Grateful Dead

dead.net the official site

You COULD do it double blind. But, you HAVE to make sure you start with the same files. Take your 24/96 or whatever file, have it professionally converted to 16-bit. Don't just get separate files to start with. Even very slight differences in volume will make a difference (louder is almost always reported as better in testing). Then get someone to help with the a/b testing. Ideally, you should NOT be able to see the other individual, and it would better if he didn't even talk if he is going to know which is which; to keep it double blind he nor you should know which is 24 and which is 16 until after all testing. Try to take no less than 100 listens. Use equipment to make sure volume level is truly identical, not the volume setting of the playback equipment, but the volume of the playback itself. And, of course, he shouldn't just switch back from one to the other. Use a random number generator to determine the order of which files to playback in what order. Ideally, you should check both files with visual analysis software so that you can really see if the conversion to 16 bit was done well. The sine wave results should be virtually indistinguishable in amplitude when overlayed. The only real visual dupifference you should be able to see would be possible content in frequency ranges above 22khz in the hi res file that wouldn't exist in the 16/44.1 file. If this is not the case you're not comparing apples to apples and the test won't mean anything. P.S professionals use 24 bit recording for reasons that have nothing to do audio quality of the listening experience of those files. It has to do with the playing room it gives for subsequent digital manipulation. I think one of the articles I linked to talks about this.

"Do frequencies (including noise purposely placed) outside the audible range change our reaction to music?" People keep missing the point that even if it's just feelings or some unquantifiable non-auditory affect, if it made ANY difference - even one you couldn't put your finger on, that would SHOW UP on the results of the double blind test. Scientifically (as far I'm concerned) they've proven that there is nothing, not even something inaudible or even supernatural, that is making a difference, or the results would be different. As far as noise, it is the EXACT same issue. Scientifically, any added noise from dithering should be inaudible unless you have a noise floor about zero, which never happens. And again, exactly as before, if it made ANY detectable difference it would skew the results of the double-blind studies - which clearly it did not; that speaks for itself. Yes, we can agree to disagree. I prefer engineering that errs on the side of not intentionally trying to take advantage of the less technically informed for a buck. And I also disagree with the characterization that this is going a "step beyond" and what it implies. You are repeating things like "demonstrably greater noise" while ignoring that noise you can't hear isn't really noise. If snake oil makes someone feel a little better it NEVER changes the original intent behind the making of that snake oil, and never will. Unfortunately, this is precisely the kind of disagreement, discussion and outcome that the folks who ARE aware of the science behind digital audio technology and are trying to capitalize on it are counting on. They have to. But, like I said, it's not my money and there are much more important things to worry about. For what it is worth, if you do spend your extra money on "hi res" files and equipment and storage space and download times, etc., I do hope you enjoy them. Especially if it's Jerry! EDIT - And, doesn't it bother you AT ALL that in the marketing on places like HDTracks and other Hi-Res sites, they are intentionally misleading. While you, after reading some of the science, have realized that the "smoothness" issue, and the "stair step" issue are bogus, even if you don't seem to see the same with the "noise" issue, it is simply fact, not opinion that there is no "stair-step" issue, but if you go look, that is precisely the kind of material using graphs, etc., that they use in their marketing. In other words, they are using something that, regardless of how you feel about so called hi-res audio files, is entirely scientifically bogus - you can see on audio sound analyzers that the music/sound waves that are produced are as smooth and identical to the originals, but these sites display graphs showing stair steps of rectangular discreet "samples" and showing more samples making a sound wave smoother, using words like giving the music a more "natural" less digital "feel" (demonstrably false). Doesn't this kind of marketing TELL you anything about what is going on??? And, in light of that, when you refer to how we don't understand everything about how humans/the brain respond to this or that, are you implying that they might be right BY ACCIDENT, that even though they're clearly intentionally lying to their buyers about much, that COINCIDENTALLY they might be selling a higher quality product?? Not buying it. I'm with the Society of Audio Engineers on this one. EDIT 2 - And, while you're talking about the (as far as I'm concerned illusory) intangible but maybe real and subtle differences, doesn't it bother you to read about the legions of people out there are who buy these hi-res files and then post about how they're SO MUCH better, you can just hear how much deeper the sound is, the cymbals are so much crisper (that would be in the AUDIBLE frequency range), the sound is so much smoother, you HAVE TO experience it for yourself! You now know how much of that is simply not factually possible (other than in the mind due to expectations), but you can still stand behind this? Sorry, I can't, I just can't. EDIT 3 - I thought of something else, too. While you appear willing to overlook the most glaring falsehoods being perpetrated on the off-chance that the "hi res" MIGHT offer some virtually intangible benefits, you appear completely ready to ignore things like the quote from the first link I sent which reads "Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space." He goes on to explain why, and I believe at least one of the other articles mentions it also - if not, I know you can find ones that do. The reasons for the slight inferiority, which have to do with the potential affects of inaudible frequencies attempted to be reproduced by sound equipment whereby the actually AUDIBLE frequencies are interfered with (something that wouldn't happen from listening to live music, like a guitar, but DOES happen due to the inherent inadequacies of speakers and headphones of whatever quality) - you seem to be perfectly willing to just ignore any negative (and in this case demonstrable) affects of using playback files that store frequencies that are not just a little but astronomically above human hearing level. Again, to quote "Neither audio transducers nor power amplifiers are free of distortion, and distortion tends to increase rapidly at the lowest and highest frequencies. If the same transducer reproduces ultrasonics along with audible content, any nonlinearity will shift some of the ultrasonic content down into the audible range as an uncontrolled spray of intermodulation distortion products covering the entire audible spectrum. Nonlinearity in a power amplifier will produce the same effect. The effect is very slight, but listening tests have confirmed that both effects can be audible." Also being ignored are the fact that virtually no microphones (certainly none in use commercially) are even capable of picking up these frequencies to begin with, so ANY frequencies in that range ARE noise introduced as part of the digital file manipulation phases, which 16/44.1 files would simply lop off, but are still contained in a 96 or 192khz file? The list goes on and on and on. And, for me, I just will never get over the INTENTIONALITY of the original deception for the sake of greed, and how it has now spilled over into otherwise well-intentioned, but misguided supporters. EDIT 4 - the argument also reminds me of psychic pay per minute phone lines. It's like hearing an argument from people who spend a few hundred dollars a month on these psychic hotlines explaining that we don't know all the capabilities of the human mind. No, we don't. Does that make it one scintilla more likely that the "psychics" on the other end of the $2.00 per minute phone call are anything but frauds? Nope. And the fact that people can and do legitimately bring up our lack of complete understanding of the capabilities of the human mind muddies the waters and gives some reasonable semblance of credence to these frauds drives me similarly batshit.

