24bit vs 16bit, the myth exploded!

[Ryokan]

the actual answer of a 100 dB trumpet plus 20 violins each at 80 dB is 100.69 dBSPL!

Makes you realise how loud aircraft are at take off, 140dBspl. Much much louder than the trumpet.

 

[71 dB]

Yes, I was trying to keep it simple so it was easy to conceptualise. The same is also true with the mantissa of a float format.

Okay. I wasn't sure how familiar you are with the subject and I wanted to make sure the "truth" is told. I agree with you about Two's complement making things harder to conceptualise, but what can you do? Demonstrating people how some things are more complicated than they assume is one way to make then realise they don't know as much as they assume.

The simple way to make this easier for someone is to just tell them half of the binary values are for the positive signal values and the other half if for negative values and then go from 16 bit to 15 bit: 111111111111111 is the max possible signal value (sign bit omitted).

4. As the decibel scale is logarithmic, we can’t just add different decibel values together. This makes dB calculations very unintuitive and misunderstood by most audiophiles. Let’s take 2 examples:

Example A:
Let’s say we have a trumpet playing loudly, which measures 100dBSPL (at say 5m away). Let’s say we then add 20 violins playing loudly, each of which individually is producing 80dBSPL (also at say 5m away). What’s the dBSPL level now? Obviously it’s not 100 + (20 x 80) but maybe we can use our knowledge of factors/ratios to figure it out? 80dB is 20dB less than 100dB and therefore 10 times less, so 10 violins would be the same (100dB) level and 20 violins would be 10 times (20dB) more, so our answer is a bit above 120 dBSPL. Does that sound reasonable/logical? That’s still completely wrong, the actual answer of a 100 dB trumpet plus 20 violins each at 80 dB is 100.69 dBSPL!

Interestingly I got 100.79 dBSPL, but the 0.1 dB difference is quite insignificant. In practice it's just 101 dBSPL.

 

[71 dB]

Makes you realise how loud aircraft are at take off, 140dBspl. Much much louder than the trumpet.

Yes, you'd need 10.000 (!) trumpets at 100 dBSPL each (or a million violins at 80 dBSPL each!) to reach 140 dBSPL. That's why our hearing is logarithmic. On linear scale the differences between quiet and loud sounds are just too massive.

 

[gregorio]

Makes you realise how loud aircraft are at take off, 140dBspl. Much much louder than the trumpet.

It all depends on the distance you measure it from, SPL decreases significantly with distance. A trumpet can also reach levels of 140dBSPL but only if played extremely loudly and measured from just a couple of inches away. At a more reasonable 20m, it’s going to be more like 95dB or so. Of course a jet airliner is going to be far louder, I assume that 140dB was measured from a significant distance, probably 50m or so.

G

 

[gregorio]

Interestingly I got 100.79 dBSPL, but the 0.1 dB difference is quite insignificant.

It depends what order you calculate it. Using the multiple sengpielaudio calculator entering 100dB and then 20 80dB entires gives 100.69dB but entering 20 80dB values and then the 100dB value (in the 21st slot) gives 100.78dB. I presume the 0.01dB difference with your calculation is a rounding error, as the sengpielaudio calculator displays to two decimal places.

G

 

[71 dB]

It depends what order you calculate it. Using the multiple sengpielaudio calculator entering 100dB and then 20 80dB entires gives 100.69dB but entering 20 80dB values and then the 100dB value (in the 21st slot) gives 100.78dB. I presume the 0.01dB difference with your calculation is a rounding error, as the sengpielaudio calculator displays to two decimal places.

G

I reduced the decibel levels by 80 dB giving me one trumpet at 20 dBref (100 on linear power scale) and 20 violins at 0 dBref each (1 on linear power scale). Adding these up on linear power scale gives 100+20*1 = 120 on linear power scale which is 20.79181246 dBref, but I need to add the 80 dB back and I have 100.79 dB. Doing the calculations this way keeps the numbers "small" and rounding errors are minimal if not completely avoided.

 

[Davesrose]

It looks like sengpielaudio calculator has pretty easy layout. It also might be pretty old: it's in HTML 4.01 which was standardized in 1999. They have their JavaScript right in the page for you to look at. Interesting that they are parsing the answer as an integer to two decimal points....with modern JS, you can have more accurate rounding with Javascript's Math.round() function. In fact, if I copy their page and remove the parse, the site's answer without rounding is 100.79181246047624 (and if I try Math.round it gets rounded to 100.79). Just a difference of fraction of a decimal (not important for audio, but an important factor for other applications). Notations for computer software is always interesting. I watch a few YouTube coding channels. A recent one showed how many calculators can't solve 6/2(1+3) correctly.

