Page 1 of 4 1234 LastLast
Results 1 to 20 of 66

Thread: The truth about high-resolution audio compared with std. CD 44k?

  1. #1
    Join Date
    Sep 2007
    Location
    England
    Posts
    250

    Default The truth about high-resolution audio compared with std. CD 44k?

    This thread is concerned with audio delivery formats and ripping.

  2. #2
    Join Date
    May 2011
    Location
    US
    Posts
    32

    Default First and foremost: recording quality is what matters

    This is somewhat of a complex subject because, as I see (hear) it, there are several factors that appear to affect SQ, whether the recording is 44.1/16 or higher. All of my music is sourced from a computer (Mac Mini). It is either ripped in AIFF format from CDs (44.1/16) or downloaded from Linn Records of HD Tracks ("hi-rez").

    1. In my limited experience, I have found that the quality of the recording is first and foremost. I have 44.1/16 rips that sound stunning, and others that sound lacking to downright poor. Some of this, IMO, depends on the type of music. For example, recordings such as acoustic guitar, vocals, and the like, sound magical because they take advantage of the mid-range capabilities of my Super HL5s. That being said, I have rock rips that also sound terrific and I owe that mainly to the recordings.

    2. I have an HD Tracks download (rock) at 96/24. IMO, the quality is poor and muddy. Many of my CD rock rips sound vastly superior at 44.1/16.

    3. I have a studio master 192/24 download from Linn (classical violin / orchestra), which sounds amazing.

    4. Whether there is any noticeable SQ difference between FLAC, AIFF, Lossless and WAV is dubious. I have read comments that claim a difference. My take is that there is little if any audible difference. As long as the rip is bit accurate and MP3 or like is avoided, I believe no one really hears a difference.

  3. #3
    Join Date
    Feb 2011
    Location
    England
    Posts
    160

    Post Nice lights and logos can make a difference to our brain (but A.S. knows that!)

    I think seeing a "96/24" or "DSD" icon light-up on your disc player/DAC can give a nice reassuring hi-rez feeling that can make the music "seem" better. Whether it IS audibly better or not, I don't know.

    I "seem" to enjoy SACDs on my new dCS player more than any other previous SACD player I've owned and usually more so than the CD versions.

    Alan showed a while back how filtered recordings of a solo violin sounded no different with several kHz cut-off the top of the frequency range. However, I continue to buy SACDs whenever I can, in preference to CDs because it justs feels like a more relaxed, freer sound with better musical flow and detail. Maybe the SACD logo has an effect on my brain's pleasure centres and my brain has ended-up associating the logo with feeling good (marketing?), leading to me preference the "sound" of SACD?

    Whether it's real or imagined, I definitely like SACD more than CD!

  4. #4
    Join Date
    May 2011
    Location
    US
    Posts
    32

    Default Testing audio formats?

    If you haven't already, (just for fun) you might try "testing" one with the other on yourself and others and see if anyone can tell the difference. Again, I would not be surprised if certain CDs sound as good or even better than certain SACDs due to the nature of the music and/or the quality of the recording.

  5. #5
    Join Date
    Feb 2011
    Location
    England
    Posts
    160

    Default Format preference

    Quote Originally Posted by QChicago View Post
    If you haven't already, (just for fun) you might try "testing" one with the other on yourself and others and see if anyone can tell the difference. Again, I would not be surprised if certain CDs sound as good or even better than certain SACDs due to the nature of the music and/or the quality of the recording.
    I have done quite a few times. Audio civilians always prefer the SACD layer, 100% of the time. And they had never heard of SACD or DSD until I told them and had no technical knowledge of electronics/loudspeakers.

  6. #6
    Join Date
    May 2009
    Location
    England
    Posts
    856

    Default Customers prefer cheaper player!

    I've done the CD v SACD comparison several times for customers, using very expensive machines like Esoteric and Ayre but with different results. On each occasion, with all types of music, the customers have preferred the ( less expensive ) CD player !!!

  7. #7
    Join Date
    Aug 2007
    Location
    Harrow, UK
    Posts
    495

    Default My challenge... is CD standard resolution good enough?

    ...that I can convert any hi-res material down to 16/44 and, under blind conditions, you will not be able to tell them apart with any kind of statistically valid consistency.

    Any takers?

