That's a little silly. Not long ago SRC was an issue, today though with such as SOX or iZotope 64 bit processing, it is likely much better than running a second recording chain.
IMO, jus as big an issue would be the question of comparability between two different A/D's if running "parallel recordings". Though I do understand the argument, and presumably each A/D would be "optimized" for it's technology...
At the very least resample vs parallel recording chains, I would say the compromises are similar, and relatively inconsequential one vs the other in the end result.
There are indeed studio DSP plugins for just such things (tube, vinyl, coupling transformer) not over used they can sound "nice" (no further comment)
FWIW, I've done the comparisons myself with some recordings... a little bit, with my "modest" TT and "modest" digital front end. Sometimes it's difficult to tell much difference. Sometimes its easy. There are so many factors as has been discussed in this thread. - not the least of which is modern mastering practice. Even between LP's afficianado's argue the merits of various pressings and mastering work...
I suppose that's what leaves so much room for all the fun discussions 😀
Excellent stuff, thanks. His last point is key- you can do the same thing in software, but someone has to spend the time to do it right.
Completely NOT silly. How many CDs do you own that were down-sampled to 44.1K with SOX or iZotope? Not likely that the liner notes will tell you.
And that's the whole point. When most people judge "does CD sound better than vinyl?" they are not simply judging the technical merits of the media, but those merits via the recordings that they bought on those media. How were those recordings transfered from the master to the version you bought?
We've talked a good bit about the vinyl mastering end of things, but not as much about the CD mastering and pressing end.
When someone records an LP onto CD at home, the process is not the same as a commercially produced CD. Maybe it should be, but it's not. It's a much simpler process with far fewer steps. Though one might hope that creating a CD from an analog master tape ought to be about as simple. Not having done this for commercial release I don't know if it is, but I doubt it.
To take an example from the fine art printing business in which I used to work. I had a few online arguments with amateur image tweakers about how hard it is the get a great scan of a painting. They protested "But I just made a scan of a color page from a magazine and it turned out great." Well, yeah. It's already been processed, tweaked, compressed and reduced to 3 colors + black. It's been predigested, so to speak. An oil painting is not such a simple thing, in fact it's vastly more complex and subtle. Getting accurate scans of bone, ash, minerals, oils, flowers, natural fibers, etc. would be a lot closer to scanning an oil painting (and even acrylic) than scanning a magazine page.
Recording something predigested is a lot easier than recording real life. And yes, the LP recording has been processed to live within the restraints of the medium. Recording it is not as hard as recording real life.
Why does that matter? The question (for me, at least) is how well do the different formats - and the processes used to get the recording onto that format - handle real life? What has to be done and is done in practice, to not exacerbate the flaws of the particular medium you are using? And how does that sound?
This sounds like angels dancing on the head of a pin. Are we really saying that there is some mysterious discrepancy between the theoretical knowledge of those who design 'ADCs' and those who design resampling algorithms? Are we suggesting that a dedicated 16/44.1 system is better than a 24/192 with resampling algorithms? I would have thought it was the other way round, and that sampling & mixing at higher resolution, then dithering, re-sampling etc. down to 16/44.1 was the way to go. Any reason to suggest otherwise?
That would be one experiment, but not a complete one. Resampling between rates that aren't simple multiples is important, because it gets done a lot.
Like I said 50 pages ago, people who prefer LPs prefer the distortions they add. Doing blind testing will only tell you what people prefer, not what is better. Plastic stamped from a piece of metal that was stamped from a piece of metal that was turned on a lathe. Every stamping gets worse. Heres an experiment for the LP lovers. Get a few copies of the same LP at various stages in the wear of the the die, and see where in the process you get the perfect amount of that wanted distortion.
You are starting to understand. Almost. Not sometimes a bad facsimile, always a bad facsimile. That is because there is no known way to capture most of what you heard and no known way to reproduce it if you could. Over 95% of what you heard was due to reflections off the surfaces but the microphones were placed close to the instruments and singers. If they weren't what you'd hear from speakers would sound like they were playing inside a tunnel and you were listening from the outside of it. Play a binaural recording though headphones and it sounds like it's coming from inside your head. The technology doesn't exist yet, the knowledge of how it works and how to engineer something like it isn't around yet.
