44 -24bitkHz vs 44.1kHz-16bit

What sonic differences should I listen for between Flac 44.1kHz 24 bit and Flac 44.1kHz 16 bit (I am streaming Tidal and Qobuz)
Chas

To my ears, higher bit rate adds space (especially depth). Higher sample rate tends to sound a bit more relaxed and “at ease”.

4 Likes

I’d be interested to hear what folks say about this, I haven’t heard anything in 24/44 that I said Yep that sounds better than 16/44 but I can’t say I’ve made any A-B comparisons. It isn’t anything like 24/88,96,176,192 etc. where one can say yep that’s hi res. at least a fair portion of the time.

I have see a number of reviews in which people report a more pronounced improvement with greater bit depth than higher sampling resolution.

3 Likes

Listen to reverb tails, micro-dynamics, and smoothness of the sound.

1 Like

I’m not sure I’d generalize the differences that way. Definitely 16 to 24 bits is a big step in micro-dynamics, etc. as IanB52 notes. Higher sample rates affect the solidity of the sound stage more IMO. Putting the other way lowering the sample rate seems to cause the soundstage to get a little more ghostly or fuzzy - same width, same height, same depth, but less precisely positioned.

5 Likes

I too have found better depth with 24-bit versus 16-bit. Not that I make a habit of it but I have bought 16-bit downloads only to later discover that a 24-bit version is available and bought it. To my ears the 24-bit versions sound better. However, maybe that’s because the engineer took advantage of the greater available dynamic range perhaps?

The files will have started out as the same recording. 96/24 is common for recording. After editing and mastering, the final file is then down converted to distribution formats if lower resolutions are desired.

Thus, a 44.1/16 file is the same as the 44/1/24 file after eight bits are truncated and dithered. That is, the recording engineer did not do anything different to record the two file resolutions.

The biggest advantage in recording in 24 bits is not the resolution (you can never use all of the bits from top to bottom) but the extra headroom. You can set your recording levels so that peaks are at -12 dBFS knowing the two bits you threw away will never matter. This way you never overload the recording equipment with too hot a signal while at the same time knowing you have not thrown away any information.

It is quite a luxury and comes in handy. For example, I long ago learned a typical symphony orchestra will play the biggest fortissimo of a piece louder in concert than at dress rehearsal. But by setting the levels at -12dBFS at dress rehearsal I have plenty of headroom for the concert and will not have an over.

With tape, where you cannot afford to throw away any signal, you have to guess so that the peaks come as close as possible to 0dBFS without going over. It can be tense.

4 Likes

Elk is right that almost every recording or remaster since the year 2000 started out as 24 bit. The CD quality version is always downconverted from the original. I actually don’t even understand why 16 bit media is still a thing considering that it all starts out as 24 bit, the files aren’t that much bigger if the sample rate is the same, every DAC plays it, and few people buy CDs. But I digress…

Many people describe 24 bit in terms of greater headroom for recording and digital processing, but that isn’t the primary thing that makes it sound different. You can record at the exact same level, and 16 and 24 bit will sound different. In the same way, when you take a 24 bit recording, add dither, and convert down to 16, you do lose something even though some will insist that dither magically preserves all the sound quality. My theory is that the very subtle and low level differences in sound contribute a lot to the overall quality, even where it might not seem to matter with the dynamics of the music.

Likewise, there are a lot of recording engineers that record 24 bit extremely hot. A couple sessions I was assisting years ago, the engineer had every track literally clipping (aka 0db) every few seconds. The overs were short enough that they weren’t highly audible, but it goes to show that using extra headroom with 24 bit is not necessarily common practice. Also, many mastering engineers will intentionally push the volume to cause the converters to clip instead of using a digital limiter to maximize loudness because they like the sound better. Take a look at any master/re-master done by Ted Jensen at Sterling sound, they hit 0db constantly.

