Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
The reason to down sample is easy. If a recording was originally recorded in let's say 24/192, you'll at least capture it at a level that SHOULD give you as good of audio quality as you can get and you would down sample it to provide the content in a smaller file size for those will limited storage or in a CD which most people can play, but you could still offer the original 24/192 as a digital file for those that have more storage available and a better system to hear the difference between 24/192 (lossless) and 16/44.1 (lossy) which is MP3/AAC which is what people are buying for mobile devices.

The discussion has nothing to do with lossy compression: in this discussion 24/192 and 16/44 are the bit depth and sampling rate of uncompressed (or losslessly compressed) sound files.

Unless you mean that e.g. the information about ultrasonic frequencies available in the 192 version being discarded in the downsample make the downsample "lossy". You are right that by downsampling you discard information, but this information is outside the accepted limits of human hearing (just like a camera discarding infrared or ultraviolet is not considered to "downgrade" the image quality).

Some record labels are recording at DSD 2x and then if you don't have DSD playback, you can convert to PCM at what ever level you want. But DSD 2x are HUGE file sizes. Can they hear the difference? Well, according to the people involved in listening tests say they can. Is it is huge difference? For the average person, they might not hear that difference, but to those that have better trained hearing with better equipment might be able to hear that subtle difference.

The xiph.org article mentioned multiple times has as basis ABX tests done with SACDs, most of them provided by HD audio enthusiasts as "best examples highlighting the superiority of SACD". Nobody was able with statistical significance to recognise the original version from a 16/44 downsampled version (not even the aforementioned enthusiasts).

When comparing 16 bit vs 24 Bit, they aren't listening to the upper frequencies, they are listening for better accuracy and spaciousness in the the frequencies they can hear. Have you heard speakers that can go beyond 20kHz and compared them to speakers that can't?

Here's a little test you can do fairly easily at any high end audio dealership that carries Sennheiser HD800 headphones. Go listen to a pair of garden variety headphones that have a frequency response of 20Hz to 20kHz (which is the normal hearing range accepted by the majority of audiologists) and then listen to a pair of headphones that have a frequency response of 14 to 44.1kHz (Sennheiser HD800) and then tell me you can't hear a difference. Odds are you will hear a difference and it's not a placebo effect.

Your test setup is wrong. You basically compare different headphones which will obviously sound different. On top of that one model is higher quality, so will likely (hopefully!) sound better than the other.

The correct test setup is to e.g. always use the highest quality headphones, and test "ultra-high-def" audio vs "16/44 downsampled" audio coming directly from the "ultra-high-def" version, so that the only difference comes from the downsampling. This way you are able to isolate whether the downsampling is transparent to the human hearing or not.

Or, if your doubt is about being able to hear ultrasonic frequencies in the first place, test with a sample with ultrasonics and the same sample with the ultrasonics filtered out (hoping that the ultrasonics don't manifest audibly as artifacts in the lower frequency ranges...).

Of course, that's basically what the above mentioned xiph.org test was about...
 
I should probably clarify something... when I said the average 24/96 release "sounds better" than the average 16/44.1 release, that's not solely because of marketing but because, on balance, the average 24/96 release will likely be produced by better recording/mixing/mastering techniques. These techniques can be, and often are, applied to producing a 16/44.1 release that'll sound identical. Where the marketing comes in is in convincing people to believe blanket statements like "24/96 sounds better [than 16/44.1]."

I would agree with that statement. Typically, but not always, a lot the 24/96 releases have been straight conversions from DSD and they typically removed most, if not all, of the audio compression. What the record labels started doing years ago is they originally used SACDs as a form of archival for old analog tapes because magnetic tape degrades over time and they wanted something to preserve the integrity of the analog tape with the technology that was available at the time, so these are to sound as close to the original 2 track masters, so these 24/96 or better are supposed to sound just like the 2 track masters (or as close as they can get). At least that's the premise behind these high res 24/96+, DSD, and DVD-A, or now they have DTS HD Masters in 24/96 and True HD, etc.
 
60dB Normal conversation
85dB Beginning of hearing damage range, earplugs should be worn
100dB Normal average car or house stereo at maximum volume

90dB is the beginning of hearing damage according to OSHA, at least and it's rated at 8 hours time before any significant damage starts. It gets shorter pretty quick for every 3dB above that. But my point is you left out the most important range. 120dB is considered instant damage. 110dB is like about 15 minutes per day exposure maximum. But Dolby Theater levels are supposed to be set to reach their absolute maximum at about 110dB for the loudest explosions, etc. but that should not go on in a movie for very long periods at the absolute maximum (or hardly ever). And a home stereo that can only max out at 100dB is a total POS (oh Bass at those levels is typically NOT harmful which is why all OSHA type measurements are "A-Weighted" which means scaled to human hearing sensitivity ranges. "C-Weighted" is evenly measured sound for all frequencies and is relatively meaningless for hearing measurements which is why nearly all sound meters have a built-in A-weighting scale option. BTW, even a Bose Wave Radio could probably put out 100dB at certain frequencies.