Does this mean we can't enjoy recordings? Of course not. But it is crazy to think that if you spend whatever sum of money you have whether it is a thousand dollars, a hundred thousand dollars, or a million dollars on audio equipment you will hear from it what you heard in that church. This is why the mindless pursuit of ever more expensive audio equipment chasing the holy grail is ludicrous. A $20,000 phono preamplifier, a $50,000 audio amplifier, a $100,000 pair of speakers, endless tweaks like expensive cables. It's like trying to build a horse drawn chariot that is streamlined enough to fly. Won't happen. Not that way. And what a market there is for those who want to expolit hobbyists who won't accept or believe it. It's like selling fake cancer cures only it's legal.
I listened to 3 phonograph records today. First time in years. Turntable and cartridge worked flawlessly. Sound ranged from awful to mediocre. Best of the bunch was Bruno Walter's reading of Dvorak's Symphony No. 9 (New World.) Spectacular performance. The dynamic compression was obvious. No pops or clicks either. There's no way to capture anything like the dynamic range of a symphony orchestra on a vinyl phonograph record.
😱 That all too mysterious 'mystery' rears its head again. Bottom line is it's not relevant to ponder the theoretical knowledge of the designers rather the resultant products that come out. Werner's already demonstrated that plenty of commercial SRC software sucks, but OTOH lots of hardware designs use half-band FIR filters. Pick your poison.
These days professional ADCs work at a multitude of sample rates, 44k1 being normally the lowest. The easiest way to engineer an ADC is have it sample at the highest rate by default then decimate down if the end user asks for a lower rate. That keeps all the switching of rates under software (firmware) control. Switching between 44k1- and 48k- multiples may well be done in hardware though. So in the end its a toss up between the filters designed by the hardware guys or those from the software boys.
Quite the opposite in my own experience - binaural recordings are the only ones that don't sound like they come from inside my head. I find I sometimes instinctively turn my head to localise the sources - uncanny.
Yes, but those distortions might be more pleasing to the ears because it sounds more realistic to us. I'm more interested in the nature & type of distortions that give us this better illusion because if we understand this & apply it scientifically we can hear something more like the real performance or more immersive. What's wrong with this? By studying vinyl replay or SET amplifiers or whatever might be a way of understanding these distortions.
Here's JJ's suggestion to how measurements might be refined to evaluate audibility differences
That is the instinctive thing to do, turn your head. But with binaural recordings played through headphones the sound turns with your head. That's why it appears to come from inside your head. If it worked, that would be the system of choice. Some have tried desperately to adapt binaural recordings to loudspeakers. That's what ambiphonics seems to be about (not ambisonics.) Then they added DSP processors and surround amplifiers and speakers. I've never heard it so I can't say what it sounds like but I haven't seen it taking over the high end audio market. It appears to have been around for at least 15 years. A sort of curiousity.
Nope, stays resolutely outside my head even when I turn it.
Would it be too much to ask for a cite on the 'desperation' aspect of this?
Ambiophonics by Ralph Glasgal -- Figure 4
Look at figure 4. Somewhere there is a photo of Ralph Glasgal's original setup with a real wall between the main speakers which are right next to each other. You put you head up against a cushion that's part of the wall like one of those 19th century optical trick toys where one eye sees a bird and the other eye sees a cage and your brain makes it seem like the bird is in the cage. Glasgal tried all manner of high end speakers, amps, processors. I think he even worked with one company to design processors along his own ideas.
There are others. One scheme has multiple binaural tracks at different angles and an accelerometer on the headphones to change tracks when you move your head so it seems like the sound isn't moving with you.
There are more. I can't cite them all. The idea has been around for over fifty years ever since two track tape recorders became available. Even Lafayette Radio had a binaural microphone arrangement offered around 1960 while JVC had microphoones built into headphones that allowed you to make binaural recordings while you were listening through the phones. This in the mid 1970s.
I think an aerospace engineer in Preinceton Chouarti or something like that name has a similar idea using speakers with his own twist. Can't find the link though.
You've never actually done what you're describing have you? Well I have and I assure you it does NOT sound like "playing inside a tunnel". I've never heard a formal binaural recording but I doubt it sounds like it's coming from inside your head. The only time I've heard something that seems to be originating inside my head was with a phase turnover I.E messed up due to incompetence.
G²
Here's a link to choueiri's video demo
Video: Princeton engineer makes 3D sound a reality ? Princeton Engineering
Video: Princeton engineer makes 3D sound a reality ? Princeton Engineering
- Status
- Not open for further replies.