  8. #8
    Join Date
    Feb 2010
    Location
    England and Cyprus
    Posts
    370

    Default Well I can't tell the difference!

    Not me.

    Take it down further to 256 kb/s AAC or OGG or WMA or even MP3 and it would still be extremely difficult to tell.

  9. #9
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,394

    Default The downgrade challenge

    Quote Originally Posted by Pluto View Post
    ...that I can convert any hi-res material down to 16/44 and, under blind conditions, you will not be able to tell them apart with any kind of statistically valid consistency.

    Any takers?
    OK, I'm up for this. Although I have my preconceptions. It would be nice if we could paste the links into your post on this thread. But that would mean all examples would have to be MP3 format. I'm not sure if the MP3 bitrate or sampling rate is fixed here on HUG for the embedded player, but I suppose 48k 320kB (or more?) may conceivably play. Thus far, all MP3s have been at 44k. Worth a try?
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  10. #10
    Join Date
    Feb 2010
    Location
    England and Cyprus
    Posts
    370

    Default Comparisons of what?

    Aren't there two issues here?

    1. That once above CD quality of 16/44.1 it would be difficult to hear any difference at higher definition (say 24/96). To test that either WAW files of losslessly compressed files (say ALAC or FLAC) would have to made available. As soon as you recode to a lossy format like MP3 you would be testing something else:

    2. That once above a certain compression bitrate (say 256 kb/s) no difference could be heard between the parent file and its digitally compressed child.

    Test 2. has been done. See

    http://www.harbeth.co.uk/usergroup/s...hlight=labarum

    or

    http://www.audiosmile.com/forum/showthread.php?t=25890

    Test 1. could be tried using sample files available here

    http://www.gimell.com/catalogue.aspx...=Studio+Master

    It is possible to hear free samples of these recordings at 16/44.1 and higher.

    {Moderator's comment: Yes. But we have no other way of presenting side by side clips here on the HUG *unless* we use the inbuilt Mp3 player.

  11. #11
    Join Date
    Oct 2009
    Location
    Australia
    Posts
    350

    Default Invitation?

    [QUOTE=Pluto;15062 Any takers?[/QUOTE]

    Perhaps someone would like to invite this (lady?) ...

    http://audioiconoclast.blogspot.com/

    the best hobbies should be left to men ;-0

  12. #12
    Join Date
    Aug 2007
    Location
    Harrow, UK
    Posts
    495

    Default The lores-hires Challenge - how I think it has to work

    I suspect that any attempt to play anything via the HUG board software will render any test invalid; it has to be files that you download and play yourself. We can work out the logistics of this after we know how many takers we are dealing with.

    One or more of you will need to provide me with some files that you believe to be a top example of the state of the hi-res art. I would suggest no higher than 24/96 as otherwise the data volume becomes huge and quite a lot of kit only functions up to 96kHz. I will convert those files to 16/44 using methods that I will not fully disclose for the time being. I will then (probably) place the converted data within a hi-res wrapper at the same sample rate as the original* and send it either to individuals or for distribution via this board, however we agree to handle the logistics.

    There has to be a degree of honesty for this test to be of use because it is always possible to examine the files forensically to determine which is which - at the simplest level, the hi-res stuff will contain programme-related information > 22kHz which a file at a sampling rate of 44.1kHz cannot. The challenge is, can you tell the hi-res material from 16/44 by listening alone.

    * The point of the wrapper is to ensure that we are testing only audible differences between the files, not your equipment's ability to play one format better than another.

  13. #13
    Join Date
    Feb 2010
    Location
    England and Cyprus
    Posts
    370

    Default Lo res at 48k?

    If the HiRes files are 24/96 wouldn't it be more appropriate to make the LoRes files 16/48 rather than 16/44.1? The downsampling is a simpler computational process, and we exclude other possible degradations to the LowRes files.

  14. #14
    Join Date
    Feb 2009
    Location
    England
    Posts
    7

    Default Sound format test

    As the author of the test linked to by Brian (Labarum) I find this an interesting proposition and would like to participate. However I agree with his comment that presenting the samples as MP3 will invalidate the test as you've immediately introduced another very significant variable.

    There are some other precautions that need to be taken. The 16/44 data needs to be presented as a 24/96 file so that are visually the same. You also need to consider that the true 24/96 file is likely to have considerable content above 20kHz. Anyone with a file editor can load them up and take a peek. So a great idea so long as there is some basic blind test housekeeping.

