View Full Version : Encoding 24bit audio (StaxRip)
latigular
8th April 2021, 00:14
Hello,
i have a problem encoding 24bit audio in StaxRip. If i choose ffmpeg or eac3to encoder the output file (eac3, dts) is always in 16bit. Is there a way to keep the 24bit depth?
Emulgator
8th April 2021, 02:16
Lossy compressed audio has no bitdepth anymore.
One can read 32bit float reported but that gives no safe way to conclude any source resolution.
The reported values can depend on the decoders, their implementations, and possibly embedded metadata.
latigular
8th April 2021, 02:30
Hello,
thank you for the reply but i have compressed lossy audio in DTS which has 24 bit, so there should be some easy way to keep it.
Audio #2
ID : 3
Format : DTS
Format/Info : Digital Theater Systems
Codec ID : A_DTS
Duration : 1 h 30 min
Bit rate mode : Constant
Bit rate : 768 kb/s
Channel(s) : 6 channels
Channel layout : C L R Ls Rs LFE
Sampling rate : 48.0 kHz
Frame rate : 93.750 FPS (512 SPF)
Bit depth : 24 bits
Compression mode : Lossy
Stream size : 496 MiB (5%)
Language : Chinese
Default : Yes
Forced : No
Ghitulescu
8th April 2021, 09:08
There is no real loss in going down to 16b.
In fact, having 24b and encoding "lossy" was already a bad decision to start with.
Keeping proportions it is like having an MP3 off a CD that has been encoded in 24b (I am not sure MP3 allowes 24b, but it's a fictional argument anyway) which then "has to be kept" 24b while re-encoding lossy again into another format.
hello_hello
8th April 2021, 20:55
While lossy audio has no bit-depth, DTS comes in a lossless flavour. In order for it to be truly lossless, there must be a reference bitdepth for decoding, and it appears they chose 24. Before DTS could be lossless, there was DTS 96/24 (https://en.wikipedia.org/wiki/DTS_%28sound_system%29#DTS_96/24). I assume the same 24 bit principle applies to DolbyTrueHD.
latigular,
Most programs would decode to 16 bit for re-encoding losslessly (flac or wave file etc) because the file sizes can get silly at 24 bit.
Can you clarify this?
If i choose ffmpeg or eac3to encoder the output file (eac3, dts) is always in 16bit. Is there a way to keep the 24bit depth?
I assume "(eac3, dts)" above refers to the source audio, but what's the output format?
If the output is lossless you'll have huge files at 24 bit. As a quick comparison I encoded 1 hour and 36 min of 5.1ch audio as flac. Once at 16 bit and again at 24 bit. The 16 bit version was 658 MB. The 24 bit version was 1.7 GB. If the output format is lossy such as AAC.... it has no bitdepth.
Richard1485
9th April 2021, 08:22
Hello,
thank you for the reply but i have compressed lossy audio in DTS which has 24 bit, so there should be some easy way to keep it.
As Emulgator has already explained, lossy audio does not have bit depth, so there is nothing to keep. There is nothing wrong with the output from StaxRip.
I wish that the MediaInfo developers would remove (or suppress by default) that stupid field from the program's output for lossy DTS. It causes nothing but trouble.
A reliable way to test for an encoding's effective dynamic range is to feed it a very quiet signal, decode with a known good player, then boost both up to audible levels and compare. Lossy encoding (mp3, ac3, dts) will only exceed 16 bits during very quiet sections that are usually inaudible, and usually never in VBR mode, because the bitrate will be automatically decreased as the levels drop below the threshold of hearing. Some encoders will also use less precision to be fast.
stax76
10th April 2021, 16:37
I wish that the MediaInfo developers would remove (or suppress by default) that stupid field from the program's output for lossy DTS. It causes nothing but trouble.
It will be hidden in the next MediaInfo.NET (https://github.com/stax76/MediaInfo.NET) release.
Richard1485
10th April 2021, 18:02
It will be hidden in the next MediaInfo.NET (https://github.com/stax76/MediaInfo.NET) release.
:thanks:
vBulletin® v3.8.11, Copyright ©2000-2025, vBulletin Solutions Inc.