I wish Owsley Stanley were still alive to debate this. He said to me that digital audio (all of it) is "a bad joke" and I tend to agree as far as in comparison to analog. The day I plugged in my (24 bit/48K) multitrack in place of my old Otari MX-70 (1-inch 16-track analog magnetic tape) was the day my studio began sounding less warm and snuggly. Of course, there are a million reasons why this is true, none of which are likely to be cured by "better" digital audio technology. I'm sure someone has tried to invent a tape emulation algorithm and I don't see that gaining any traction. That aside, virtually all professional studios use 24 bit recording, even knowing the product will end up as 16 bit. I have the choice but have never used 16 bit multitrack. Maybe I'll try that. It won't be double blind, but it could be revealing if I use a MIDI source, drum machine and/or other "pre-recorded" sources so there will not be any performance cues. I could even transfer a song from an old LP and hear it both ways. I'll report back with results. I am not down with false marketing of 24-bit audio. The science should not be tampered with to make a buck. PONO makers and the like should just explain what they have done and see what the market will bear. I don't plan to buy one, but I could change my mind.

Well, we are going to have to agree to disagree on the "snake oil" issue. If 24 bit has demonstrably lower noise, it's not snake oil, even if subjects in a double blind test can't "hear" it. The effect of audio on humans can only be measured to a certain degree. The rest -- call it "feelings" if you must -- is in the ear and brain of the beholder. Do frequencies (including noise purposely placed) outside the audible range change our reaction to music? I don't know, and no test can prove there is no effect. I'm sure that Warlocks box "sounds" great on paper. It apparently met whatever specs were used to produce it. I prefer engineering that errs on the side of quality. I want digital audio to go a step beyond the old 16/44.1 design, and now it is going there. And it is unlikely to go further in that direction, if that is any consolation to anyone thinking this will never end.

