Hi-Res Audio: Don't waste your money!

Music labels and hi-fi companies alike are pushing the virtues of high resolution audio, but is it any better than CD audio?

In this video, we break out some science to get behind the hype, and hopefully help you decide whether you need to spend loads of money buying all your music again in a hi-res format!

TRIGGER WARNING: IF YOU ARE A Hi-RES AUDIO FAN, YOU MAY NOT LIKE THIS VIDEO - YOU HAVE BEEN WARNED! (but keep an open mind and you may be surprised.....)

=======================
Thanks for Watching!

Please don't forget to subscribe to the channel - let's build a great community of audio nerds who don't take themselves too seriously!

Music:
Creative Commons — Attribution 3.0 Unported — CC BY 3.0

Follow me on twitter: @AudioFixated - this is the best way to contact me!
========================
#AudioFixation #HiRes

[robertlakay]

People don't realise how high quality music the CD format can offer. Most times when CDs sound bad it's the mastering and not the format.

[mxbishop]

I've spent most of my life recording and listening to music. I started out recording from LP to cassette tape, then from CDs to MDs. Then from CDs to MP3s. Then downloading CD-quality files. And then, hi-res downloads. And then, streaming hi-res music. I've searched high and low for the best recordings, and the best way to hear them. I've listened to all the formats, on all kinds of equipment, on all kinds of speakers, in all kinds of environments. From portable players, to very nice car stereos, to very high-end gear in controlled listening rooms. My conclusion: The mastering/mixing is everything. Sure, a poor quality cassette tape, in a poor quality cassette player - is going to sound terrible. But a poorly mastered/mixed song is going to sound terrible - no matter what format, sampling rate, or bit depth, one is using to store it. To better illustrate what I'm talking about, consider this: I have a new CD by Melody Gardot called "Sunset In The Blue". Just a plain-old Redbook CD, nothing special about the audio format: 2-channel stereo, 44.1 KHz sample rate, 16 bit word depth. And it sounds freaking amazing. (Sony CDP-X77ES player, McIntosh MA6900, rendered on the original B&W 703 speakers, with premium interconnects and speaker cables.) It's like Ms. Gardot is in the room, singing to me, personally. The music is alive. I hear no noise, no artifacts, nothing thin, weak, muffled, or fuzzy. Just great sound. That is what the standard 44.1 16 bit format is capable of - with an expert recording, that is mixed properly. Having gone down the high-end, hi-res road for decades, and discovering what I know now, I would say most people will be better rewarded by searching for those tracks that are expertly mastered/mixed - and spending less time trying to find the hi-res holy grail. And perhaps, with the money saved by not purchasing hi-res files, more funds are available to purchase additional music, and/or to make equipment improvements. This has been my experience.

[BleakVision]

CD- Audio is one of the most future-proof technologies in history. The engineers did everything right, and then some.

[CFL]

Wow. This video has finally helped me find peace in my soul lol. I've been collecting CDs for a long time but I always thought i needed to spend some serious money in upgrading to Hi-Res if I really wanted to get the best audio possible. Now I realize the collection have been spending so much time and dedication putting together is more than enough for me and instead of changing to the high-res format I need to spend that money on equipment that will unlock all of the potential and squeeze every drop of quality from those CDs. Thank you Audio Fixation!

[ryugatsuchiya]

Simple and easy explanation, saved me a lot of head scratching and reading useless articles in the internet.
Earned my sub.

