PDA

View Full Version : Touch and 24/96 clock



firedog
2009-09-09, 05:12
Hi-

My main interest in the Touch is 24/96 ability. Is this a hardware clock based function or a synthetic clock based ability?

Thanks

Phil Leigh
2009-09-09, 07:55
Hi-

My main interest in the Touch is 24/96 ability. Is this a hardware clock based function or a synthetic clock based ability?

Thanks
Not sure I understand what you mean... it plays 24/96 natively (i.e. perfectly) and has a suitable clock circuit for that purpose, derived from a seperate crystal to 44.1 playback...

JohnSwenson
2009-09-10, 00:09
There are two crystals, one which is used for 44.1 and 88.2 and the other used for 48 and 96. The master clocks are divided down by integer amounts for whatever frequencies are needed for the sample rate in question (word clocks, bit clocks etc).

John S.

jimzak
2009-09-16, 17:15
What are the implications of the 24/96 clock?

All of my music is in .flac format.

What else is needed to gain the benefits of this clock?

A high-end amplifier?

seanadams
2009-09-16, 17:55
What are the implications of the 24/96 clock?

All of my music is in .flac format.

What else is needed to gain the benefits of this clock?

A high-end amplifier?

At the risk of stating the obvious: you need 24/96 music.

pfarrell
2009-09-16, 18:11
seanadams wrote:
> At the risk of stating the obvious: you need 24/96 music.

Hi Retired Guy.

You mean that the device doesn't synthesize back the missing values for
frequency and resolution?

--
Pat Farrell
http://www.pfarrell.com/

firedog
2009-09-17, 01:26
It means that you can play 24/96 hirez files in their native format.

Some people upsample 16/44.1 (CD)files to 24/96 and store them that way b/c they think it improves the sound in playback; you can also download tracks produced in 24/96 from sites like HDTRACKS.COM

Many think hirez files sound better.

The other SB units (excepting Transporter) play 24/96, but down sampled to 24/48 by SqueezeCenter before conversion to analogue.

jimzak
2009-09-17, 01:53
How does one rip music to 24/96?

Teus de Jong
2009-09-17, 02:51
How does one rip music to 24/96?

In general one doesn't, although there are some music video's with 24/96 audio tracks. The same may be true for DVD-A (I don't have any of these).

But a lot of people download 24/96 files from e.g. linn records, gimmell, hdtracks, etc.

Teus

firedog
2009-09-17, 06:07
How does one rip music to 24/96?

Some people do a standard 16/44.1 rip of a CD and then upconvert it. You need a ripping program that has a "resampler" function. Examples: dbpoweramp - windows or Wave Editor for MAC

Some artists (see the newer Neil Young discs) put their work out in high res formats, and with a bit of work, you can extract high res files from them.

Otherwise, they can be donwloaded from the sites mentioned above.

jimzak
2009-09-17, 08:31
Ok. Thanks. I understand.

Not much music available at this clock rate and mostly classical, DVD, etc.

We'll see what the future holds.

Pneumonic
2009-09-17, 10:53
Still, others do needledrops (digitize) of their vinyl albums and save them as higher resolution files for play back on systems capable of reproducing them. This is why I bought my Transporter in the first place.

Phil Leigh
2009-09-17, 11:05
Still, others do needledrops (digitize) of their vinyl albums and save them as higher resolution files for play back on systems capable of reproducing them. This is why I bought my Transporter in the first place.

also a few DVD-A's and SACD rips @88.2/24 from modded players... :-)

mortslim
2009-09-25, 14:29
The other SB units (excepting Transporter) play 24/96, but down sampled to 24/48 by SqueezeCenter before conversion to analogue.

The new spec of 24/96 on the new Touch is not something to have Gear Lust about. For all practical purposes most consumers won't miss it.

By definition, CD's are recorded at 16 bit/44.1 sample rate. So if you ripped your own CD's, 16/44.1 is the best quality you can ever get. Yes you can "upsample" to the higher spec with the appropriate software (e.g. Cakewalk's Sonar) however the resulting quality will be the same as the original source material. You'll get a good copy, but that's all it is, a good copy of the original.

Unless you are a teenager with perfect hearing, you won't be able to tell the difference between 24/48 and 24/96 anyway. You are limited by the ability of the human ear to hear. After reaching the age of 20 years, your hearing (sorry to say) gradually decreases every year in terms of the frequencies it can hear.

DVD's (not CD's) can be recorded at 24/48 or 24/96. So if you ripped music from a DVD, you will need software (e.g. Sonar) to check what the sample rate was on the DVD.

If you ripped from a vinyl record, yes you can rip to 24/96 if you have the appropriate software and hardware (analog to digital converters). The reason is that you are not really "ripping", you are doing straight recording in real time. But what a lot of people forget here is that vinyl degrades over time as it involves a needle scrapping the vinyl, resulting in degradation of the sound. So even with the best recording equipment, it is doubtful that recording at 24/96 will be a benefit over say 24/48.

Future music: even though DVD's have given us the promise of higher quality sound for many years now, music is still published for the most part at CD quality. Why? Probably a realization by the music industry that it would increase costs to record to DVD and the industry understands that most people have CD players in their car, not DVD players, so there would be an extemely limited market for DVD quality music.

It was a revolutionary jump from vinyl to CD in the marketplace. Going from CD to DVD for sound was evolutionary not revolutionary and going from DVD to Blueray DVD was even less evolutionary for sound for most of the market. The industry looks at CD quality as "good enough" for the vast majority of the market. The industry's market anaylsis is that it isn't worth the effort to publish music to DVD quality.

There is some DVD music available but it is in the very small minority.

