Ideal and practical DACs resolution

[SergeG]

Hi,

16 bit data theoretically provides with 96db Dynamic Range (DR) (20*log(2^16) )
24 bit - 144db 20*log(2^24)
32-bit - 192db 20*log(2^32)

But, parameters of the award winning ES9038PRO 32-bit DAC
provides with "up to 140dB"
(ESS Technology :: ES9038PRO)

Actually 140db DR can be achieved with just 24 bit data.
What are the 8 (32-24) LSBs for?
Anyway signal determined by the 8 LSBs would be below the practical noise level.

The 8 LSBs can't affect THD as well, ES9038PRO's THD is 122db
And 24 bits is more than enough to make that.

Is it just a marketing as "the more bits the better"?
Or just a technological consequence as "even number of bytes are more technological"?

What am I missing?

Thank you,
Serge

 

[iperv]

https://www.analog.com/media/en/training-seminars/tutorials/MT-003.pdf

Enjoy the reading.

Enob is the keyword here.
And marketing the second one.

Edit:
https://www.analog.com/media/en/technical-documentation/technical-articles/ADI-data-conversion.pdf

Edit2:
from the last document, enob=(140-1.76)/6.02 bit = 22.3bit
So i think you have to read that 140dB of snr as a measurement of enob, that means it has circa 22,3bit of effective resolution in bits.

 

[SergeG]

Thank you!

 

[planet10]

The Stereophile measurements on DACs suggest that the practical limit on magnitude resolution is something like 21 bits.

dave

 

[iperv]

The Stereophile measurements on DACs suggest that the practical limit on magnitude resolution is something like 21 bits.

dave

Over that value, the electronic noise floor is predominant.
Over that value, in fact, is marketing.