Redbook compared to hi rez format

[magiccarpetride]

A few years back I got very enthused about the promise of the high definition (a.k.a. hi rez) digital format. This enthusiasm got fortified for me after doing a comparative listen between the Getz/Gilberto famous album from the early '60s (the one containing "The Girl From Ipanema"); I listened to the hi rez and the regular CD version, and found that hi rez sounded way superior.

However, upon further examination, I've realized that what was contributing to the superior quality of the hi rez version was a different mastering, not necessarily the difference in format. In other words, the mastering on the CD was sub standard, while the mastering for the hi rez version was done properly, which resulted in a much more musical presentation.

Disappointed, I became dismissive of the purported qualities of the hi rez format, chalking it up to the marketing b.s.

But the little worm in my head wouldn't leave me alone, so yesterday I took the time to downsample one track of the Weather Report "Heavy Weather" hi rez version, and to play it back to back with the hi rez version. The track was "The Juggler". First, I played the crunched down redbook version (16 bit/44.1 kHz), and then played the hi rez (24 bit/96 kHz) version. This time, the hi rez version sounded convincingly better, and it wasn't due to the different mastering for the CD format. We're talking the same source material, only different in the fact that one copy was dithered down to 16 bit. But I was really able to hear how the 24 bit version sounded fuller, more musical, with softer yet at the same time more present details.

Am I imagining things here, or had any of you also noticed the differences in two formats?

[Werner Berghofer]

Am I imagining things here

Yes, I think you do. Head-Fi: 24 bit vs 16 bit, the myth exploded! (http://www.head-fi.org/t/415361/24bit-vs-16bit-the-myth-exploded)

Werner.

[magiccarpetride]

Yes, I think you do. Head-Fi: 24 bit vs 16 bit, the myth exploded! (http://www.head-fi.org/t/415361/24bit-vs-16bit-the-myth-exploded)

Werner.

I'm well aware of the theoretical explanations. What I'm more interested in here is learning about the empirical deviations (if any) from the theory.

Or, as one eminent politician in the former Soviet Union had declared, after visiting some leading edge scientific research facility in Moscow: "This new invention obviously works well in practice, but does it work in theory?"

[Cans]

I reckon the hi res recording of an album commercially available is a more pleasant listen in the vast majority of cases than the redbook. The logic my head has come up with is there are more 0's and 1's so more information can be stored within each track. More information brings the music closer to the master recording. That's how my mind works on it.
I've listened to the fleetwood mac's back catalogue , the police's back catalogue, elton john's madman across the water. Stunning. There are loads of others at hdtracks , if you enjoyed the cd version , it's an easy way to get a bit more money spent. A wothwhile upgrade, changing the software rather than the hardware :-)
My mind isn't closed to this by any means , it's a voyage of discovery, I will soon be ripping my own vinyl and shall see where that takes me , but I won't be spending 1000's on it :-)

[bobbasrah]

Other factors to consider beyond assumptions that downsampling is a true mirror of the hirez, is on playback - Are the various resolutions handled in the same way by the conduit (usb/spdif/?), treated in the same way by the DAC, or filtered differently before it hits the analogue chain?
It is not necessarily so simple an answer.....

[dave2010]

It's difficult to be sure what's going on, if anything. It is possible that in many cases that those who claim to be able to tell a difference between "hi res" formats and standard CD formats actually can't, and it's also the case that often the mastering of different formats of the same music may give either subtle, or sometimes, not so subtle differences.

I did some simple "experiments" years ago with adding noise to audio signals. Obviously if the noise levels are high enough it is perceived as noise, but if it is low then it probably won't be. Signal plus noise was always perceived as being quieter, with less dynamic range than the same signal without the noise. This arguably fits in with a model for ear/brain processing in which our ear/brain systems compute a noise floor, and then "measure" signals relative to that. It follows that perceived loudness of a wanted signal is not just dependent on the total energy within the audio, which should be greater if noise is present.

A more extreme example of the effect of noise is trying to listen to music while in a car. As the road and other noise increases, the music will be submerged in the noise, and in order to perceive "loud" sounds, the levels have to be turned up. When the car stops the music will often be considered to be too loud.