The future of music is not DVD or Blueray DVD. It is online. And online music is in a compressed format, e.g. mp3. So if anything, the future of music, at least for the short-term future of the next few years, is lower quality, not higher quality. E.g. Rhapsody, which streams at a higher sample rate than most other services, still is not even streaming at true CD quality.

So enjoy your Squeezebox Classic or Boom or Duet and don't worry about the new "theoretical" specs.

mortslim
2009-09-25, 17:51
Here are somoe interesting observations from Cakewalk (professional audio recording software company) on the issue of bit depth and sample rate:


"Its also worth noting that not all audio projects benefit equally from the use of higher bit-depth/sampling rates. Most professional sound engineers will tell you that 24-bit, higher-sampling-rate audio is most beneficial when working with music that focuses on natural acoustic instruments and/or vocals recorded with very high-quality microphones. The benefits of 24-bit/high-sampling-rate recording are less audible when working with highly synthetic or highly compressed program material, often found in contemporary popular music. You must make your own judgment about whether the benefits of higher-capacity audio justify the extra demands it places on your computer.

Note: Red Book Audio CDs are still 16-bit (44.1 sampling rate), so if your goal is to burn a CD, you will have to mix down to 16-bit."

http://www.cakewalk.com/Support/Docs/24BitTips.asp

Mnyb
2009-09-25, 20:12
Well I would order a Touch asap, I'm ripping my 100 DVDA's right now.
And have an DAC/processor that can handle 24/96 .

I also got a spool of cat5e wire lying around, if my wifi is not up to 24/96 flac to one player and transcoded 24/48 to my other wifi connected player.

mortslim
2009-09-25, 22:12
I have six Squeezeboxes (3 Classics and 3 Booms), so I am a big a fan of Logitech’s products. My only point is that most owners of a Squeezebox shouldn’t feel left out of the party if they don’t own a Touch just because of the new specs of 24/96 DAC’s because in the real world they wouldn’t hear any difference.

Of course if you are in the market for a Squeezebox anyway (e.g. don’t have any Squeezeboxes yet or want to expand to additional rooms), the Touch does look attractive.

On the issue of DVD-A source material, even here one has to pause before gear lust overtakes reality.

Here are some excerpts from a manufacturer of audio equipment on the issues of what you can actually hear from a DVD player’s digital outs due to copyright issues:

“many DVD transports currently on the market fail to transmit the exact digital audio data bit-for-bit as it appears on the DVD. Of the DVD transports Benchmark has tested (8 as of this writing), all of them modified high-resolution audio data before sending it to the digital output”

“Transports modify the high-resolution data in order to comply with digital copyright obligations to the publishers and distributors of music and video discs. The music and film industries are trying to prevent bit-for-bit copying of the recordings. Consequently, manufacturers of DVD transports have been forced to reduce the resolution of digital outputs to prevent digital duplication.”

“However, the sonic consequences of these modifications can be devastating to audio quality. We are able to measure sample rate, bit depth, and conversion quality with our Audio Precision test equipment. In almost all cases, the data modification includes poor quality sample-rate conversion (SRC) and word-length truncation. These processes defeat the benefits of high-sample-rate and/or high-bit-depth source material, and severely degrade sonic quality. Although the SRC and word-length reduction is being applied to reduce the quality of the digital output to CD resolution, it actually degrades the sound quality far beyond that of properly mastered CD resolution. In fact, these transports will deliver higher performance from a CD than from a high-resolution DVD.”

“Unfortunately, the SRC processing in the transports tested was of poor quality. The distortion that results can be devastating to the quality of the playback system. Also, it is often re-sampled to 44.1 kHz, which inherently eliminates all the advantages of playing high-sample-rate audio discs.”

“Most DVD transports reduce 24-bit digital recordings to 16 bits by truncating the 8 'least-significant-bits' (LSB's). In other words, these 8-bits are simply removed without dithering. This type of truncation adds quantization errors that result in significant levels of distortion.”

http://www.benchmarkmedia.com/discuss/feedback/newsletter/2009/07/1/your-transport-giving-you-all-bits


Here are some excerpts from a manufacturer of audio equipment on the issues of bit depth and sample rate:

“96kHz sampling also causes problems with the noise performance of low frequency EQ stages (because the differences between adjacent samples are smaller), so a 96 kHz system typically requires a longer wordlength to achieve the same noise performance as a 48 kHz one.”

“Measurements which specify the dynamic range of the ADC or DAC in isolation should also be treated with caution, since these are often “data sheet” numbers supplied by the IC manufacturer which are rarely if ever achieved in practice.”

“The ADC and DAC parts that we use are both "nominal 24-bit" items, but this is essentially meaningless. If a manufacturer claims that they have a "24-bit converter" in their product, then the next question to ask is how you should measure the unit to confirm the 144 dB dynamic range that this implies. In practice no-one is achieving even 20-bit noise performance (=120 dB dynamic range) from a digital system of this kind at the present time. The DN9848 achieves >114 dB dynamic range or "19 bits" overall from input to output. Note that this is an unweighted figure (i.e. flat frequency response). Some manufacturers quote "A-weighted" figures which flatter the unit's performance significantly by applying a psycho-acoustic curve to the measurement. Measurements which specify the dynamic range of the ADC or DAC in isolation should also be treated with caution, since these are often “data sheet” numbers supplied by the IC manufacturer which are rarely if ever achieved in practice.”

“In the end, the one-sentence summary is "don't worry too much about the bits and sample rates - trust the same real-world performance measurements of noise and distortion that you would apply to analogue".