[skesinis]

Great video and very thorough! I have only one addition regarding the higher sampling rate recordings:
One reason about why the 44.1kHz sampling rate was considered “low” and why the first CD players that were ever created sounded a bit “metallic” is this: While the highest frequency that can be sampled is 22.1kHz, the whole spectrum of frequencies that can “pass” from the same sampled points is mirrored above that frequency. Even though these are clearly higher frequencies and we normally can’t hear them, their differences between the actual captured ones below the 22.1kHz and those mirrored above that fall into the hearing spectrum and these are the actual ones which can be heard. So there is a need for a filter to block anything above 22.1kHz in order to counter that effect. The first CD players had 6th to 8th degree filters (1st degree = 3db/octave 2nd degree = 6db/octave etc) because their filtering curve needed to be flat from 20Hz to 20kHz and then drop dramatically between 20kHz and 22.1kHz. Those filters have a dramatic effect in phase distortion which is something that you can clearly hear as a metallic sound in the high frequencies since this is where our hearing is more sensitive in phase differences for detecting sound direction. Later on, the oversampling was introduced, by interpolating samples using math in-between the actual recorded samples, resulting in a much higher “spectrum mirroring” frequency which now needed a simple 1st degree filter to have a flat response between 20Hz-20kHz.
The actual question now is whether you can possibly hear with your equipment in a blind test, the mathematically interpolated samples of a 44.1kHz or 48kHz recording created by a DSP and the actually recorded samples of a higher sampling rate recording at 96kHz-384kHz or even 768kHz, and tell any difference in a blind test.
As you mentioned in your video, anyone most likely can’t.

[navarrmh]

"and if you do, you are Batman" hahahaha

[briannacluck]

Honestly appreciated this video. Now I don't feel guilty that some of my albums are "only" 16 bit FLAC files, and will stress myself out less trying to find places with 24 bit audio

[evtyler]

This is a topic I've struggled for years to understand and now I finally get it. Thank you so much for your brilliant and simple explanation!!!

[outisaudio]

Really loved hearing about all this stuff again. I took a course on Nyquist-Shannon, sampling, etc. in college and you did a great job summarizing it quickly (from what I remember). The major takeaway is that the people recording your music have already thought all this stuff through, way back in the eighties when this stuff was being invented by people with PhDs. We've understood the science of sound for quite a while, and while audio equipment has gotten way better for way cheaper, CD-quality audio is still just fine.

[tabloidannouncer]

Every time I get the urge to purchase hi-res audio equipment, I come back and rewatch this video to remind me of the limits of human hearing. Thank you for saving me a bunch of money.

[klaushaunstrupchristensen]

When CD Washington introduced in the eighties it was marketed AND engineered as perfect sound. And I think they got it about right. The 16 bit 44 kHz dithered will cover the requirements for perfect transmission of a Mahler symphony without compression. I think rather than focusing on lossless transfer of ever higher resolution the industry should upgrade the resolution on wireless transmission (like Bluetooth).
Nowadays when data storage and transmission has become it cheaper than it was, it’s interesting to speculate on what would the data format have been if it was introduced now. For absolute perfection my pick would be 18 bit 48 kHz. 18 bit will give a dynamic range and noise floor of 108 dB before dithering and 48 kHz will give a bit mor space for high frequency filtering. Not that I believe this format would improve the sound, but it would be technically sweet and keep some reserve .
Thanks for sharing your thoughts 😃

[lighteningwawa]

A lot people have the misconception that digital music are reconstructed in a zig-zag fashion. Meaning instead a smooth wave, the reconstructed wave has shapes like staircases. And that is completely not true at all! According to Nyquist-Shannon theorem, sampling at double the highest frequency of the input would contain enough information to reconstruct the smooth original signal. What's needed after the "staircase signal" is simply a low-pass filter.

[JapanSun]

Completely agree with the assement, a good lossless 16bit Flac is all you need. I like browsing for used CDs and digitizing them in FLAC. I did an A -B-X Test once with 32bit vs 16bit (Sennheiser HD600 and ifi DAC Zen), i couldn't tell the difference.

[hokiemonproductions]

That cleared up some of the technical aspects of Dankpods’ flac player reviews. Appreciate the physics/tech lesson!

[timharbert]

CDs cheap. Encourage folks to stay away from CD.