It takes a doubling of power for every 3dB gain. My system can do about 115dB maximum C-weighted and that's with a 180W amp for my ribbons and a 350W amp (both in to 4 ohms) for my 10" woofers on my Carver Ribbon system. Most music I play is in the low to mid 80s average A-weighted for the loudest parts of the music with peaks possibly at 90+ for short periods. Unless one listens all day long, this should be relatively safe. Most people with hearing damage listen to headphones that are turned up to overcome other noises (e.g. on airplanes) and this leads to average volumes that are way too high.

16 bit recordings offer dynamic range of 96 dB, that means for example a range of 10 db to 106 dB. But even that base level of 10 dB is far-fetched for most listening rooms because they usually have some kind of equipment like lighting, air ducts or amplifiers that produce hum of up to 20 dB.

Actually, the typical average house noise level is typically in the 40s to 50s for a "quiet" room. Thus, noise-masking is a significant issue. No one listens to music in an anechoic chamber except scientists for testing so the idea that 96dB dynamic range isn't enough is ludicrous given the masking effects. To hear all 96dB in an average room, you would probably need to be well over 105dB, possibly 110dB or higher for many frequencies. This is very bad for your hearing. Furthermore, probably 99.5% of all recordings don't have ANYWHERE NEAR 96dB dynamic range to begin with, particularly with the trend for louder and louder mastering. Even the BEST classical recordings I've seen may very rarely achieve recorded dynamic range levels around 98-100dB at most and you can actually reproduce that on CD with noise-shaping fairly well. Most humans hardly notice volume differences in the 1-3dB range without constant comparisons anyway. I don't know what's supposed to be "HD" about going from dead silence to Space Shuttle loudness volume changes anyway. Only movies typically even come close to using that much dynamic range and I don't want to hear ANYTHING that might damage my hearing (i.e. 120dB is instant damage) anyway.

Humans cannot hear much if anything over 20kHz even when they're young and most people over 40 would be lucky to hear 15-16kHz reliably at volume levels that aren't obscene. So many audiophiles rave about (analog) LPs on the best record players sounding better than digital music, particularly CDs, but they have a maximum dynamic range perhaps in the mid-70s at best and that's with 12" singles. Typical 12" 33RPM albums around 60-65dB dynamic range, not even CLOSE to CD capable levels and hardly anyone notices that they're missing since MOST recordings (particularly rock ones) don't have anywhere near 65dB dynamic range. In fact, most rock albums average dynamic range between "semi-quiet" sections and "loud" sections are probably less than 10-20dB these days. It's terrible.

But then is "Music" rated on its ability to go soft to loud quickly or is it based on the actual musical pitch/note changing content? Personally, the difference to me between good speakers and bad speakers has more to do with EVEN (e.g. +/- 1dB) and FAST (ability to handle quick transients) frequency response along with good room matching. What is called "detail" comes from being able to hear the frequencies associated with smaller frequency changes at the start of the envelope relative to louder "noises" like the ringing of a long note or overly emphasized bass, etc. Most cheap systems simply cannot put out quality bass or have things like "metallic" sounding highs (over-emphasis of certain high frequencies given an unnatural "crispy metallic" sound to instruments that didn't have them in the recording. NONE of this has a lot to do with really large scale volume changes (dynamic range) or higher frequency response than human hearing allows (>20kHz).

THUS, I'll say what I've said before on this subject one more time. MOST people would do 10x better to get better speakers with enough power to drive them cleanly and/or work on room response problems than to worry about high-end DACS, "HD Music" or any other crap that doesn't matter a bean in hell compared to the least common denominator which is their speakers, most of which are nowhere NEAR the quality that minute tiny tiny differences would even be revealed. In other words, who CARES about 0.05dB difference in a high-end DAC compared to a "low" one (which are the high ones from 10 years ago) when their speakers are +/- 3dB (i.e. up to a 6dB difference between what the frequency is supposed to be and what is coming out of the speaker, let alone the room response which could potentially make that 10-20dB or more for some frequencies, particularly bass where uneven room response and standing wave issues are a BIG problem for typical rooms which were never designed for audio listening.

In short, "HD Audio" is a marketing gimmick that will be used to separate you from your money when you should be spending it on better speakers and room treatments.

I would like to say one final word about iPhones and the like. The REASON many of you get such god-awful audio from your iPhone has to do with two things. One is the cheapy-little op-amp (amplifier on a chip, basically) used in such tiny flat phones. In short, they put out very little clean power. The second is the other end of it which is the headphones used with the iPhone. Whatever headphone you use is going to impact the music both through its frequency response (those tiny ear buds will NEVER sound very good even with all the clean power in the world) and its ability to be driven with a cheap op-amp (low power). The best thing you can do short of buying a headphone amplifier is to get headphones that both have good frequency response AND are very efficient (e.g. the higher the number the better). In other words, if a headphone doesn't need much power to put out a lot of sound then it can be driven cleanly (high quality) with very little power. THAT is what you want with a device that puts out very little power. The same is true with home systems. Something like Klipsch horn-loaded speakers rated at maybe 100dB per 1 watt at one meter will perform as well with 12.5 Watts of power (e.g. maybe a tube amp) as a 91dB rated speaker will perform with 100W of power. This is extremely important in car speaker systems where no dash head unit on earth can put out more than about 14.5 Watts RMS without an outboard amplifier that uses a DC-to-DC power transformer type amplifier. Reducing noise in the environment can make it sound louder at lower volumes as well and save your hearing in the process.

44.1/16 is CD quality and 96(or higher)/24 is HD. All music is done 24bit these days so downsampling is not 2014.

"HD" is a bad label for audio given there is typically ZERO audible difference between 16/44 and 24/96. It is a gimmick on the playback side, but very useful for headroom on the recording side. Sadly, gullible people that believe everything they read or hear from any "news" source will often buy into any snake-oil sale they read about.
 
on balance, the average 24/96 release will likely be produced by better recording/mixing/mastering techniques.

Not to mention that even if the original 16/44.1 master is fantastic, there's an incentive to make a 96/24 version sound different just so that people hear a difference when they buy it. If they released 24/96 versions and nobody could hear a difference, that wouldn't really sell, would it?

Typically, but not always, a lot the 24/96 releases have been straight conversions from DSD and they typically removed most, if not all, of the audio compression.

First, you don't really remove audio compression. If a release has too much compression, you go to a mix/master before compression was applied and work from there. Second, you're really hung up on DSD. Some recordings probably were digitized to that format at some point. But no reason to assume that most did that, or that most of the 24/96 releases now are coming from that format as opposed to going back to analog tapes. If a DSD copy was made, depending how long ago that happened, it's possible that a fresh digitization now might sound better simply because A to D converters have improved so much over the years. And it's not like 24 bit is that recent a development, especially at the level of mastering houses.
 
For instance, a young person will hear a signal at 20kHz. To capture that signal, you need to sample at at least 40kHz. Then, that young person will be able to hear something, but you will have lost a lot of characteristics of the signal - for instance, you will not be able to know if the original signal was a sawtooth, a square or a sinusoid. So, significant information will have been lost.

This is total garbage, absolutely None of the signal will be lost! It's a simple concept in layman's terms; one needs at least 2 times the sample rate to encode/decode an analogue signal into digital (and back again) but, and it's a BIG BUT, having 3, 10 or 10,000,000 times the sample rate does NOT make the reconstructed signal more accurate!

That's why CD recordings sounded metallic at first. The solution, which is applied on all CDs, was to cut the frequency around 16kHz to avoid the destruction of the characteristics of the signal around Shannon frequencies.

There is no such thing as "Shannon frequencies", you just completely made that up, just as you have completely made up the "fact" that 16k is cut on all CDs?!

That's why 96kHz is interesting, because it keeps quality in the upper part of the spectrum.

96kHz has resolution in the 20.05kHz - 48kHz audio band which CD does not. Great for dogs, bats and dolphins but not so useful for human beings who cannot hear above 20kHz (which is why the CD specs were chosen in the first place)!

Moreover, 24-96 is not only about 96kHz, it's also 24 bit. And there, you gain a lot.

No, you gain absolutely nothing whatsoever!!

The result is that when you go at the bottom of your intensity, you have a very very low resolution in your sample, while the human ear (or eye) still have a good resolution.

No, on playback you have perfect resolution, many times greater than human hearing!

It's the same problem with audio: CD killed the dynamic range (hence the loudness war), because it's not that good when you have a lot of dynamic during the low volume ports.

You're joking right? The highest dynamic range commercial recordings on the market have a dynamic range of about 60dB which is roughly 1,000 times less than the dynamic range available on a CD! So how on earth has CD killed dynamic range?

If you are going to correct someone on matters of digital audio theory then you really should learn at least the basics of digital audio theory!!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.