“And after that, there are always your ears...”

http://www.klarkteknik.com/faq-04.php

Finally, a recent study published in the journal of the Audio Engineering Society concluded that not a single listener could consistently distinguish between high-resolution audio and audio played at the CD standard.

http://www.aes.org/journal/online/comment/?ID=14195

Phil Leigh
2009-09-25, 23:36
However, ripping a DVD-A (using DVDAudioExplorer) avoids all the issues mentioned and produces a bit-perfect copy at whatever the native sampling rate/bit depth is on the disk.

The Benchmark paper is talking about what appears at the S/PDIF output of a normal DVD player. In the context of Squeezeboxes, this is unimportant.

There is a well known paper that describes a test where a 44/16 ADC/DAC is inserted into the replay chain and no-one could detect it.

The reason I am interested in ripping DVD-A's is because they often contain a superior sounding master mix to the equivalent CD. However, the same is often true of Japanese pressings of standard redbook CD's (especially the recent SHM-CD's).

Having done quite a lot of side-by-side comparisons of the same track at 24/96 vs 24/48 I agree there is no perceptible difference on the Touch.

However, something appears to me "better" about the Touch compared to the Classic when driving a DAC, and whatever it is is material-independent. The difference isn't huge but it is there.
I was running the SB3 and Touch side by side for about 3 months (into the same DAC) and quick A/B's were not categoric, but over longer periods the Touch won me over.

Mnyb
2009-09-26, 00:40
Yes to most of the above .

Imho some (few ) DVD-A sound better than any CD I heard, many don't ,some simply have a better master as phil pointed out .

However I have the material and now when it's possible i rip it to use all over the house (disc are so awkward the never get any playtime ).

The Touch would come in handy here. Native support for 96k means that transcoding will not be happening so often as it loads the server quite a bit.
My lean little server can do 24/192 to 24/48 for one player/stream but thats the limit, synk can be weird with 24/192 source ?
The sligthly improved digital output may or may not be heard in my system.
So Touch is cheaper than a server overhaul.

A squeezebox would not suffer from the mentioned problems neither would the DVD-A player I usually use.

It has a proprietary link to the pre/processor via it's own protocol over three paralleled coax cables(6ch 24/96 ).

Yes 6ch 24/96 is bliss, I wish there where a multichannel squeezebox.
DVDA really shine in multi channel.

Processor then handles speaker management drc and subwoofer filter other stuff and volume in the digital domain.

So my signal is going to be post processed in some way, better to keep it in it's native 24/96 format then. I've learned that all processing shall take place in much higher resolution than the finished "product", I think my processor uses 64bit floating point internally to be transparent even with 24/96 material.
The nontransparent part is of-course DA and the rest of analog parts

In the future I will get active speakers with digital crossover these will probably also accept 24 bit input, so I only starts to loose sound quality after xo and in the speakers .

Mnyb
2009-09-26, 00:56
PS you can see that flac is less effective with this material most likely because the last bits are random noise from mixing consoles microphones and such.
You can clearly spot old analog masters :)

firedog
2009-09-26, 01:05
Unfortunately for your post about "you can't hear the difference" -

I can - and so can lots of other people. The AES paper you note showing "no one" can tell the difference has some serious methodology problems. Just because it is supposedly a "scientific, double blind" test doesn't make it a good test.

BTW, there are high res files available for legal download from sites like HDTracks and Chesky that are in 96k and above. So obtaining hi-res isn't dependent on ripping from a CD. DVD-A also exists. Some artists, such as Neil Young, are releasing their music in hi-res formats.

I'm sorry you can't hear the difference. Maybe your stereo isn't good enough, or your ears aren't. I'm over 50 and the difference is often striking. It isn't just a matter of high frequencies, but many other aspects of how music sounds. Ever notice that some people exposed to hi-quality video playback say they "can't see the difference"? Or that some people say mp3 and CD's "sound the same"; well the same thing applies here. Listen to 24/96 for a while on a suitable system, and the difference will be striking when you switch back to 16/44.1. You obviously haven't done anything like this, or you wouldn't be posting on the topic purely on the basis of things you've read.

If you want to believe the high school physics about why "we can't hear" anything outside of what a 16/44.1 CD can reproduce, go ahead. But the truth is that psychoacoustics (what we hear) is way more complicated than that. Do a little research and you'll find academic and engineering papers that show otherwise.

Quote from Robert Hartley at Avguide.com:

"Bob Stuart (degree in psychoacoustics) has shown in a series of AES papers why 16-bit/44.1kHz is insufficient to encode all the information humans can hear. He bases this thesis on models of human hearing that are generally accepted in the psychoacoustic literature.

Secondly, although 44.1kHz sampling is perfect in theory for encoding a signal with a bandwidth of 20kHz, in practice 44.1kHz is too slow because of the requirements it puts on filter design. The anti-aliasing filter needs to have no attentuation at 20kHz and more than 100dB of attenuation at 22.05kHz. Such steep filters introduce time-domain distortions. See the AES papers of the early-to-mid 1990's by Mike Storey.

Third, there's been a suggestion that although we can't hear sine waves above 20kHz, we can detect the steepness of transient signals that implies a bandwidth greater than 20kHz. We use these steep transient in localizing sounds. Part of the HDCD process (the patent application makes for fascinating reading) encodes in the hidden information channel indicators that a signal's transient leading edge is less steep on the 44.1kHz/16-bit signal compared with the high-res original from which the HDCD-encoded compatible signal is derived. A conjugate process in the decoder restores the transient's orignal rise time."

If you're happy with standard CD's, that's great. But don't tell the rest of us what we can and can't hear. And at least admit the POSSIBILITY that maybe standard CD's aren't the best medium we have available, and hi-res adds something we can perceive.