A funny thing with tape, is that a lot of the time vocals and instruments are recorded at between +1 and +4 in order to fatten up the sound. With tape you can actually go over 0db and not have horrible clipping sounds, but soft saturation from the tape and electronics. Stuff like cymbals or acoustic guitar, they might keep below 0db to preserve the treble, but vocals, drums, bass, electric guitars, etc are often recorded above zero. This is possible because 0db in analog is actually -18db in digital, so there is a huge amount of headroom transferring tape to digital. I once saw a punk album recorded at +9db VU analog, something I would never do because so much of the high frequencies are lost.

3 Likes

Gentlemen, thank you for your information; great discussion. I would love to hear more.
Chas

I agree, you do lose something when going from 24 bit to 16 bit. Dither masks the noise/quantization errors but does not magically restore the sound. I also agree files should be offered in 44.1/24 for downloading.

I, too, have seen many push digital levels as if they were working with tape. I understand it is hard not to do when you have experience working with analog and move to digital. I do not worry about losing a bit or two to headroom as the lower bits of a 24 bit signal are lost in noise in any event.

3 Likes

Much of the Bandcamp stuff I download is 24 bit :slight_smile:

2 Likes

Good to know about Bandcamp. I have purchased many singles and a few albums from them.
Chas

1 Like

Elk,
I greatly appreciate your insights.
Chas
PS thanks to @Tedsmith and @IanB52 also for their information.

2 Likes

This is the sort of comparison that really needs blind, or double blind testing. I’d be shocked if any human could do better than a monkey flipping a coin to differentiate reliably about the difference between 16/44 and 24/44. Theory is one one thing, a genuine test on the street is something else. It may require someone with particularly golden and talented ears to consistently discern the difference. But the only way to know, is to conduct genuine tests, not hobbyist switching back forth between tracks.

1 Like

You’re probably right there, as with most audiophile decisions relating to technical principle rather than actual enjoyment factor, but on principle I like getting a 24 bit not a 16 bit when I download a track (of course I do wonder sometimes if the 24 bit is just the standard 16 bit distribution file with an extra 8 bits added :wink: )

1 Like

I’ve been of different minds on this topic. For many years I was in the higher resolution DSD to the moon camp. But I began to wonder how much of it was based upon actual listening, and how much of it was based upon exposure to such high res belief systems on internet forums!? I began to wonder how much my enthusiastic endorsement of all things higher resolution as better was based on a mindset I had adopted by reading too much. I conducted a non scientific not blind test of my best recordings, be they redbook, high res PCM, or DSD. In this informal survey I concluded that I enjoyed them all equally. A well recorded piece sounds great no matter the resolution. That led me to sell off my entire high res rig in favor of a more forgiving DAC.

I’m not declaring any one route the winner. I think it really comes down to personal preference. But I can’t help but be skeptical, even within myself, of saying the 24/358 sounds better than 24/192, even though in my high res days I’d without question pick 24/358. But I wonder if that enthusiasm was mostly because 358 is a higher number than 192 so it must be better :joy::joy::joy:

1 Like

There is probably a lot of that, but I think sometimes it’s cool to do so purely on the principle of having “a better scientific instrument”, although this assumes that they sound equally nice to my tired old ears. Hence I am happier downloading 24 bit rather than 16 bit just on principle (and I know they were likely originally recorded at 24 bit anyway). I just don’t claim (again to these tired old ears) one is better sounding than the other. I wouldn’t like to say that others can’t tell the difference, even in blind tests, I have met some folks with astoundingly astute hearing over the years, who plainly could tell the difference (I used to try and catch them out when audio engineering by sneakily switching i.e. blind testing, and a few of them could usually tell, at least at the start of the day, which was which.
Whatever makes me happy is fine :slight_smile:

2 Likes

But, if the 44.1/16 is a CD, quite often the CD was mastered by a different person than non-CD releases. Or at least it used to be that way. I haven’t kept up with stuff like that for a few years, so that may no longer hold true for more recent releases.

Is that still true when recording at 30 IPS?

Remember, the faster the tape is moving across the head, the lower the noise floor will be, the more dynamic range capability the tape can handle, and possibly slightly higher recording level. I would think that would also help retain most if not all treble extension.