    I'd say make files available for download by the listener.

    Rob.

  15. #15
    Join Date
    Aug 2007
    Location
    Harrow, UK
    Posts
    495

    Default Reducing to the lowest common denominator

    Quote Originally Posted by Labarum View Post
    If the HiRes files are 24/96 wouldn't it be more appropriate to make the LoRes files 16/48 rather than 16/44.1?
    The whole point of the exercise is that we reduce the hi-res to the lowest common denominator; 16/44.

  16. #16
    Join Date
    Aug 2007
    Location
    Harrow, UK
    Posts
    495

    Default Comparisons - game rules

    Quote Originally Posted by RobHolt View Post
    You also need to consider that the true 24/96 file is likely to have considerable content above 20kHz. Anyone with a file editor can load them up and take a peek. So a great idea so long as there is some basic blind test housekeeping
    Already considered...read message #12 in this thread!

    Ultimately, there is nothing we can do to stop people cheating by examining the audio bandwidth within the test files and claiming that "file X definitely sounds worse than file Y". But remember, it's only a game!

  17. #17
    Join Date
    May 2011
    Location
    India
    Posts
    591

    Default Cables - if I can't hear a difference ....

    The last week has been spent in getting digital wireless sound into my reasonably hi fi systems - which are:

    1. Marantz SACD Pearl,Quad 99/909,Harbeth Compact 7 - main system
    2. Denon M38,Spendor S3/5 se - bedroom

    For domestic reasons the latter sits on a long wood desk in the bedroom next to my mac, not the best place for the speakers, but it works quite well for all that, with some vibrapods below the speakers, which are also placed close to the front edge of the desk.

    What has been interesting is that my ears can't hear any difference in sound quality, regardless of what I do at the source/cables, as long as that source is free from gross distortion like hum or static.

    A list of sources/cables tried, in getting to a stable wifi signal feed, including those tried for comparison reasons.

    1. Itunes streaming to Airport Express, connected to the amps via analog wire, and also via optical Toslink
    2. Itunes connected via computer headphone socket to aux in of the amp by analog wire.
    3. Ipod touch streaming via dock connector to a made for ipod USB socket on the amp.
    4. Itunes/mac connected by USB wire to a USB for computer socket on the Marantz.
    5. SACD on Marantz.
    6. LPs on Rega P5.
    7. CDs on Marantz.

    It seems that none of these influence the sound to a degree I can hear in my home. And I have now realised that what I do not hear for myself, isn't worth the trouble or the expense. Of course, the sound is hugely different for the same source in the main living room, and in the bedroom - no surprise there, different speakers, different placement, different room acoustics.

  18. #18
    Join Date
    May 2011
    Location
    US
    Posts
    32

    Default Cables only need to met basic criteria ... speakers/rooms vastly more important

    I'm not surpirsed. I think you've nailed it on the head. Cables need to be decent not exotic / expensive (i.e., they need to be undamaged, of appropriate gauge, and designed satisfactorily to get their jobs done).

    Speaker placement and room acoustics are vastly more important IMO and experience.

  19. #19
    Join Date
    May 2011
    Location
    India
    Posts
    591

    Default External DACs etc. - another inaudible improvement?

    Quote Originally Posted by QChicago View Post
    I'm not surpirsed. I think you've nailed it on the head. Cables need to be decent not exotic / expensive (i.e., they need to be undamaged, of appropriate gauge, and designed satisfactorily to get their jobs done).

    Speaker placement and room acoustics are vastly more important IMO and experience.
    And it isn't just cables. To my ears ( and to the benefit of my wallet!), I also now find that even external DACs are of no audible value. One of things I needed to do in my systems was to connect the mac to the Denon. I was busy researching - which USB DAC is best etc etc - when I thought I should first just run a 8 GBP analog wire from the mac's headphone socket to the RCA aux inputs on the Denon amp. With the mac not injecting any static or hum into the signal, I find that to my ears, it sounds just as good as a CD played in the Denon. There is a reduced volume of sound, I suspect that the current from the mac out is a little lower than usual, but the denon amp has enough head room for my needs in the room to take care of that. It is once that is done that I do not realise the difference between the two sources of course. Another feature of this connection is that the volume can be controlled at the mac as well, though this has its downsides as well.