Mnyb
2009-09-26, 01:32
Unfortunately for your post about "you can't hear the difference" -

I can - and so can lots of other people.

BTW, there are high res files available for legal download from sites like HDTracks and Chesky that are in 96k and above. So obtaining hi-res isn't dependent on ripping from a CD. DVD-A also exists. Some artists, such as Neil Young, are releasing their music in hi-res formats.

I'm sorry you can't hear the difference. Maybe your stereo isn't good enough, or your ears aren't. I'm over 50 and the difference is often striking. It isn't just a matter of high frequencies, but many other aspects of how music sounds.

If you want to believe the high school physics about why "we can't hear" anything outside of what a 16/44.1 CD can reproduce, go ahead. But the truth is that psychoacoustics (what we hear) is way more complicated than that. Do a little research and you'll find academic and engineering papers that show otherwise.

Quote from Robert Hartley at Avguide.com:

"Bob Stuart (degree in psychoacoustics) has shown in a series of AES papers why 16-bit/44.1kHz is insufficient to encode all the information humans can hear. He bases this thesis on models of human hearing that are generally accepted in the psychoacoustic literature.

Secondly, although 44.1kHz sampling is perfect in theory for encoding a signal with a bandwidth of 20kHz, in practice 44.1kHz is too slow because of the requirements it puts on filter design. The anti-aliasing filter needs to have no attentuation at 20kHz and more than 100dB of attenuation at 22.05kHz. Such steep filters introduce time-domain distortions. See the AES papers of the early-to-mid 1990's by Mike Storey.

Third, there's been a suggestion that although we can't hear sine waves above 20kHz, we can detect the steepness of transient signals that implies a bandwidth greater than 20kHz. We use these steep transient in localizing sounds. Part of the HDCD process (the patent application makes for fascinating reading) encodes in the hidden information channel indicators that a signal's transient leading edge is less steep on the 44.1kHz/16-bit signal compared with the high-res original from which the HDCD-encoded compatible signal is derived. A conjugate process in the decoder restores the transient's orignal rise time."

If you're happy with standard CD's, that's great. But don't tell the rest of us what we can and can't hear. And at least admit the possibility that maybe standard CD's aren't the best medium we have available, and hi-res adds something we can perceive.

I think mr Stuart's research shows that circa 20bit 55k ? would suffice, so choosing 24/96 gives the music producers some slush margin.
Consequently his own brand of Equipment ( Meridian ) don't do 192kHz other in the players, downconverting to 96k before sent to speakers or processors.
As 192kHz would only double the processing power needed and hog bandwidth with no obvious gain in sq.

Firedog: keep in mind that recording and playback are different animals.
The filter problems are worts in recording, in fact all things are accentuated in recording AD is *much* harder than DA .

Even people that say 16/44.1 is enough does not doubt that recording and producing in 24/96 is the way. Just that for playback purposes 16/44.1 is ok.
In fact in most cases it is.

Imho, many audiophiles have bougth the mythos that 24/96 is going to save their favorite analog recording, this is bull in most cases a better master and transfer would do that regardless of output format.
It is possible if they get their hands of the original 24 or 48 tracks and remix it from ground up digitally.
But no in most cases good old 16/44.1 is better than the old analog tape.

But I find that the few DVD-A's that are truly recorded at 24bit from the mic stands and onward has an edge here.

mortslim
2009-09-26, 10:40
A PhD student at the MRC Cognition & Brain Sciences Unit, Cambridge, UK recently wrote a two-part article on:

Audiophiles and the limitations of human hearing

The other week Gizmodo posted an amusing rant about a set of $7250 speaker cables, and the gushing review they received. Among other things, the reviewer referred to the cables as “danceable.” James Randi soon popped around to offer his $1 million prize to the cable company, if they could prove that their cables outperform “normal” Monster cables in a double-blind test.

This is actually an issue of the limits of human perception. Is it really possible to tell the difference between normal high-end equipment, and equipment that veers into the audiophile range? It’s clear that according to many audiophiles, the answer is going to be yes. Wikipedia informs us that there are actually two schools among audio enthusiasts: the objectivist school, which favours double-blind testing, and subjectivists, who favour a more philosophical approach. The review that caused so much ire comes from Positive Feedback, an online magazine that concerns itself with the “audio arts” – guess which school they subscribe to.

Among subjectivist audiophiles, there is a belief that almost any change to the stereo setup results in a perceptible difference in sound. This results in bizarre behaviours, as in this picture from the Positive Feedback staff page:

(see picture below)

Note how the speaker cables are carefully propped up on stilts to keep them off the floor, and how what looks like power amplifiers are propped up on massive slabs of wood (I can assure you those didn’t come from the local lumberyard). Another nice example comes from an article in Hi-Fi magazine Masters on Video and Audio:

“The [product] tightened up the sounds of a wide variety of equipment, the improvements often most noticeable in the bass. Imaging and focus usually improved, as did the interstitial quiet, which raised the level of overall palpability, air, and transparency.”

The product? Shelves.

There is one obvious objection to raise here: judging by the pictures of the reviewers on sites such as Positive Feedback, most of them are in their 40’s and beyond. As the following figure shows, this spells trouble:

(see graph below)

I grabbed this from a lecture handout, so unfortunately I don’t know the source. The lines plot performance at detecting sounds over age, with each line representing a frequency. In this case “hearing level” is a standardised measure where normal hearing is at 0 dB. The clear pattern is that the higher frequencies disappear with age. This figure only goes up to 6 khz, but it’s worth noting that the human ear can hear up to 20 khz, and that the loss is more dramatic the higher up you go.