I'm not sure what to say. While the Warlocks sound has issues, are they mastering issues? Mixing issues? One thing we know is that it is not a 16/44.1 vs 24/96 issue. We know that that is not the problem. In the tests (talked about in one of the links) where they did a double blind test where they inserted a 16.44.1 loop, they didn't even bother dithering. Dithering is NOT the issue. It moves quantisation error/noise into the mostly inaudible regions of the frequency range. Part of the problem is that by asking, "So why not go 24/96 from here on out?", it's like hearing someone listen to a snake-oil pitch - snake-oil that won't do any harm, but costs major bucks and for which an entire industry is ready to sell you lots more of it and lots of extremely expensive accessories to go with it. You're asking, what's the harm? And, part of the ability for them to do that is predicated on people having the same preconceptions and and misunderstandings about digital audio that were in your original post - believing in things like "granularity", a "smoother" sound because you have more discrete samples (probably the most frequently heard misunderstanding), greater "depth" to the recording because you have more bit-depth (COMPLETELY off), the idea it is closer to analog, the idea of that what you get is a "stair-step" sound wave and having more samples makes for more steps, and smoother sound wave, etc. Even many audio professionals who don't deal directly with the technical aspects of how the files work buy into this demonstrably nonsensical understanding of what is going on - and this is CRITICAL for the people who want to take your money unnecessarily (many of them probably belive it too). As long as there are folks bringing up ambiguity (similar to "the snake oil coulnd't HURT), as long folks repeat nonsense like "well, the extra frequency range in 96khz recordings may not be in the audible range, but the harmonics created by those frequencies probably affect the way the music FEELS". If that were true IN ANY WAY the double blind tests would fail - people would be able to pick out the difference. In any case, the train's probably already left the station. The idea of "high resolution" is probably already too firmly entrenched, and I expect many people will buy into it. I guess there are worse things, but the snake-oil thing drives me batshit. P.S. Edit - I recently found out that, contrary to what I implied in an earlier post, unlike in the early years of digital audio, modern DAC's (digital to audio converters), even the most inexpensive ones are virtually perfect. There is no longer really any such thing as a "better" or "higher quality" DAC. They all virtually perfectly reproduce an analog sound wave that is identical to the original.

Thank you for the links. The common caveat seems to be "if properly dithered". I am sure I have heard many digital recordings that lacked proper dithering (or other treatment) because they sounded obviously harsh. So we can't necessarily assume we are always talking about properly dithered recordings. Some sound terrible and it is clearly a digital issue as you don't hear analog recordings sounding this way (although they can obviously have their own problems). Also, John Siau says in his article, "Long word lengths do not improve the amplitude "resolution" of digital systems, they only improve the noise performance. But, noise can mask low-level musical details, so please do not underestimate the importance of a low-noise audio system." So if 16/44.1 is "good enough", it is just barely "good enough" and sometimes probably isn't. So why not go 24/96 from here on out? We will never need to go higher than that. Relating this to the Grateful Dead, the release "Formerly the Warlocks" sounds terrible to me, and I am nearly certain this is a digital issue. I have never heard an analog recording that lacked this much "depth" and sounded this harsh. By "depth" I am not talking about dynamic range nor frequency range. There is something missing throughout the signal. I can't measure my dissatisfaction with this recording -- all I have for instruments are my ears. But I am sure some other listeners hear what I hear in this recording. I'm not blaming it on 16/44.1. I am blaming it on poor digital engineering of some kind.

