24bit vs 16bit playback

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
As an engineer I kinda tend to agree, at least until I hear a difference with my own ears.

Imho, it's a very wize statment :) !

But also the experience can have a different origin in the result due to some no controlled factors as well (I don't talk of psycho or blind test which are always the easy answer when the experiment does not match the theory to find an answer).

That said, the final result is more the result of the whole layout of the DAC ( and also the others Hifi devices too). I agree with Esgigt. A simple 16 bits with a good layout will give you a very good sonic, more than you need or the whole hifi can play (often with a very good DAC there are more weak links than the DAC itself).

But as always, each are looking fort different results. One likes non amplified event, others want to hear the teeths of the singer... In each case it gives you a different tonal balance than you must adapt to the room, your tastes.... For instance a musician will not set up the whole like an other. But He may be happy with 16 bits as most of the dynamic behavior of the reccording.

Now I'm not the sure the opposite is false. I'm not sure Neil Young is controlling all these links and knowledges. I don't see why a 192/24 should sound worst !

Why a Soekris Dam does not sound like a Total DAC e.g. has more to see with layout than the bit depth or sampling rate.

I would like to listen to the last DAC of ECEDESIGNS which has a 16 and 24 version, but dunno if the layout around the discrete R2R is exactly the same; a simple different cap can change all the result !

My simple 2 cents as I understand it and if the interpretation of my experience is correct (which is very not a certainty or a definitiv one :D).

I have an other certainty : the markett ! The market says : most gainq with downloading will be made in mobile situation with Smartphones... so for the moment you need light datas. Firstly to raise the margin (the more the datas is heavy, the more it cost you to manage it when you are a web enterprise !

Btw...April fool D in this european part of the world.... Neil has an Android not an IPhone ! :p
I don't see why a 192/24 should sound worst !
Neither do I...

IMHO, case is the implementation quality from recording to listener.
As I stated earlier, 16 bits is more than sufficient under most listening circumstances. The sampling frequency is an other ball-game, since what we hear is not entirely made up of "raw" frequencies between 20 and 20Khz, but also of the IM products of higher harmonics.
But that direct sound is only a part of the listening experience, because people are also sensitive for changes in pressure (lower than 20 Hz, sensing "space") and a form of contact sound in ultra-sonic regions..

Agreed, it's not sound as defined, but the analogy with light (in the article) has a big problem because we are not aware of the side-bands of visible light (except for the infra red, which we can feel as heat), but we are aware of the audio side-bands to some degree. From that perspective higher sampling frequencies than 44.1kHz are defendable.
Recording and processing in 24 bit can be very useful since the signal levels aren't yet defined. For playback, that's a lower noise floor than needed since the signals will already be set so that the highest peaks are at or near 0dBFS. 16 bits give a noise floor that's much better than needed in domestic environments.
In my living room I'm rarely below 40 dB noise (there is always a little noise : wood furnitures, external event through the Windows, etc...) just when breathing !

If the reccording can allow it, I will be happy to have 24 bits to have more lively sensation on some milliseconds peaks than acoustical Jazz and classical music allow!

For that I will need a speaker able of 100 or 130 dB effcienty with chair at 1 meter to profit of some transcient on some Wagner appertures or Jazz bands !... while the current level is only between 80 to 90 dB !

If :

The room allow it,

The true 24 bits

The true non compressed reccording with chain process at 24 bits with sota engineer à la Chesky !

But I really think one should have if liking lifely acoustical music a 115 dB efficienty, the more you have the more you avoid thermal dB compression of the drivers (thermal issues, impedance chgts, phase behavior beetween drivers), because it's easier to make a good amp with few watts, because cones whith low exursions distorss less on dynamic peak (which can drill your ears) and because some rare reccordings Worth it !

Could I be happy with 96 db efficienty ? Certainly, for the moment I live with a good sensivity of 85 dB at 2.83 v /m !
Last edited:
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.