However, here we are presumably not discussing anything as crude as sound in a noisy environment, so the interesting question is whether different levels of background noise can have an effect. When we discuss CDs and assume a noise level of 96dB relative to the signal, the first question to ask is can we hear or detect the noise relative to maximum level. The answer could well be "no". Then ask, what happens if the signal (wanted) is lower? Some recordings are quiet to start with, and within such quiet recordings, parts will be quieter still. With PCM recordings, low level signals (wanted) may not be very much above the noise floor if 16 bits are used, and in that case the effects could be audible. This is slightly speculative, but it seems much more likely that for some quiet music that 24 bits would really make a difference, as there would be hopefully virtually no chance that noise in the low order bits of such signals would have an audible effect.

I don't know the answer in this particular case - whether 16 bit is acceptable or not, and indistinguishable from higher resolution recordings. Could we consider a slightly different process - take a 16 bit (or a 20 bit recording if you can get one), and then gradually degrade the signal by knocking off the low order bits, until the point at which some/most/all people can hear the difference. The music will probably still be quite recognisable at 8 bits resolution, though sounding somewhat rougher. At 12 bits or higher resolution, many people may not be able to hear a difference, at least with casual listening. Although standard CDs have always been 16 bits, it was noted by some recording engineers and technical people that early equipment was really only working with 13 or 14 bits, as the equipment was not good enough. Also recording engineers who were used to working for LPs did not notice some low level problems with the audio until some while later, once they had become more accustomed to digital working. This suggests that there can be audible differences up to around 16 bits resolution, at least for people with trained ears. Whether anyone can really hear much difference once the resolution used for recording goes above 16 bits is still an interesting question.

The modern practice of using more bits during recording and processing, and also higher sampling rates does make some sense, as this can minimise errors arising from digital mixing and filtering.

Mp3s have been denigrated (and perhaps rightly), but I have a strong suspicion that if an mp3 file is burned to an audio CD that the quality degrades further. Even allowing for the "fact" that (some) mp3s are horrible, they are capable of being processed with more than 16 bits. Most players will work with floating point operations (more than 16 bits) My feeling has been that taking an mp3 file and forcing it down to 16 bits to fit on a CD can result in an audible further loss of quality. This is again only an opinion, but could indicate that 16 bit resolution is not quite sufficient for good quality audio - at least under non optimum conditions.

We also have to consider playback levels, which correspond in some ways to magnification in visual systems. Although the new iPad has a better display than previous models, and there is a clear difference when you look at a magnified image (see the Apple web site - http://www.apple.com/uk/ipad/features/ - look at the pages about the Retina Display), at the normal magnification levels used on plain text it is actually hard to distinguish between them. If you go looking for noise, by magnifying in a visual system, or turning the volume up in an audio system, you may well find it, but under normal conditions it may not be perceptible.

[The Vinyl Adventure]

I have to say, I'm so far beyond caring for the arguments for and against
I have high res music which sounds better than the cd version and for me that is fact
This could be for the following reasons
High res is actually better
The version mastered for high res was mastered with more care and attention
My equipment (the linn ds) prefers high res files (this might be due to linn wanting to sell more high res recordings; especially I you like a conspiracy)
Knowing I'm playing a high res file might positively effect my mind set and increase my listening pleasure
The little pixies in my linn that feed on bits get more to eat and fart better tunes
For all I care any of the above might be true or complete crap ... I don't care, if it works for me I will spend my cash if I see fit
That said, I have spent about 3 weeks now all day every day listening to spotify through my caiman and grados whilst writing thousands of lines of HTML and CSS ... I've found I can listen to music quite attentively under those circumstances and not once has the quality from spotify bothered me ...
Maybe that makes me deaf
I don't care about that either
I'm comfortable with the money I spend on music, the quality of recordings I buy, the quality of both my hifi and headphone system regardless of source
I don't want to sound smug, (I do a little) but whilst people worry about that sort of thing some (a lot) of other people are just buying, listening to and enjoying music
It's those of us who are winning (and maybe the record companies who are getting a few extra £££ for selling us "high res" recordings ;))

[StanleyB]

I have to say, I'm so far beyond caring for the arguments for and against
I am getting to that stage as well. Why bother heh ;)?

