1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

iPhone AAC vs. Aptx and Aptx-hd real world

Discussion in 'Sound Science' started by neil74, Oct 4, 2017.
5 6 7 8 9 10 11 12 13 14
16 17 18 19 20 21
  1. Monstieur
    aptX HD just uses more bits per sub-band than aptX. It's still esentially transmitting ADPCM which is grossly inefficient compared to psychoacoustic codecs.

    All aptX implementations I've heard have audible frequency response distortion - this should not happen if it was just bit-depth reduction, so it's most likely an implementation flaw. The AAC implementation on the same Bluetooth devices don't have this problem.
  2. Monstieur
    There is nothing wrong with the technique itself - it's the low bit-depth used in SBC over Bluetooth that makes it audibly degrade the sound, especially when used without psychoacoustics.

    If you simply reduced the bit-depth by 1 bit, it would still technically be SBC and probably be audibly transparent.
    Last edited: Jul 24, 2018
  3. bigshot
    I'm guessing that you know that you aren't arguing on point and you are afraid if you answered, you wouldn't come out on top. That's the wrong way to approach a forum like this. Talking with people in sound science isn't about winning or losing. It's about sharing information. I have information you might not have, and I invited you to share information that might show me a different side. The fact that you can't focus your argument and discuss things straightforwardly is kind of sad for you. Oh well.
    Last edited: Jul 24, 2018
  4. Steve999
    I have a question--would Apple Airplay and Apple Music be the best way for me to stream music?

    If I am using Bluetooth, is there any way that I as a person who does not understand that technology can do as a rule of thumb so I get a high fidelity stream? Kind of an HWmonitor for your bluetooth stream or something to see what's going on? Wifi and over the network hard-wired I understand decently.

    On my home stereo I usually stream Apple Music over Apple TV. Good? Bad? Is there a way to optimize it? If they can match what I have on my computer they will just use the Apple 256 AAC VBR over my network sent straight from Apple, if I understand correctly. If Apple does not have a match they have uploaded my file from my computer and stream that.

    If I use Spotify over a Roku through my home stereo is that worse?

    If I use Spotify on my home computer to my Airplay + Bluetooth speaker is that better than using Apple Music in any way?.(My family of 5 shares a spotify account.)

    I'd like to be confident that my streaming is giving me something transparent or very close to it. Honestly I am just sort of asking what button to press (figuratively)--I don't have the wherewithal to figure it out. I am guessing sticking to Apple and its ecosystem would give me the best results but would be very grateful for any knowledgeable contrary opinions.. With so many variables I cannot imagine figuring it out for myself. I get a little concerned when I read about all of this transcoding and different codecs and aptx etc. I just want to do the simple thing that gives me the best stream.

    I know that's a lot of questions but I am really just looking for a simple answer based on knowledge and quality of the technologies if that's at all possible. If it requires a little study time or experimentation or configuration or a little extra effort i can do that to. I want to get to set it and forget it and just know I got the best or close to the best stream I could.

    I must often stream from my computer to a Bluetooth speaker or from my Apple TV hard-wired to my nice stereo. My bluetooth speaker can do line in, ethernet in, Airplay, or Bluettooth or Bluetooth Aptx.

    Thanks anyone!

    Last edited: Jul 24, 2018
  5. bigshot
    I've found that bluetooth isn't perfectly transparent for the most serious forms of listening. It's fine for casual listening and most of the time, it isn't a problem. But AAC 256 VBR, which is the standard in the Apple Store is perfect for human ears. I use Airplay to stream around my house from my Mac Mini media server to an Airport connected to each stereo. It works great. The jitter rating of Airports is a little higher than normal, but it's still well below the threshold of audibility. I'd recommend a system like the one I use. The only problem I get is some skipping when someone in the house is hogging the wifi bandwidth with a chunky download. That doesn't happen too often.
    Last edited: Jul 24, 2018
  6. PiSkyHiFi
    We've already addressed this. You mean sounds the same on the same equipment when you say audibly transparent, I don't think I would pass an ABX test with 256 Kbps AAC, even on my best equipment - I'm not even interested in that because I just explained why it's not the full story, you didn't pick that up because you're focused on getting me in a box where you can't be challenged.

