The sole variable when I created the different WAV files from my turntable is the saving format in Audacity. This suggests that the differences in the 24-bit and 32-bit sound is attributable to the format since the recording platform, both hardware & software, was unchanged. In fact, the two files, one for 24-bit and one for 32-bit music, were created minutes apart, with just a simple 'Save As'. |
sorry, I read that as you used the RME at 32 bits and the Lynx at 24 bits. You didn’t state that you recorded at 32 bits and saved as 24 bit files until now. However, the software has to do something to convert the file from 32 to 24 bits. That doesn’t mean you will get the same results using a native 24 bit chip rather than a 32 bit chip that has to be converted to 24 bits. Perhaps something is lost (distorted?) in the conversion. There are still too many variables to definitively state that recording with a 32 bit chip is better than a 24 bit chip. All you can say is that when this program converts your 32 bit file to 24, you prefer the original 32. Maybe if you record at 24 and convert to 32 that will be equally good? If you are insisting on 32 bits, I believe your choices will be severely limited... good luck in your search |
a few more thoughts.... any time you manipulate a digital file (whether it is to apply EQ, change the bit depth or the sample rate. or filter or change the file in any way) you create artifacts such as pre-ringing, or post ringing, or phase anomalies, and/or other unwanted results. That is why a program like HQ Player offers many options for manipulating a file. Some are listed below.. they all sound different. It is why DCS offers various options for applying digital filters. Why Ayre has a listen and a measure mode on some of their DACs. Why Roon has many options and settings when changing sample rates. My point being, taking a 32 bit file and converting it to 24 bits is not as simple as it seems. Recording at 24 bits and converting a 32 bit file to 24 is not the same thing. The only difference may be the way Audacity saved the file, but it had to manipulate the data to do that. The sound will therefore be changed in ways you don't know. Resampling filters: ‣ 3 intermediate phase ‣ 11 minimum phase ‣ 4 impulse optimal ‣ 3 closed form Dithers and noise-shapers: ‣ 5 noise shapers Delta-Sigma conversion: ‣ 70 oversampling filters (64x – 2048x) |
Check out the new miniDSP ADept ADC and Phono Premap 123dB SNR / -117dB THD+N ultra low noise and distortion
|
Herman: "Of course, this is with my gear." YMMV. For my recordings, I have reduced the variables to the bare minimum: from the raw data in Audacity, Save As 24-bit or Save As 32-bit. Errors could still be present, but I am satisfied with the conclusion that 32-bit uncompressed music files sound better than their identically recorded 24-bit versions. Audacity is set to record at 32-bit float, 384kHz. Windows saves as 32-bit Signed. Maybe HQ Player has better rate conversion than Audacity/Windows, but I’ll leave that question for others to answer. Thanks for your input. |