[dudexyt]

When MP3 became popular in the mid-late 90's, I was able to hear a difference between a high quality MP3 encoding of a CD track and the CD itself, using your standard consumer (Sony) headphones. And today, after getting some decent midrange hifi gear, I can hear a difference between, say an Apple mastered AAC and a CD recording of the same track, but it takes really dedicated listening and lots of repeat listening to hear the subtle differences.
This is good news though, because it means your large playlists on Spotify, Apple Music (or preferred streaming service/download service), or your large compressed library of CD rips are enjoyable as is and you're not going to miss much unless you have the equipment to reveal the differences. And even then those differences are subtle, unless you spend the money where those differences can be magnified. This also means games can still be enjoyed even if they might compress sounds and music a bit more for space/memory savings.

What does matter, though is how it's mastered. Sometimes a hi-res track might use new mastering that makes it sound better than the original CD recording. But say that a streaming service offers different mastered versions of the same song, that can have an appreciable difference in how the music sounds and how enjoyable it is.
One example is Pearl Jam's debut album Ten. I've never really enjoyed the overall sound of Ten because of the mastering, and it took a lot of listening to really appreciate the music (and band) because it didn't have the "kick" or "vibe" of the "grunge" era artists of the period, but when they issued a new mix of it 10 years later, I enjoyed it more. In the remastered version of "Ten" you can hear Eddie Vedder's voice clearly when he mumbles or speaks softly during some parts of the song, and it has more of a hard rock feeling and vibe. It totally fits the sound of the period and it's more enjoyable.

[carljung]

a lot of audiophiles are into numbers. actually listening to the music is secondary. and they do not like blind testing! my god if you want to make some heads pop show how $2000 cables are only for the well off and gullible...

[juanjogarcia]

Respectfully disagree. A lot of simplification here. First. when you talk about perception limits you consider only frequency domain, not time domain. Second. All Digital Analog Converters have filters for high frequencies caused by analog recreation from sampling, it is known that when you apply a filter you get less digital aliasing as high is your sample rates. Old low sample rate synths for example, produced aliasing in high frequency even SR being double or more of perception limit. This aliasing effect when filtering is clearly detected by most people in blind test. Modern digital synths use oversampling to unrequired levels by our limits of perception for avoiding perceptible aliasing when filtering in high frequencies reproduction. Three. When you consider bit depth you say music never goes from 0 to 100. First, Decibels are not like grams, it is not absolute measure, it is relative to a level. In case of DAC or preamp equipment we are talking db relative to line level, but then there is your amplifier. Decibels here is not absolute value, it is relative to amplification level (Pre amp, line level db). 24 bit means the different amplitudes the sample may have 24^2 and it corelates completely with the concept of HDR in video (levels of light from white to black). With all the respect, i disagree, digital 44.1-16 can be cold, metallic, in certain types of music. And there is a lot of difference from some DACs to others regarding how to manage aliasing in high frequency digital oscillation. Only complete argument (not oversimplifies) its blind test one. But if i can hear aliasing in a digital synth oscillator compared to the analogue counterpart, i can hear aliasing made by filters of my DAC, in certain high frequencies. Some DACs oversample internally 44.1-16, being the source at this read, to avoid filter aliasing. Also there is an important argument when you are resampling an analogue tape (Especially in Jazz and Classical Music) you capture a thing that when converted to analogue the line level after DAC is much more close to the output to line level of the original historical recording tape. And as you know, all tapes are going to die eventually. Including Miles Davies tapes ;-)

[goodsound]

It’s a common mistake that the big advantage of Hi-Res audio is the higher frequencies but it’s a better time resolution and less problematic digital reconstruction filters (noise shaping), so you don’t need to be Batman to profit from real Hi-Res.
And actually we had a similar discussion when moving from standard TV to HDTV. People claimed that they won’t see the better resolution or the more colors. And now the same again with UHD TV. Just because you can’t see or hear the advantage of Hi-Res doesn’t mean others cannot as well. It’s individual, not absolute.