    I challenged your logic on dynamic range and you ignored it - it's impossible for AAC to have as large a dynamic range as CD - that's just basic logic

    I said AAC is a fantastic lossy codec, it's one of the best, go back and check.

    I don't think it's reasonable to say that no one could ever tell the difference between AAC 256 and CD, yet you seem to pointing at that as if it's fact, like there is no evolution of sound systems at all.

    Honestly, it seems absurd to me that while I'm trying to make the best out of the sound equipment I have, I would deliberately reduce the quality of the source material because some research said I couldn't possibly hear the difference.

    The only reason I used AAC in the past is due to limitations on storage at the time.

    If I'm going to audition different sound equipment, do you think it's sound science that I would reduce the digital representation to barely adequate deliberately? I would be just introducing new ways to prevent revealing flaws in my equipment.

    Transparent means the same as without a sound system.

    Because transparent means more than comparing just 2 sound system scenarios, it's about being able to close your eyes and see if your senses can be tricked by a sound system into feeling like it isn't there.

    I don't know about you, but I've never encountered a sound system that actually did that fully - partly it's because that standard of transparent requires fooling subconscious mechanisms that are extremely good at piecing together a model of what is happening around our ears, so even if we can't verbalise what's happening and distinguish differences consciously, we can certainly tell that it's not real and that's sufficient to show that ABX is simply not enough to actually trick us.

    Don't rely on ABX, rely on careful, considered listening many many times over and possibly never reaching a definite conclusion, because you can already tell it's a sound system and not real.

    I don't think you're ever going to get that, your own definition of audibly transparent won't allow it.

    In conclusion, I don't store files in AAC any more because I don't need to, it would be absurd to reduce the representation in any way unless it was the only way things could work.

    Why not mention something about audibly transparent in case I missed it.
  7. bigshot
    Since I’m on my phone, I’ll take it one point at a time.

    AAC and dynamic range

    AAC compression is not dynamic compression. AAC files have the same dynamic range as CDs.


    People who can tell the difference

    I cited two controlled tests above that showed that neither age, trained ears nor high quality equipment were a factor in being able to tell the difference between AAC and lossless. If there are people who can discern a difference consistently, we haven’t found them yet.

    Transparent means it sounds real

    You are misusing the term then. Transparent means that the sound going in sounds identical to the sound coming out. A transparent amp is described as “a wire with gain”. Reality is a function of recording techniques, not file format. The reason your system has never sounded real to you is because reality isn’t the goal in commercial music recording. The goal is to organize the sound for clarity and creative contrasts. Neither AAC nor lossless nor HD audio present reality. That isn’t a limitation of AAC specifically.


    You may not need to store AAC files because you have a small library and you never stream or listen to music wireless. That is fine. But that doesn’t mean that your lossless library has better audible fidelity than the same music encoded in high bitrate AAC. I have a large library with over a year and a half of music. It all fits on a single hard drive, making it simpler to do scheduled backups. I also have a media server that streams music and movies to every room in my house. I could have shelves and shelves full of CDs in my listening room, but I don’t have to put up with that clutter. All my CDs are ripped and the discs are in boxes in the garage. Much more convenient. The same sound quality.
    Last edited: Jul 24, 2018
  8. PiSkyHiFi
    AAC is not dynamic compression, like there are people who thought otherwise.

    I've been talking about data rates and codecs, there is zero evidence to suggest I think of AAC as an analog audio compressor.

    That article concludes quote "The frequency response and dynamic range are essentially unchanged, there's just more junk in the signal"

    So.... what about dithered mastering then... that's the counter example I gave you to prove that dynamic range can be lost.

    If it helps you at all, I'm sure the person that wrote that article would agree with it - a source that is mastered down to redbook can achieve slightly better than 96 dB dynamic range with careful use of dithering, it must be bit-perfect or be processed by higher precision math to maintain any of the perceived range above 96 dB - encoding to any 16 bit based lossy format will destroy any dithering that was present in the mastering.