In other words, it’s not really worth trusting an audio reviewer who is older than you are, because there is a range of higher frequencies that you can hear while they cannot.

Apart from the overall lack of evidence and the sheer physical implausibility of some of the products, there is some classic research in social psychology that have implications for this topic.

Cognitive dissonance theory was primarily developed by Festinger. Briefly, the idea is that when the individual finds himself in a state where internal beliefs conflict with reality, there is dissonance, which is an unpleasant state. The individual may then employ a number of mechanisms to get around the dissonance, ranging from simply acknowledging that the beliefs were wrong to attacking the reality of external events, or devaluing the conflict.

The implication for consumer behaviour is that when your green $7250 cables arrive in the mail and you plug them in, finding that they do nothing would result in unacceptable dissonance. In fact, cognitive dissonance theory predicts that the more you pay for the cables, the more inclined you will be to conclude that they sound good, regardless of the actual quality of the cables. In this context, it is worth noting that the Positive Feedback website states that their policy is that reviewers should own the equipment they review, which is a very unusual policy in light of cognitive dissonance theory.

There is another classic social psychology study that is relevant here: Sherif’s investigation of the autokinetic effect (1935). To observe this effect, place yourself in an absolutely dark room with a single, faint light source. The spot of light will appear to move around as a result of small eye movements that your brain normally filters out. Sherif’s participants didn’t know about this however, so they really thought the light moved.

When the participants rated when the light moved individually, there was considerable variation between the participants in how far the light moved. Sherif then placed participants in groups and asked them to call out the movements of the light. Now, there was a convergence effect, so that the estimates of the different participants came closer to each other, and remained close in subsequent individual re-tests.

If you and your friend are listening to a new stereo and she mentions that the low bass sounds a bit flat, you are going to hear it too. The sound itself is ambiguous, not to mention the terminology that audiophiles use, so Sherif’s study suggests that in such situations, you will align with the group. You can imagine that this tendency to conform is quite useful in many real-life contexts, but it does mean that wine sampling and stereo testing are unlikely to reflect anything other than your tendency toward conformity. That doesn’t mean it can’t be fun, of course.

You can test this out yourself if you ever find yourself at a wine sampling. Make up associations: say the wine tastes like blackcurrant (always a a winner), sandal wood, tobacco, myrrh. As long as your ideas aren’t too far off, you will find that others suddenly experience the taste too.

While our senses are rather limited, our ability to fool ourselves is almost endless.

http://phineasgage.wordpress.com/2007/10/13/audiophiles-and-the-limitations-of-human-hearing/#comment-420

mortslim
2009-09-26, 10:40
Here is part 2 of the article:

As a continuation of the recent post on audiophiles, let’s look closer at how good we are detecting the compression in digital music formats.
Most music formats, such as MP3 or the AAC format used by iTunes, define the rate of compression as the number of bits that is used to encode each second of music. The standard bitrate, as used by the iTunes Music Store and elsewhere, is 128 kbit/s. Music geeks (myself included) tend to use slightly higher bitrates, while the proper audiophiles use lossless formats that compress the file without actually removing any information. Recently, Radiohead released their new album as a free download, only to experience some fan backlash for their choice of a 160 kbit/s bitrate. Critics bemoaned the fact that this was half as much as the 320 kbit/s rate that is used on the mp3s available for purchase on their website. By comparison, the bitrate of a normal audio CD is approximately 1411 kbit/s, so clearly a lot of information is removed.

But can you tell the difference? I dug out a few non-peer-reviewed sources to get an idea – if someone knows of peer-reviewed studies into this, I’d be interested to hear about them. The most serious source is probably this 1998 report from the international organisation for standardisation (PDF), which reports some evidence that participants could distinguish 128 kbit/s compression from the original, uncompressed source. Unfortunately, no tests were made above 128 kbit/s. More recent, but less rigorous tests have been reported by Maximum PC and PC World.

Maximum PC elected to report their results participant-by-participant, and with a sample size of 4, maybe that’s just as well. There isn’t enough data reported in this article to actually run a binomial or another significance test, but the overall conclusion seems to be that none of the testers did well at distinguishing 160 kbit/s from the original source.

PC world’s test actually contains some descriptives, and used a sample size of 30. However, they used some fairly obscure ways of reporting their results. Clearly, in a case like this one, the optimal method is to ask the participants to guess which file is the mp3 and which is the cd, and run a number of trials without feedback. With this approach, you can easily assess whether performance is over the level of chance (50%) for each bitrate. With this in mind, here are their results:

(see spreadsheet below)

The percentages represent the proportion of listeners who “felt they couldn’t tell the difference” – once again, this measure is far from ideal. While we have no idea which of these differences are significant, the trend is that the differences in ratings flatten off: there appears to be no difference in quality between 192 kbit/s and 256 kbit/s, and in the case of MP3s, no real difference between 128 and 192.

These studies aren’t exactly hard science, they do seem to indicate that those complaining about Radiohead’s 160 kbit/s bitrate wouldn’t necessarily be able to distinguish it from CD quality, let alone a 256 kbit/s mp3. This illustrates the human tendency to overestimate our own perceptual ability – if we know that two things are different, we will find differences, imagined or otherwise. Blind testing is the only way to establish whether a genuine difference in sound quality exists, yet, this is very rarely done.

If you want to test your own ears, try these examples. With the above in mind, it would be best to get a friend to operate the playback, so that you can’t tell from the outset which file is which. If you run a large number of trials, you can also look up whether your performance is above chance in this Binomial probability table. In psychology, .05 is the commonly accepted p value, so as an example you would need to get 15 out of 20 trials correct for your performance to be significantly better than chance at this level.

Update: Dave over at Cognitive Daily has answered my prayers by carrying out a nicely designed test of performance at discriminating different bitrates. In a nutshell, his results confirm the ones reported here – Although there participants rated the 64 kbit/s tracks as significantly poorer in quality, no differences appeared between 128 and 256 kbit/s. Read the complete write-up here: http://scienceblogs.com/cognitivedaily/2007/11/few_listeners_can_distinguish.php.

http://phineasgage.wordpress.com/2007/10/16/hearing-limitations-pt-2-distinguishing-mp3-from-cd/

firedog
2009-09-27, 04:49
Mortslim-

I've seen all those "scientific" tests. Basically, they prove nothing.
Why? Again, what equipment was used, and how was it set up? Is it really good enough to bring out the differences? Some of the tests you bring up fail on these counts.

Again, who are the listeners? Audio perception is a learned skill. Lots of people can't hear differences, that doesn't mean they don't exist. One of my friends can't hear the difference between my audiophile setup and his ipod with 128k files. So what? That "most" people can't hear something only proves that "they" can't hear it; it proves nothing about what others can hear. I couldn't find the link just now, but I saw a test done with music writers and musicians (i.e., professional listeners) who correctly identified changes in source material or equipment 90% of the time.

Finally, there is even the question of the validity of the testing method itself. How do you test audio perception? Quick A/B comparisons? Long term listening? Just a few of the questions involved. Double blind testing as done in some of these tests is not necessarily the best method for testing.

Hearing and age? Again, not necessarily meaningful. You seem not to be able to understand the concept that hi-res reproduction isn't simply about hearing higher frequencies. BTW, I had my hearing tested, and scored about 20 years younger than the chart you think is gospel. The data is based on average or typical results. Means nothing regarding a specific individual.

Finally - your "point" about audio snake oil. So since some audiophiles may fool themselves, the conclusion is then that everything hi-end is bullshit?
Sorry, not a logical conclusion.

Again, from everything you write, it seems clear you've made up your mind without ever hearing good hi-resolution sound reproduction. A closed mind doesn't listen any more accurately than one that is easily suggestible. Maybe you're the one fooling yourself.

firedog
2009-09-27, 05:26
Here is part 2 of the article:


Update: Dave over at Cognitive Daily has answered my prayers by carrying out a nicely designed test of performance at discriminating different bitrates. In a nutshell, his results confirm the ones reported here Although there participants rated the 64 kbit/s tracks as significantly poorer in quality, no differences appeared between 128 and 256 kbit/s. Read the complete write-up here: http://scienceblogs.com/cognitivedaily/2007/11/few_listeners_can_distinguish.php.

http://phineasgage.wordpress.com/2007/10/16/hearing-limitations-pt-2-distinguishing-mp3-from-cd/

I clicked on the link and read the write up. Suprise - your synopsis of the survey distorts what it said. Yes, most listeners didn't differentiate. But, and I quote:

"There was, however, a small, significant correlation (.09) between listeners who had purchased their own (presumably better) external speakers and ability to discern the difference between the Copland recordings."

"Those rating themselves as more extreme audiophiles were more likely to be able to detect the difference between the different data rates of the Santana MP3s. This was not attributable to their having better headphones; they simply appear to have better knowledge or hearing ability than those who aren't audiophiles. Even so, the correlation between audiophilia and ability to detect the better Santana recordings, though significant, is not very strong: just 0.17.

Part of this may be due to the particular excerpts I chose for the study. As many commenters pointed out two weeks ago in the survey thread, the easiest way to discern artifacts due to MP3 encoding is in the cymbals, and neither the Copland nor the Santana had cymbals. I suspect the acoustic guitar in the Santana has some of the same properties as cymbals, which is what made encoding differences easier to detect there than in the Copland."

In short, the authors of the study confirmed: (a)that their methodology and choice of music was possibly flawed; (b)that some individuals (such as audiophiles) appear to have better knowledge and ability to detect differences. So actually this study proves the points I've been making, and not yours.

RabanePaco
2009-09-27, 09:40
I've seen all those "scientific" tests. Basically, they prove nothing.
Why? Again, what equipment was used, and how was it set up? Is it really good enough to bring out the differences? Some of the tests you bring up fail on these counts.


You also probably have to be under 25 years old to hear any major differences and you should not have been listening to 100 dB pop all the time. Over 50 your hearing greatly deteriorates.

Most standard FM broadcast don't seem to be much better quality then 128 kb/s.

Anyone who uses an $10, or even standard iPod earplugs also can ignore this discussion completely.

I rip CD's standard on 192 kb/s for popular music or older recordings up to 320 kb/s for complex classical music. Once I had a rip I didn't like (192 kb/s) for which I could hear a small but discernible difference with the original CD. I re-ripped at 320 kb/s. I think now it is o.k.

Radified Guide to Ripping & Encoding CD audio (with EAC & the LAME MP3 encoder) (http://mp3.radified.com/) has references into research of and listening to ripping.

[PS. I wished I still had the hearing of 25].

Phil Leigh
2009-09-27, 09:49
Most standard FM broadcast don't seem to be much better quality then 128 kb/s.

.

Then you need a better broadcaster, tuner or both!

RabanePaco
2009-09-27, 15:42
Re: Most standard FM broadcast don't seem to be much better quality then 128 kb/s.


Then you need a better broadcaster, tuner or both!

Do you have a more reliable quote for measurements about FM radio quality? [Maybe is is 192 kb/s, but it certainly is not CD quality.] A better tuner doesn't help is for a bad source.

Raptus
2009-09-29, 05:07
There are two crystals, one which is used for 44.1 and 88.2 and the other used for 48 and 96.
Having separate/native crystals for 44.1 and 48 KHz is more important then support for hi-res. Thx for doing something right.

Raptus
2009-09-29, 06:03
stuff
Do you at least acknowledge the necessity of a means to filter out bias when making comparisons or assessing quality?


I couldn't find the link just now, but I saw a test done with music writers and musicians (i.e., professional listeners) who correctly identified changes in source material or equipment 90% of the time.
Without details about the test this statement is absolutely meaningless. Given enough "freedom" anyone can setup a test that produces exactly the intended results.


Finally, there is even the question of the validity of the testing method itself. How do you test audio perception? Quick A/B comparisons? Long term listening? Just a few of the questions involved. Double blind testing as done in some of these tests is not necessarily the best method for testing.
So, what is "the best method" in your opinion?


Finally - your "point" about audio snake oil. So since some audiophiles may fool themselves, the conclusion is then that everything hi-end is bullshit?
No, but quite a few self entitled "hi-end" companies are making money from the lifestyle they associate their products with through marketing, where the actual engineering of the product is secondary. In fact, you can observe that quite a few hi-ends introduce wanted sonic non-linearities into their products, as to create a characteristic, recognizable sound, justifying their exclusive status, much like sportscar companies tune the sound of their engines. At this point the concept of HiFi has been replaced by something else. That is not to speak of the cases where the product has an extremely polished exterior, but when measured out reveals catastrophic qualities, with minimal effort and cost put into the parts of the system that really matter to sound quality.

So no, not all audiophile gear is BS, but if you want HiFi, you are much more prone to find a good product in the pro gear section.


it seems clear you've made up your mind without ever hearing good hi-resolution sound reproduction.
a) what is a "good hi-resolution sound reproduction" in your opinion?
b) why do you think you can assume people have never heard "good hi-resolution sound reproduction"?


Audio perception is a learned skill
Agreed.

firedog
2009-09-30, 03:28
Do you at least acknowledge the necessity of a means to filter out bias when making comparisons or assessing quality?





So, what is "the best method" in your opinion?






a) what is a "good hi-resolution sound reproduction" in your opinion?
b) why do you think you can assume people have never heard "good hi-resolution sound reproduction"?



Filter out bias - sure, I'm in favor. But as most of the supposed "scientific double blind tests" are done, that is all they do. They don't actually test what people (at least some people) can hear on a quality setup. Best method - I'm not sure. But again, most of the tests commonly referred to as "proving" what people can't hear have obvious problems with the methodology - either in listeners chosen or in equipment. Beyond that, there is the obvious problem of setting up a test that actually mirrors how people perceive music. And I'm not at all sure that quick A/B tests of two files are the way to do it. In fact, I could make an argument that that's the worst way to do it.

A from above)something with enough quality to reproduce the supposed differences between the files being tested. Mid-fi systems won't do it.

B from above)I made no comment about "people" in general . I was referring specifically to one poster, and from his posts it could be inferred that he'd never actually spent much time listening to hi-res files, yet "knew" that they couldn't sound better than 16/44.1 to human ears.

Raptus
2009-09-30, 04:43
Filter out bias - sure, I'm in favor.
Good, then we can agree on the fact that tests that do not filter out bias are worthless.


But as most of the supposed "scientific double blind tests" are done, that is all they do. They don't actually test what people (at least some people) can hear on a quality setup.
You're ignoring that these tests actually do produce meaningful results, being extensively used by the more serious/honest companies in developing their equipment, or in the development of audiocodecs.

Also, you are assuming that a better setup will expose more flaws. This is not always true. Specifically in lossy audio compression the calculation of masked components is based on the assumption of a linear playback chain. Thus worse, less linear chains end up revealing more quantization noise. I experienced this a couple of times.


But again, most of the tests commonly referred to as "proving" what people can't hear have obvious problems with the methodology - either in listeners chosen or in equipment.
Well ofc you need to look at the test in detail, you can't generalize.

Also, there are obviously snake oil products on the market, or ones, where the differences are measurable, but not audible. It shouldn't be surprising when the tests confirm exactly that.


Beyond that, there is the obvious problem of setting up a test that actually mirrors how people perceive music. And I'm not at all sure that quick A/B tests of two files are the way to do it. In fact, I could make an argument that that's the worst way to do it.
Since we agreed on the necessity to filter bias, I hope you mean a blind form of A/B.
Blind tests don't have to be quick. In fact, if you are serious about testing, you could easily make them long term. Blind tests are not as hard to do as many audiophiles make them out to be. Except ofc, when they confront audiophiles with the truth ;)


A from above)something with enough quality to reproduce the supposed differences between the files being tested. Mid-fi systems won't do it.While this tends to be true, I strongly oppose to this as being the law (btw, how do you define mid- and hi-end?). Typically a 200$ headphone + decent amp will reveal much more flaws coming from lossy compression then any five figure setup.


I was referring specifically to one poster, and from his posts it could be inferred that he'd never actually spent much time listening to hi-res files, yet "knew" that they couldn't sound better than 16/44.1 to human ears.
You or somebody mentioned the problems with using the steep filters required to do the necessary filtering above 20Khz. Whiles this is correct, it ignores that by now almost all players/DACs use oversampling, which moves the cutoff much higher while allowing filters with reduced steepness.
If there is obvious differences between the same recording at 44.1 and 96Khz there is a problem somewhere else in the chain, not the recording resolution itself.

firedog
2009-10-01, 02:22
Hi-

1. "You or somebody mentioned the problems with using the steep filters required to do the necessary filtering above 20Khz. Whiles this is correct, it ignores that by now almost all players/DACs use oversampling, which moves the cutoff much higher while allowing filters with reduced steepness.
If there is obvious differences between the same recording at 44.1 and 96Khz there is a problem somewhere else in the chain, not the recording resolution itself."

Don't agree about "problem somewhere else in the chain". From the beginning my sole contention was that hi-res files can sound better than standard 44k files. Other posters disagree. The "double blind tests referred to in proving that "people can't hear this or that" often have obvious flaws when you check how they were conducted, so I don't take them seriously.

As far as obvious differences between files, what do you call obvious? I can hear differences when I listen to 96k files. Are they huge - like the differences between a 64k mp3 and a CD? No. But they can be heard. I don't know if that makes them "obvious" or not, and I don't care. I care only that I can hear the difference.

Filters: Read up on pre-ringing and newer players using aphodising filters - there are only a few very expensive ones on the market. All reports I've heard give these players high marks for improving the sound of CD playback. If you read up on pre-ringing, you'll also find out one of the reasons hi-res playback is superior, at least in theory.

What system? I'm not sure why you're stuck on this and asking me about a dollar figure for the equipment involved. The point of the discussion has been about hi-res files, not lossy reproduction. You need a fairly high end system to reveal the advantages of these files. If a test were conducted, we would need to agree on the playback system used ahead of time. A typical home stereo probably isn't good enough to really show the differences between 16/44 and 24/96.

Keymaster
2009-10-01, 03:47
"Trust your ears, not your eyes"

-A prof. of mine when studying recording engineering.

Raptus
2009-10-01, 03:51
Don't agree about "problem somewhere else in the chain". From the beginning my sole contention was that hi-res files can sound better than standard 44k files.
Yes they can, but let be show you just two examples of cases where it wasn't the recording resolutions fault:

- Studios promoting hi-res formats like DVD-A or SACD have been found guilty of using inferior mastering for the standard CD audio layer of those discs, as to promote the superiority of the hi-res format.
- Soundcards/DACs with just one frequency crystal and mediocre resampling algorithms. I've come across such soundcards, which produce obvious differences for signals with much high frequency content. The first one I stumbled upon made 44.1 sound significantly different than 48KHz. Couldn't believe it myself at first, until I found that the card had only one 96Khz crystal. Must have been using simple padding/truncating to "adapt" the signal...
Thats why I applauded SlimD/Logitech before for using two crystals.


Other posters disagree. The "double blind tests referred to in proving that "people can't hear this or that" often have obvious flaws when you check how they were conducted, so I don't take them seriously.
You can't discredit a good method just because someone made bad use of it.


Filters: Read up on pre-ringing and newer players using aphodising filters
I am educated in signal processing so thank you for reminding me of pre-ringing and <buzzword>.


What system? I'm not sure why you're stuck on this and asking me about a dollar figure for the equipment involved.
I am asking for your definition of hi-end because you keep throwing the term around, as if there were consesus on what it means.
Also, I am stating that, for revealing imperfections coming from lossy audio compression, the "hi-endness" of the system is not as important as you say.


The point of the discussion has been about hi-res files, not lossy reproduction.
You mean lossy coding. Every reproduction into the analog world is lossy.


You need a fairly high end system to reveal the advantages of these files.
There it is again...


If a test were conducted, we would need to agree on the playback system used ahead of time. A typical home stereo probably isn't good enough to really show the differences between 16/44 and 24/96.
This is leading nowhere. Very probably you are never going to do a double blind listening test in your room, with your setup. And you will keep telling us you hear a difference, while everyone that can't confirm, must either be using a lesser-end system or have untrained ears. In the scientific world, for findings to get credibility, they must be repeatable by peers. If your findings work only in your room, your system and your ears, what general validity have they? (And that is assuming flawless testing method and execution).

EDIT: Actually thats pretty convenient, if you are the only one in the world that hears "it", you can be happy to be part of a very exclusive circle.

tommypeters
2009-10-03, 08:35
The problem with double-blind tests can be illustrated by the triangle below:

http://www.eyetricks.com/0406.gif

...and also by the black dots (how many are they, try to count them) here:

http://eyetricks.com/0101.gif

When changing back-and-forth between different equipment your ears will start to fool you the same way as your eyes in the illustrations above. Because it's neither your ears nor your eyes, but your brain that does the fooling. If there's no easily-detected distortion, but rather "lack of information", your ears will soon fill that in and you will seemingly switch back-and-forth between similar-sounding audio.

Meridian did a test of double-blind tests and concluded that they couldn't use it for internal tests. And since internal tests aren't to prove anything for anyone it's perfectly alright for them to use other methods. But for a reviewer, while flawed, double-blind tests may still be the way to go. But he/she should take them for what they're worth.

firedog
2009-10-04, 02:39
The problem with double-blind tests can be illustrated by the triangle below:

http://www.eyetricks.com/0406.gif

...and also by the black dots (how many are they, try to count them) here:

http://eyetricks.com/0101.gif

When changing back-and-forth between different equipment your ears will start to fool you the same way as your eyes in the illustrations above. Because it's neither your ears nor your eyes, but your brain that does the fooling. If there's no easily-detected distortion, but rather "lack of information", your ears will soon fill that in and you will seemingly switch back-and-forth between similar-sounding audio.

Meridian did a test of double-blind tests and concluded that they couldn't use it for internal tests. And since internal tests aren't to prove anything for anyone it's perfectly alright for them to use other methods. But for a reviewer, while flawed, double-blind tests may still be the way to go. But he/she should take them for what they're worth.

yes, that's why I don't believe in quick A/B double blind tests, or even in just non-scientific testing with quick A/B switches.