In a different topic about HIFI design a couple of us touched on the subject of hearing such things as the TV line frequency and the FM stereo pilot tone. Of course, nobody has so far dared to suggest that anyone who claims to be able to hear those frequencies must be mistaken or their FM tuner of TV is faulty. But a mention of audio file formats suddenly makes everyone an expert, with a couple of them even going so far as to imply that people are claiming things.

The word 'claim' in itself makes clear that one is challenging the honesty of the person who made the 'claim'. i.e false prophets. This doesn't make for pleasant topic participation. I would go so far as to 'claim' that is is a turn off for many who would have loved to reply and offer their own opinion, but who are of the opinion that it is just not worth the hassle if their integrity would then be questioned by learned friends who seem to have all the answers.

[morris_minor]

There's also the link I posted here (http://theartofsound.net/forum/showthread.php?t=16828).

I think so much in "hi-fi" is down to psychological factors. We're convinced of something because we expect it to sound/act the way we want it to.

[dave2010]

I am getting to that stage as well. Why bother heh ;)?

In a different topic about HIFI design a couple of us touched on the subject of hearing such things as the TV line frequency and the FM stereo pilot tone. Of course, nobody has so far dared to suggest that anyone who claims to be able to hear those frequencies must be mistaken or their FM tuner of TV is faulty. But a mention of audio file formats suddenly makes everyone an expert, with a couple of them even going so far as to imply that people are claiming things.

The word 'claim' in itself makes clear that one is challenging the honesty of the person who made the 'claim'. i.e false prophets. This doesn't make for pleasant topic participation. I would go so far as to 'claim' that is is a turn off for many who would have loved to reply and offer their own opinion, but who are of the opinion that it is just not worth the hassle if their integrity would then be questioned by learned friends who seem to have all the answers.Stan

Sorry you don't like the word "claim" - I don't think it has the pejorative meaning you ascribe to it.

If you wish you could claim that you can hear the buzzing of TV line transformers (do they still have them), or fluorescent lights "whistling". I could probably not disagree with you, as I've not been able to hear most of these kinds of sound for quite a while, though just sometimes I can walk into a lab with computer displays turned one, and I find that somewhat unpleasant and go round turning them off.

I used to get on buses which had very pronounced whistling from fluorescent tubes, and they were definitely audible. I know people who can detect those devices which output high frequencies to deter mice. I'm pretty sure I can't do that.

[dave2010]

I have to say, I'm so far beyond caring for the arguments for and against

For all I care any of the above might be true or complete crap ... I don't care, if it works for me I will spend my cash if I see fit

That said, I have spent about 3 weeks now all day every day listening to spotify through my caiman and grados whilst writing thousands of lines of HTML and CSS ... I've found I can listen to music quite attentively under those circumstances and not once has the quality from spotify bothered me ...
Maybe that makes me deaf
I don't care about that either
I'm comfortable with the money I spend on music, the quality of recordings I buy, the quality of both my hifi and headphone system regardless of source
I don't want to sound smug, (I do a little) but whilst people worry about that sort of thing some (a lot) of other people are just buying, listening to and enjoying music

It's those of us who are winning (and maybe the record companies who are getting a few extra £££ for selling us "high res" recordings ;))Mostly this represents the voice of sanity, though touching on money, if I'm really happy enough with CD quality why should I buy recordings again such as these?:

http://www.amazon.co.uk/Mozart-Last-Symphonies-Signature-Edition/dp/B0079J27SS/ref=sr_1_2?s=music&ie=UTF8&qid=1333450339&sr=1-2 (I've probably got some of these already)

http://www.amazon.co.uk/Elgar-Concerto-Pictures-Farewell-Signature/dp/B0079J26XO/ref=sr_1_1?s=music&ie=UTF8&qid=1333450339&sr=1-1 Doesn't every classical music enthusiast have at least one copy of this performance already?

http://www.amazon.co.uk/Beethoven-Triple-Concerto-Signature-Edition/dp/B0079J27JW/ref=sr_1_4?s=music&ie=UTF8&qid=1333450339&sr=1-4 Yet another which I have already

and finally

http://www.amazon.co.uk/Bruckner-Symphonies-No-8-Signature-Collection/dp/B0079J26S4/ref=sr_1_8?s=music&ie=UTF8&qid=1333450339&sr=1-8 which I don't have, and I don't know the performance.

There are still people around for whom "even" £10 is quite a sizeable sum, and it seems pointless to duplicate existing CDs unless there is a significant benefit in audio quality, with the additional premium for these SACDs - not to mention of course the need to have an SACD player in the first place if the SACD layer(s) are to be used.

I do have SACDs, I do have an SACD player, and they do, generally sound slightly better. Despite that however, I hardly ever play them, and have now mostly moved back to CDs which I rip and play via my Squeezebox.

[bobbasrah]

I know people who can detect those devices which output high frequencies to deter mice. I'm pretty sure I can't do that.

Deter Mice you mean Dave?:D


There are still people around for whom "even" £10 is quite a sizeable sum, and it seems pointless to duplicate existing CDs unless there is a significant benefit in audio quality, with the additional premium for these SACDs - not to mention of course the need to have an SACD player in the first place if the SACD layer(s) are to be used.

I do have SACDs, I do have an SACD player, and they do, generally sound slightly better. Despite that however, I hardly ever play them, and have now mostly moved back to CDs which I rip and play via my Squeezebox.

SACD prices were always high, but downloads in hirez from admitedly restricted catalogues are pretty eye watering pricing by comparison with the selection you posted. My SACD player is in a bag on a shelf, the discs are all in archive boxes, neither of which have been in use for over 5 years now.;)
Only have the vinyl left to do.....the saga....:(

Having collected the best reviewed classical performances on the basis of quality of play and recording, there are only a few performances left to get. I agree with your own position Dave, if I already have the album, there will NOT be any pressing desire to pick up another version, even for a Tenner. :mental:

[morris_minor]

I do have SACDs, I do have an SACD player, and they do, generally sound slightly better. Despite that however, I hardly ever play them, and have now mostly moved back to CDs which I rip and play via my Squeezebox.

Ditto.

:)

[dave2010]

Bob

No - my attempts to deter/scare away mice generally fail. I've given up, and gone back to good old fashioned mouse traps. Nasty, but it it's them or me, it's not going to be me.

SACDs are also a bit of a pain, as it's not easily possible to rip them. I believe it can be done, but only with professional equipment, so we'd lose limbs in the process. The CD layer can be ripped, but that's not really the point, is it?

Maybe I'll go back to SACDs in a while - though the whole anti-copying business (anti-ripping business) puts a downer on it. Perhaps it is possible to extract a digital channels using an HDMI link and other matching kit. Bit of a shame really.

[WOStantonCS100]

The logic my head has come up with is there are more 0's and 1's so more information can be stored within each track. More information brings the music closer to the master recording.

It's not just your logic (which I'm glad you pointed out). It's simply how things work. This is the one point that all arguments against high definition/resolution digital always fail on. Everything else is a side step; playback chain, cabling, mastering, etc. etc. That's all well and good; but, it doesn't have anything to do with the real issue which remains that high definition/resolution digital, by virtue of it's higher sampling frequency, naturally and indisputably, retains more of the original analog waveform in the recording process than Red Book. That's just fact. Usually, when this tidbit comes up in ye ol' debate... it's glibbed over, side stepped or blatantly ignored. It quickly turns from the objective capturing of more samples (proven) to the subjectivity of whether one cares about or can hear the difference that makes. Strangely, rarely if ever does one find such arguments regarding lo res/hi res video.

[Reid Malenfant]

SACDs are also a bit of a pain, as it's not easily possible to rip them. I believe it can be done, but only with professional equipment, so we'd lose limbs in the process. The CD layer can be ripped, but that's not really the point, is it?
It can be done with an early Sony PS3 that hasn't had updated firmware as far as I'm aware. The DSD stream can simply be ripped to a HDD :eyebrows:

[magiccarpetride]

It's not just your logic (which I'm glad you pointed out). It's simply how things work. This is the one point that all arguments against high definition/resolution digital always fail on. Everything else is a side step; playback chain, cabling, mastering, etc. etc. That's all well and good; but, it doesn't have anything to do with the real issue which remains that high definition/resolution digital, by virtue of it's higher sampling frequency, naturally and indisputably, retains more of the original analog waveform in the recording process than Red Book. That's just fact. Usually, when this tidbit comes up in ye ol' debate... it's glibbed over, side stepped or blatantly ignored. It quickly turns from the objective capturing of more samples (proven) to the subjectivity of whether one cares about or can hear the difference that makes. Strangely, rarely if ever does one find such arguments regarding lo res/hi res video.