    I don't think it requires a genius to get that, since the data is reconstructed into 16 bit and will contain random errors compared to the original dithered master, encoded error diffusion will be ignored - perhaps, if one is careful, attempts to reconstruct by using higher precision math in the compression encoder and then dithering down again might work, but it's guaranteed to not be as high dynamic range as the original source potentially could be, simply by being different.

    Dynamic range is lost, or should I say potential dynamic range is lost as it will depend upon the mastering.

    That's just one example of how dynamic range is lost through lossy encoding... this is science.
  9. bigshot
    Perhaps I'm having trouble sorting out your points. That's why I keep trying to pare the conversation back to one thing at a time. Could you please explain what you meant by this quote?
    I am confused because you then say this...
    I'll move on to other points when I figure out what you're getting at with this one. I originally pointed out that AAC doesn't affect dynamics earlier. You replied saying I ignored your challenge. So I cited an article that shows that AAC compression does not alter the dynamic range of music. Are you agreeing that it *is* possible for AAC to have as large a dynamic range as CDs now?

    Dithering shouldn't be an issue. The same dynamics are the same dynamics. Lossy codecs are designed to throw out sound you can't hear and leave the sound you can hear. Most commercial music doesn't exceed 50 or 55dB in dynamic range anyway. CDs and high bit rate lossy are overkill when it comes to dynamic range. I only would use dithering when I downsample. When I rip a CD, it should carry across with the same audible sound of the CD, wouldn't it?

    This isn't an argumentative trick. I'm honestly trying to discuss this with you and I can't figure out what you are saying. I keep talking about audibility, and you keep going back to talking in theory about things that aren't audible.

    All of us use our stereos for the same thing... listening to music with human ears. Theory is great as far as it affects that particular purpose. But theory for the sake of theory that doesn't affect audibility doesn't matter for that purpose. You can say that you feel better about having larger file sizes or that you worry about smaller file sizes. I'll accept that as a psychological block you have to using lossy. But when it comes to audible sound or the purpose of listening to recorded music in the home, I can't see any advantage to lossless. It's bigger, less convenient, less cross platform, it doesn't stream and it sounds no better.
    Last edited: Jul 24, 2018
  10. PiSkyHiFi
    The article itself was mainly pointing out that compression in audio could mean dynamic range compression (analog audio compressor) or it could mean information compression like lossy encoding and just to keep in mind that these are 2 completely different concepts.

    No, I am correcting that article or extending it.... AAC cannot mathematically achieve the same potential dynamic range, I've just shown it logically based upon dithered mastering.

    When I say potential dynamic range, I'm referring to the range of precision - how closely it can represent the original signal in all it's dynamics and how much error is associated with it. AAC will be less for the reasons I outlined above - I don't think there is any argument about that.

    Audibly transparent is listening to a live performance, not whether a track sounds the same on a particular sound system to it's compressed doppelganger.
  11. PiSkyHiFi
    It certainly does stream, just not with Apple. For example, Google Chromecast Audio can stream lossless, although not always reliably - nothing to do with the wifi I think, it's just the software needs better buffering management and I haven't really explored this much yet.

    Honestly, I cannot see any reason to revert to lossy for storage - lossless is just as convenient, it's even more cross platform and it might sound better, especially after listening many times over and not trying to compare what we can verbalise through ABX testing.

    I have AptX HD for streaming at home and in the car, home streaming works better on the App side if it's Bluetooth for me - but when lossless is ubiquitous across hardware, I'll shift to that.

    We are arguing over where to draw the line... and honestly I think we're actually pretty close, I have used both MP3 and AAC before for storage, it's just deprecated now.
  12. Monstieur
    AirPlay is lossless, so your source will be reproduced exactly.
    If you use the native music app to play from Apple Music / iCloud Music Library on the Apple TV, then you get 256 kb/s AAC.
    If you use Bluetooth, AAC is the best codec since it defaults to 320 kb/s, at least on macOS.
    Last edited: Jul 24, 2018
  13. bigshot
    Streaming lossless isn't practical for the majority of people who listen to music. Compressed audio sounds the same and works pretty much flawlessly on any computer or mobile device.