Hi One Man, Respectfully (seriously), there are too many factual errors and misunderstandings about digital audio technology in your post to reply without writing another tome. I will instead point you to some links that explain some of it. http://xiph.org/~xiphmont/demo/neil-young.html http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf http://benchmarkmedia.com/blogs/news/15121729-audio-myth-24-bit-audio-h… http://productionadvice.co.uk/no-stair-steps-in-digital-audio/ http://en.wikipedia.org/wiki/Audio_bit_depth http://www.head-fi.org/t/415361/24bit-vs-16bit-the-myth-exploded In particular your understanding of the relationship between how digital audio technology works, and what you are referring to as "granularity" is simply incorrect, but conforms to "common sense" in the sense of how most people believe digital audio works. If you're interested in the topic I would suggest reading those links in their entirety (I believe they have references to many other locations for further information as well). Taken together, I think these go a long ways to a good explanation of some things that are not intuitively obvious, things like, from that last link: "So, 24bit does add more 'resolution' compared to 16bit but this added resolution doesn't mean higher quality, it just means we can encode a larger dynamic range. This is the misunderstanding made by many. There are no extra magical properties, nothing which the science does not understand or cannot measure. The only difference between 16bit and 24bit is 48dB of dynamic range (8bits x 6dB = 48dB) and nothing else. This is not a question for interpretation or opinion, it is the provable, undisputed logical mathematics which underpins the very existence of digital audio." You will also see, as explained in the article on bit-depth, that each "sample" as represented by a 16-bit (or 24-bit or 2-bit) binary number ONLY encodes the amplitude (volume) of the signal. Frequency is controlled ENTIRELY by sampling rate. When you have a particular "volume" measurement played back 1000 times a second, you get a sound frequency of 1000hz at the volume specified. It's easier if you think of each "sample" as encoding a virtually instantaneous "tick" sound where the number of bits controls only the volume of the tick. How fast the ticks are made produces a tone. While it is true that 16-bit encodes 65,536 different possible numbers, and 24-bit encodes 16,777,216 different numbers, the granularity you refer to I don't think is granularity as you believed it to mean. The difference between 65,536 and 16,777,216 is ONLY the difference of how many VOLUME levels can be encoded. While there is some controversy over whether frequencies over human hearing can affect what we hear (there shouldn't be), there is no controversy that no one can detect the difference in volumes from one level to the very next at the granularity level of either 16-bit or 24-bit, so their "smoothness" is identical to human hearing. For instance, LP's are the equivalent of about 11-bit recordings (they have to compress the dynamic levels so the lowest volume to loudest fits within this range due to the limitation in groove/needle technology). Assuming with the most modern technology, the newest LP's can be equivalent to 12-bit (and I have no reason to think this, but let's assume they've improved), that means LP's as you knew them had a "granularity" of about 2,048 volume levels with newer ones MAYBE having up to 4,096. I don't think the "granularity" of 65,536 is a problem and certainly NOT distinguishable from 16,777,216.

It's partly my fault this board has digressed into a long discussion about digital audio. Sorry about that. But I must say (at least) one more thing. Saying that bit depth only affects dynamic range is way off the mark. Bit depth is the number of values available for each digital sample of the waveform. So the granularity (resolution) of the sound is dependent on bit depth. Sure, it ends up as a sound wave by the time it reaches your ears, but the shape of the wave is modified by digitizing it. Take the logic to the extreme. If you could have a 2 bit recording, each sample could only be assigned to one of 4 values. Imagine how raw that would sound. The number of available values is the number 2 raised to the power of the bit depth. So, an 8 bit recording has a "granularity" of 256 available values per sample. A 16 bit recording has 65,536 available values per sample and at that point is getting quite a bit more resolved. A 24 bit recording has 16,777,216 available values per sample and is thus 256 times more resolved than 16 bit. I'm not saying everyone can hear the difference between 16 and 24 bit. But people can certainly hear 8 bit vs 16 bit. So some people - maybe not enough to statistically skew the even odds stats - probably can hear 16 vs 24. I can tell you from my experience that my analog studio tape machine sounds noticeably better than my high-end 24 bit digital recorder with excellent AD and DA converters. And anything that approaches analog by providing higher resolution is a move in the right direction, even if Neil Young is a grumpy old man having a mid-life crisis about 2 decades late.

I suspect that this is based in some degree on the fact that Neil can be a rather abrasive personality and people will take shots at him when they can. There is also probably a bit of a reflexive distaste for the pricing and kickstarter campaign that came with the pono rollout. As we see here often, any time a product is priced above what a kind veggie burrito cost in the lots at SPAC 1985, people bitch and moan.

I realized after the fact that every time I referred to uncompressed CD quality files I should have referred instead to lossless CD quality files, as some might not get it that FLACs and SHNs are digitally identical to the uncompressed wav files at playback. I agree about the need for greater availability of lossless downloads. It drives me batshit that iTunes doesn't offer FLAC, and even most sites that have the largest selection of classical music still only offer mp3's. You would think that classical music places would be the first places to realize the demand for lossless download purchases, but I guess not. I create my own high quality mp3's so that I can fit my entire music library on several 160GB portable devices, but I like to have the originals on my home playback library.