    I believe that there are products in the market that are sold for improving this mac to amp connection, for prices ranging from 200 GBP to 10000 GBP. Exotic cables apart! I could never afford the 10000 GBP stuff, but before I became a member , and read about this subject in some detail here, I would have put down up to 400-500 GBP for this, including for the two different kinds of cables needed...so it is good to not have golden ears?!!:-)

    I also realise that there may some computers that will inject electrical noise into the signal from the headphone output socket, but these would be the exception I should think. After all, if they did that, it would be heard on headphones as well, rendering that socket pretty much useless.

  20. #20
    Join Date
    Mar 2011
    Location
    USA
    Posts
    35

    Default The mechanics of the CD - an overview and data reduction sonics

    Last evening I read an interesting passage on the operation of a CD transport. I am reading Robert Harley's "The Complete Guide to High End Audio" concurrent with "Principles of Digital Audio", Ken Pohlman's definitive text on the subject. My purpose in doing this is to gain a better fundamental understanding of digital audio playback. The fact that music's passion, drive, perspective, and expression can be captured in a stream of one's and zero's, I find absolutely fascinating.

    Prior to this study, I thought that the DAC section in a CD player was the real workhorse; responsible for the majority of processing. It turns out that this is not the case at all. The transport does a lot more than just shine a laser beam on a spinning disc! First of all, the data contained on a CD is not in the familiar SPDIF format. It is stored in what's called EFM or "Eight to Fourteen Modulation". Here, a 8 bit blocks are assigned 14 bit words along with merging bits which are responsible for the data stream. This stream contains not only the audio data, but also the clock signal, table of contents, and error correction data. Pits and Lands are embedded in the optical disc. Both of these result in a binary zero, while the transition from pit to land (or vice versa) correlates to a binary one. A device called an optical decoder divides the audio info, from the subcode and reads and interprets these bits.

    Another assumption that I would have made prior to this reading was that, much like a turntable, the data was read at the exact instant it was converted. This is not so, a certain amount of buffering has to be implemented so that error correction can work properly. This is not to suggest that disc rotation is not precisely controlled; the transport's servo allows for very high degrees of precision. The optical signal coming off the disc is very analog in nature. After all the information is extracted, de-interleaved, and decoded; the end result is the familiar SPDIF bitstream and a clock signal.

    Due to the great complexity of a CD transport, I am convinced that they would account for many of the audible differences heard between various CD players. Another interesting conclusion that is supported by Mr. Harley is that bypassing a transport altogether, and having a computer read the SPDIF data, can result in a more accurate analog output. This is a big advantage music servers have over disc players.

    Since this thread is concerned with digital playback and differences among formats, I thought I would post here. The first 16 entries of the thread seemed to be leading up to an objective comparison of audio files. Then, we got back to the ongoing cable debate. It appears that some such test were already performed with less than definitive results. Even if the testing were to proceed with the most stringent possible regulation, I am doubtful that any objective conclusion would be reached. In a very general sense, it makes sense that more information would translate to better sound quality.

    Lossy compression algorithms have shown that certain information is more musically relevant. A 128Kbs Mp3 stream has less than 1/10th the bit-rate of Redbook CD, yet it does come fairly close to representing the original. Does it sound 10 times worse than Redbook? Would a 256Kbs stream sound twice as good? Does a 24/176 file, recorded at this sample and bit rate, sound 6 times better than CD? Well, I think we've crossed over from an objective to a subjective response with this. I also think there is something very similar to Economic's Law of Diminishing Returns at work. For example, I enjoy listening to Linn Radio which is a 320Kbs stream. This is roughly a quarter of the data rate for CD, but it is considered "Hi-Res" by streaming standards. For the most part, I don't notice anything lacking in the sound. On the rare occasion that they play something that I own, I will listen to it and then play back my file for comparison. This is not a "blind" test and it only accounts for 2/3 of the ABX process. Lets just say that I have to listen very hard for differences. If they exist at all, its not something that I would be willing to bet on being able to detect. On the other hand, does this mean that I should go ahead and compress my entire Library to save on the space occupied by the bits that I just admitted were of little significance?

    Well, maybe, but I'm not going to be doing it, and I wouldn't even if storage were at a premium.

Page 1 of 4 1234 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •