heavily bombarded with false information about digital recording, they usually guess that the analogue recording is digital because it sounds better. I want to run the experiment again though using a PET scan to see the difference in brain activity.
Not that I can tell. A well tuned machine is going to produce less noise than your room itself. Now if you don't have a head shield, the tape can become microphonic and pick up vibrations in the air. No kidding, it's possible to record to tape without even using mics or preamps. But a proper shield over the head stack pretty well eliminates the problem. Punches are a slightly different matter. When you punch IN a correction, you will have the old wave form and new waveform overlap eachother for a period of about 100 ms at 15I/S which is actually quite usefull. Punching OUT a correction will cause a 100ms gap between the new and old takes. This is due to the space between the erase and record heads. So if I punch out a correction, I wait till there's a rest in the instrument so the gap is not noticable.
There IS still information missing at lower frequencies, though I think it's an acceptible amount. I'm mostly concerned with 3KHz and higher.
You can reconstruct it, but it won't be the same as the original. The higher the resolution, the more accurate, but it's NEVER an exact reconstruction unless you're using pure digitally generated sine waves. Hardly a real reflection of music.
No, this experiment shows that you can't trust FFT measurements. In both examples, there's obvious displacement in the time AND intensity domains seen as the conical formations that should be single, needle thin lines in this case.
That comment came out of left field. Nobody will argue against that. I personally prefer not to overdrive anything, hence why I record 400nW/m on GP9.
They exist and can be had for $15,000. Actually, it's whole turntable unit with 4 lasers and an LD-style load tray. Which sucks because I came up with the idea in high school.
That's very true which is why I tend to use digital tools for long term mastering jobs. I can recall settings and it'll sound the same as it did 6 months ago. But I still prefer my analogue gear if I'm going to master a whole album in a single day.
mshilarious
Banned
As apl explained earlier, you can change the smoothing window on the FFT if you like. I posted all the waves so you may do so. If you do, you'll see the needle-point lines you like, along with lots of little -120dBFS spikes. Those are quantization distortions that exist in the original noise-free digitally generated signal. I could have eliminated those by creating the sample wave in 32 bit float and dithering to 24 bit, but I regarded that step as unnecessary, because the downsample and upsample faithfully reproduce those same QDs.
But if you are fastidious, here is a dithered 24 bit file so you can repeat the experiment--the Hanning FFT looks like this:
Those couple little -160dBFS peaks above 19kHz are what you can see of the dither. If I run that on the 32 bit float file, they aren't there. But it's not possible to listen to a 32 bit float file, so I think that is a more realistic FFT.
QD_free_4_sine_wave.wav
I confess I am disappointed you did not review any of the actual recordings I made for audibility of QD. That is a crux of your argument, so it should be easily audible, long before the pipes get far into the fadeout. It was easily audible in the test signal, how about the real recordings? You strongly implied if not flat out stated earlier that 16 bit CDs have less than a 54dB dynamic range due to QD, and the sample I posted should disabuse you of that notion.
There you go again. You haven't posted any audio or analysis that supports your assertion, so instead you go on pretending Fourier was wrong. Actual music is EASIER to reconstruct than a series of high frequency sine waves, which are about as tough a test as you can give a converter AND its surrounding analog circuitry, making those poor opamps slew as fast as they can!
But it doesn't matter, I posted "actual music". Where is the flaw?