In Reply to: I keep hearing that the difference between 24 bit and 32 bit digital recording is negligible posted by Goober58 on June 28, 2025 at 20:53:24:
By design 0 dBFS is the loudest digital signal possible, say all bits are on.
16 bit can solve details down to -96 dBFS, 24 down to -144 dBFS.
Almost each and everybody records in 24, it simply gives you more headroom than 16.
Almost all recording software uses 32 bits, not because of the dynamic range but to keep the quantization error down.
You might analyze a couple of recordings. Most of the time 24 bits contain musical information up to bit 17 or 18 so not much better than a CD (16).
There are recordings with musical information up to bit 20 even 21 but this requires a very, very silent recording chain.
We have the same problem with playback. If your power amp has an SNR of 100, it can resolve CD (16 bit) at ease. If you want to resolve 20 bits, you need one of those ultra quiet modern Class D amps with a SNR close to 120.
Obvious exporting 32 bit makes no sense as it is far in excess of what our systems are able to resolve.
In practice LP's have a dynamic range of 50 dB...-..
The Well Tempered Computer
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Bit depth = dynamic range - Roseval 03:49:03 06/29/25 (11)
- RE: Bit depth = dynamic range - Goober58 09:16:50 06/29/25 (10)
- Another thought - 13th Duke of Wymbourne 00:21:29 07/01/25 (2)
- You're the man! - Goober58 09:54:44 07/01/25 (1)
- RE: You're the man! - Goober58 20:41:51 07/01/25 (0)
- My mother says - 13th Duke of Wymbourne 11:19:16 06/29/25 (6)
- Coloring - AbeCollins 08:28:58 06/30/25 (5)
- Let me recant that thought - 13th Duke of Wymbourne 17:14:45 06/30/25 (4)
- RE: Let me recant that thought - AbeCollins 11:12:56 07/01/25 (3)
- My experience - E-Stat 16:25:18 07/01/25 (2)
- Yes, same here - AbeCollins 22:07:36 07/01/25 (1)
- RE: Yes, same here - E-Stat 05:38:17 07/02/25 (0)