    There is no reason to revert to lossy for storage if you have already ripped to lossless. Transcoding a large library would be time consuming and would just result in a bigger footprint on your hard drive with two libraries.

    Lossless is definitely *not* more cross platform. Your computer does FLAC easily. Mine does ALAC easily. We can jury rig our computers to play the other format, but it isn't easy. However both of our computers play compressed audio easily.

    Direct switched, line level matched ABX testing is the best way to determine whether two similar sounds are perceptually different or not. Long term testing is subject to problems with auditory memory. Humans are unable to accurately determine differences between similar sounds after as little as a second or two. Our ears adjust, our echoic memory fades and the discernment is reduced to pretty much random chance.
  14. PiSkyHiFi
    This probably comes down whether sound is something that happens without being heard or is it entirely subjective. If a tree falls in the forest.

    Sound equipment has many issues - as you pointed out, a sound system isn't necessarily trying to be transparent, actually, I think that is the aim and we are just used to justifying what we have already as if we couldn't possibly do any better.... transparency is a lofty goal.

    Lossless isn't more cross platform - well, I disagree, FLAC is open source too and usually works on DAPs even when they haven't included AAC. The Xduoo X2 is an example from a while back. The main thing though is lossless can convert losslessly between formats, so using either ALAC or FLAC is not an issue.

    i am adamant that ABX testing is not even half of a decent evaluation of sound quality, time to absorb the details is essential and forming conclusions maybe not even be possible, you may remain uncertain but open like myself.

    Quote "Our ears adjust..."
    Surely that's even more reason to focus on the experience rather than the analysis and leave room for different experiences of the same reproduction, I would like to leave room. Sound equipment can achieve frequencies we can't even hear and yet it can't achieve transparency of imaging and quality, that tells me that digital sound storage needs to have a fair amount of headroom to allow for different experiences upon repeated listening.

    For me, sound burns into my brain, I can recall sound very well and I know this because I'm also a musician that plays by ear.
  15. bigshot
    A sound system definitely is designed to be transparent. That's the goal of music codecs too. We're really fortunate today because we can achieve audible transparency with just about any amp or DAC you can buy. I remember when I first started in this hobby over 40 years ago. Amps were rarely transparent. LPs weren't transparent. Often, even reel to reel tapes weren't. I test every component I buy for transparency, and I haven't found anything that wasn't since the early 1980s. I bought a $40 DVD player from Walmart and it was transparent too. Aside from transducers, transparency is a given now.

    What I said before is, recordings aren't made to be realistic. They are designed to sound *better* than real. More balanced, more organized, more variations and contrasts. You can't compare a live performance to a recording because they are two completely different things. But you can compare the sound of a recording contained in a WAV file and compare it to the sound of the same recording in a compressed format. If the sound of the source is the same as the sound at the other end of the chain, it's transparent. With electronics, just about everything is audibly transparent.

    FLAC isn't supported natively on Macs. You have to download a third party player to do that. iPhones don't natively support FLAC either. ALAC isn't always easy to play on PCs. There is still a format war when it comes to lossless. But there is no format war for compressed audio. Just about anything can play a 320 LAME MP3. And that is an audibly transparent codec/data rate.

    Well, you better be glad that the people who test medicine don't feel that way. Controlled testing is the backbone of the scientific process. You can feel free not to believe in it, but your subjective impressions are infinitely more subject to error than controlled tests. You should probably do a little googling on human perception and the effect of bias. Bias is real. All humans make decisions every day that are colored by bias. The best way to eliminate bias and come up with objective answers to our questions is through controlled testing. If you reject that, you're in the wrong forum right now, because this is sound science. The scientific method is important here. Subjective impressions based on bias are fine in the rest of Head-Fi. In fact, a lot of high end audio salesmen prefer you avoid controlled testing and measurement. Subjective impressions are really good for the profit margin!

    By the way, ears take time to adjust. So the longer the sample you listen to, the less likely you are to discern a real difference. Direct A/B switching is the best, because you can put one sound right next to another sound and hear smaller differences clearly.
    Last edited: Jul 25, 2018
5 6 7 8 9 10 11 12 13 14
16 17 18 19 20 21

Share This Page