Good point. Another point, which is arguably DAC-dependent, is that DACs seem to use different filters depending on the resolution of the source material. Higher resolution material (say, 24 bit/96 kHz) will allegedly elicit different treatment at the DAC level than the 16 bit/44.1 kHz material. Even if it can be proven beyond a trace of a doubt that the extra information supposedly encoded in the high resolution format is superfluous and makes no theoretical difference, this extra information embedded in the hi rez still is not ineffective, since it solicits, provokes different type of processing from a DAC. And that difference may account for the difference we're hearing when comparing redbook with hi rez.

[Reid Malenfant]

Good point. Another point, which is arguably DAC-dependent, is that DACs seem to use different filters depending on the resolution of the source material. Higher resolution material (say, 24 bit/96 kHz) will allegedly elicit different treatment at the DAC level than the 16 bit/44.1 kHz material. Even if it can be proven beyond a trace of a doubt that the extra information supposedly encoded in the high resolution format is superfluous and makes no theoretical difference, this extra information embedded in the hi rez still is not ineffective, since it solicits, provokes different type of processing from a DAC. And that difference may account for the difference we're hearing when comparing redbook with hi rez.
Yes, so what you hear from higher resolution wider bandwidth material is intermodulation from stuff outside of your hearing range affecting what you can hear inside the hearing range.

While you can never hear a 30KHz tone, you could certainly hear the effect it has on a lower frequency (that you can hear) due to intermodulation.

This only applies to material with a wider bandwidth like **/96KHz, 24/44.1 or 24/48 will have far less of this effect as the brick wall filtering is just above audible frequencies.

The bigger bit depth of say 24 bit will have much more of an effect on what you hear. 24 bit is 256 times as accurate as 16 bit audio :)

[WOStantonCS100]

The bigger bit depth of say 24 bit will have much more of an effect on what you hear. 24 bit is 256 times as accurate as 16 bit audio :)

...and there is a gradual shift to 32-bit float.

The other day I walked into a department store looking for boxer briefs. ;) The sales person told me that the "A" pair was 256 times more likely to retain it's original shape and not shrink up to the point where it causes intense pain, ultimately leading to my passing out on the floor, over set "B". Needless to say, I didn't go for set "B". :eyebrows:


Good point. Another point, which is arguably DAC-dependent, is that DACs seem to use different filters depending on the resolution of the source material. Higher resolution material (say, 24 bit/96 kHz) will allegedly elicit different treatment at the DAC level than the 16 bit/44.1 kHz material. Even if it can be proven beyond a trace of a doubt that the extra information supposedly encoded in the high resolution format is superfluous and makes no theoretical difference, this extra information embedded in the hi rez still is not ineffective, since it solicits, provokes different type of processing from a DAC. And that difference may account for the difference we're hearing when comparing redbook with hi rez.

A tiring (and somewhat boring) study of the topic at hand revealed an insight on another frequently used argument for Red Book. Namely, the one about it being sufficient enough to recreate frequencies within the realm of human hearing. I'm not the first to point out that music is more than just reproducing a frequency. A frequency generator can spit out A440 or anything else all day long; but... how many of us want to listen to a frequency generator rather than music. Music is a complex beast, in other words, many frequencies at the same time with varying degrees of loudness, timbre, etc. etc. and on it goes. I have no doubt that algorithms can recreate frequencies. However, it seems rather obvious that what is not being recreated, at lower bit rates/sampling frequencies (in comparison to higher rates) is the full complexity of the audible analog spectrum present when said recording was made. If the crux of hi res digital was only about frequencies above 20kHz than I would say "pish" to it as well. But, it's not. It's about retaining and illuminating as many colors and textures between 20Hz and 20kHz as possible, as Mark eluded to in one of the reasons for wanting the ability to capture higher frequencies.

I would also agree that a system (DAC, filters, etc.) designed from the ground up (not reconfigured as an after thought) to reproduce high res digital, as a mission critical goal, is most likely the superior option.

[dave2010]

It can be done with an early Sony PS3 that hasn't had updated firmware as far as I'm aware. The DSD stream can simply be ripped to a HDD :eyebrows:Does that capture the DSD stream, or a PCM near equivalent? If DSD (which could be good), then we need a player, either hardware or software, to play it. Maybe the Sony PS3 you mention does that.

[Reid Malenfant]

As far as I know it captures the DSD stream... Someone on here mentioned it, might have been Twelvebears though don't quote me on that. Actually I'm not sure about the member at all. I suggest you post a new thread about the PS3 & DSD ripping & see what you get!

It's certainly someone I know, but not personally. It was just something that I couldn't help remembering as it kind of stuck out if you know what I mean :D

[morris_minor]

The bigger bit depth of say 24 bit will have much more of an effect on what you hear. 24 bit is 256 times as accurate as 16 bit audio :)

Mark - this is what I've always thought. The opposing stance though is that higher bit rates contribute a wider dynamic range (most of which is unusable in home playback).

I'm thoroughly confused :scratch:, though TBH a well mastered redbook recording can sound as "good", i.e. be as enjoyable on musical and SQ grounds as any hi-res recording. Is not the crux of this issue the fact that the mastering has to be done right in the first instance?

[WOStantonCS100]

I'm thoroughly confused :scratch:, though TBH a well mastered redbook recording can sound as "good", i.e. be as enjoyable on musical and SQ grounds as any hi-res recording. Is not the crux of this issue the fact that the mastering has to be done right in the first instance?

A 750 HP engine is great... a 750 HP engine in a tuned suspension, aerodynamic roadster is an altogether different deal.

Mastering is immensely important. If that's not done right, the format is of little consequence. If the mastering is done well and presented on a format that can't deliver it's full potential, it will sound good while being subject to the limitations of the format. Great mastering on a format that can deliver much of what the master had to offer in the first place is another story entirely.

I recently did an LP transfers of a pristine direct to disc pressing. First I recorded it to cassette (normal bias albeit good quality cassette on a 3-head deck with external dbx Type II). I lived with this recording for a week or so. It's a very good recording and reasonably close to the LP. As analog sources are usually devoid of such, there was no listening "fatigue". Then I recorded the same LP to 7 ips 1/2 track open reel (RMGI SM911 tape also with dbx Type II). With the dbx 200 program route selector I was able to level match the cassette and reel deck and then switch back and forth) while listening with cans on. The cassette was very good. But, it was when I could do a direct comparison with reel that I realized cassette wasn't giving me all there was to get. As it goes, chrome than metal cassette got closer, scarily so.

I've done this test with 16/44, 16/48, 24/44 and 24/96 digital resolutions. I do not think 16/44 is horrid. Nevertheless, in direct comparisons, the 24 bit higher resolutions, to my ear, always sounded closer/truer to the input signal.

[magiccarpetride]

I'm thoroughly confused :scratch:, though TBH a well mastered redbook recording can sound as "good", i.e. be as enjoyable on musical and SQ grounds as any hi-res recording. Is not the crux of this issue the fact that the mastering has to be done right in the first instance?

Yes, and if you have a good source (i.e. properly mastered track), then the format you use for the playback of that track can affect your listening experience. My original example (Weather Report "The Juggler") speaks to that. Both the redbook and the hi rez formats were cut from the same decent master. But if you listen to them side-by-side on a high quality system, you will be able to hear how hi rez offers more warmth, more softness, more ease, and is overly more effortless in delivery. Less congested, shall we say. The redbook version isn't missing any details, they're all there, it's just that these details are not presented in as musical vein as with the hi rez.

[TopBalcony]

Yes, and if you have a good source (i.e. properly mastered track), then the format you use for the playback of that track can affect your listening experience. My original example (Weather Report "The Juggler") speaks to that. Both the redbook and the hi rez formats were cut from the same decent master. But if you listen to them side-by-side on a high quality system, you will be able to hear how hi rez offers more warmth, more softness, more ease, and is overly more effortless in delivery. Less congested, shall we say. The redbook version isn't missing any details, they're all there, it's just that these details are not presented in as musical vein as with the hi rez.

MCR, how were you listening to the hi rez version ? I thought this had only been released on sacd ? Would like to know if there was a pcm download available.

[magiccarpetride]

MCR, how were you listening to the hi rez version ? I thought this had only been released on sacd ? Would like to know if there was a pcm download available.

A friend gave me the PCM rip off the SACD. Sounds close to listening to the LP.