To modernize the site: it's also hard coded to have a loop of up to 30 source inputs. It would be simple enough to change the interface for you to add the number you want, set ranges of the same values, and then set the loop to whatever the final number of sources you've set to calculate.

 

[Vamp898]

After i liked some of their songs i gave two Albums from HoneyWorks a try. After testing the normal Version with 1 Song and comapred it to the Hi-Res Audio Version, i decided for the Hi-Res Audio Version and again, there is such a big difference in Mastering and Mixing. I have to go like 18 Volume levels higher compared to the Standard Version due to the highter dynamic range. And everything sounds much more organic and harmonic. Especially with the IER-M9 (or any other professional monitor, pretty sure) its such a joy to listen.

Its a strange feeling as they are very Jpop songs, so songs where people would normally not expect such mastering/mixing, but Japanese is just a world for itself^^

But again, the typical Pop Mixing/Masterin is reserved for the standard version and the people who invest into Hi-Res Audio get provided with an significantly different and better version of the songs.

So even though it makes no difference on paper and you should not buy Hi-Res Audio expecting to hear a difference between its 24bit or higher bit-rate, with a lot of music there can be drastic difference.

So often times (not always, i also experienced cases with identical mixing/mastering between the two versions) you get much higher quality with the Hi-Res Audio Version. For the wrong reason, but still. Don't refrain from buying Hi-Res Audio due to the lack of audible difference on a techincal level.

 

[bigshot]

As an example, The Rolling Stones SACDs have tracks that are completely remixed and original wire reverbs, slapbacks and 60s vocal futzes have been replaced by digital equivalents that don't sound at all the same. The effect is to soften the edge of the music, which makes it sound cleaner, but goes against the whole sound of The Rolling Stones. On some of the SACDs, the albums had different mastering on the CD layer than the SACD layer. It was obvious when they made the mistake of putting the single version of one song on the CD layer and the album version on the SACD layer. One had horns and the other didn't!

Mixing and mastering is a subjective process and it can be done well, and it can be done poorly The fact that it's hires is no guarantee that it's the same or better. There are cases when the hires is definitely worse. Usually more attention is paid to an album on first release where the whole crew is together supervising and approving it. 30 years later, an engineer who just gets the master tossed on his desk with a post it that says "Fix This" may not put the same heart and soul into it. Whenever they remix or drastically remaster without the input of the original producer and artists, it's a coin toss whether it's good or bad. This is true of multichannel mixes too. I've got hundreds of 5.1 and quad discs and the batting average is about 1/4 good, 1/4 bad and half ehhhh...

The only way to know which version sounds best is to ask collectors who have compared all the versions. Sometimes it's the original LP, sometimes it's the first release CD, sometimes it's the remastered CD, and sometimes it's the hires track. There's no way to predict the quality based on the format.

Believe it or not, there's a Frank Zappa album that is best on 8 track! He didn't like the way it changed tracks in the middle of songs, so he extended and moved stuff around to fit onto the four equally timed tracks and made it better in the process.

 

[GearMe]

...Believe it or not, there's a Frank Zappa album that is best on 8 track! He didn't like the way it changed tracks in the middle of songs, so he extended and moved stuff around to fit onto the four equally timed tracks and made it better in the process.

Seems apropos...

'twould be sacrilege to have the tracks switch right in the middle of 'Don't Eat The Yellow Snow'

 

[gregorio]

But again, the typical Pop Mixing/Masterin is reserved for the standard version and the people who invest into Hi-Res Audio get provided with a significantly different and better version of the songs.

There’s two issues here:

1. You get a more dynamic/better version on Hi-Res because audiophiles were gullible enough to fall for the false marketing and pay more for it. If audiophiles were not so easily suckered, then the Hi-Res version would be available in 16/44 format.

2. More dynamic range isn’t necessarily better. In fact, in most music listening circumstances more compressed (less dynamic range) is actually better.

G

 

[Ghoostknight]

1. You get a more dynamic/better version on Hi-Res because audiophiles were gullible enough to fall for the false marketing and pay more for it. If audiophiles were not so easily suckered, then the Hi-Res version would be available in 16/44 format.

arent the high res version tailored to audiophiles? this would mean that without audiophiles "falling for it" there wouldnt be a master like that

2. More dynamic range isn’t necessarily better. In fact, in most music listening circumstances more compressed (less dynamic range) is actually better.

well if we speak low listening levels you are right, but usually the dynamic range is compressed because of this, casual listeners just listen on "crap gear" and on lower volume levels where compression sounds indeed "better"

 

[gregorio]

arent the high res version tailored to audiophiles?

Sort of but not necessarily to audiophiles, just those with good equipment, a good listening environment and some critical listening skills.

this would mean that without audiophiles "falling for it" there wouldnt be a master like that

Why? Audiophiles and those mentioned above would still provide a market for “a master like that” but it would be distributed in 16/44 rather than say 192/24 because audiophiles have “fallen” for the false marketing of Hi-Res.

well if we speak low listening levels you are right, but usually the dynamic range is compressed because of this,

No it’s not. Usually the dynamic range is compressed for exactly the opposite reason! Because:

casual listeners just listen on "crap gear" and on lower volume levels where compression sounds indeed "better"

No. Casual listeners often listen at relatively high levels, say when driving, walking, exercising or travelling by other means, etc., because the noise floor of the listening environment is high and obviously they want to hear the music above that noise floor. Compression sounds better in such cases because uncompressed (or lightly compressed) peak levels would be uncomfortably loud/unlistenable. Also, with the possible exception of the headphones, casual listeners do not listen on “crap gear” in fact it’s usually as good as expensive audiophile gear or better.

Heavy compression is also “better” at low listening levels. It’s only in a relatively small range of listening scenarios where light compression could be better.

G

 

[Vamp898]

As an example, The Rolling Stones SACDs have tracks that are completely remixed and original wire reverbs, slapbacks and 60s vocal futzes have been replaced by digital equivalents that don't sound at all the same. The effect is to soften the edge of the music, which makes it sound cleaner, but goes against the whole sound of The Rolling Stones. On some of the SACDs, the albums had different mastering on the CD layer than the SACD layer. It was obvious when they made the mistake of putting the single version of one song on the CD layer and the album version on the SACD layer. One had horns and the other didn't!

Mixing and mastering is a subjective process and it can be done well, and it can be done poorly The fact that it's hires is no guarantee that it's the same or better. There are cases when the hires is definitely worse. Usually more attention is paid to an album on first release where the whole crew is together supervising and approving it. 30 years later, an engineer who just gets the master tossed on his desk with a post it that says "Fix This" may not put the same heart and soul into it. Whenever they remix or drastically remaster without the input of the original producer and artists, it's a coin toss whether it's good or bad. This is true of multichannel mixes too. I've got hundreds of 5.1 and quad discs and the batting average is about 1/4 good, 1/4 bad and half ehhhh...

The only way to know which version sounds best is to ask collectors who have compared all the versions. Sometimes it's the original LP, sometimes it's the first release CD, sometimes it's the remastered CD, and sometimes it's the hires track. There's no way to predict the quality based on the format.

Believe it or not, there's a Frank Zappa album that is best on 8 track! He didn't like the way it changed tracks in the middle of songs, so he extended and moved stuff around to fit onto the four equally timed tracks and made it better in the process.

Yes, with music that is that old, i would be very cautious. Especially when the people who did make the music where not involved in the process.

 

[Vamp898]

There’s two issues here:

1. You get a more dynamic/better version on Hi-Res because audiophiles were gullible enough to fall for the false marketing and pay more for it. If audiophiles were not so easily suckered, then the Hi-Res version would be available in 16/44 format.

2. More dynamic range isn’t necessarily better. In fact, in most music listening circumstances more compressed (less dynamic range) is actually better.

G

The difference is not just the dynamic range, so you maybe should not focus too much on that. Its the overall sound that is different.

But beside that, why would you release the audiophile targeted Version in 16/44? Why take the additional step and convert it? If you record in 24/96, it makes much more sense to just export it and be done. You don't have to care about all the settings how the conversion is done and check if everything is still intact and so on. And the size difference isn't that big with FLAC anyway.

If all people would have the exact same version (like it is done on bandcamp where you always get the highest quality for the same price), than all people would get the Hi-Res Audio Version and nobody would get an 16/44 because... again... why. You can just convert it yourself when you burn it on an CD, there is no need to have it converted beforehand.

But i doubt that everyone would want the Hi-Res Audio Version. Most pop music is optimized for Airpods and Mono Bluetooth speakers and car audio and so on. Especially on these cheap and rather bad playback devices, you want an loud and compressed version with stronger bass and pushed upper mids, otherwise the voice will just mush together with the rest of the music and there will be no feelable rhythm. But you don't want the version with the pushed voice and boosted bass with High End earphones, but you can't have both at the same time.

That is why two versions exist. I think this is an pretty good solution to be honest. You have one version that is optimized to sound well on cheap earphone and mono speakers and you have one version, that is optimized to sound as good as possible.

That some companies use this to rip off people is a negative side effect, yes. There should be some indication and/or better explanation what the differences between the versions are. But if this exists, i think its good to have both and not just everyone gets the same