Ok I’m confused after reading threads on bit depth?

So for the threads I’ve read here, and checking a file, when you resample , you resample to 26/16 but the last 4 bits are ignored giving you the 12 bit sound.

If I do that and put the sample to highfi mode, it sounds like the original , how can it be upsampling ?

Also what’s the difference between resampling to 26/12 and just turning a sample to classic mode ?

Cheers

Edit: did some more testing when putting on highfi on a sample I’ve resampled , there is a difference.
I was sitting next to one of my speaker not realising the other had gone into standby. With the loop I was using the difference was much more obvious once it was playing in stereo.
Which begs the question …

I think after the resample the track is put in to classic mode automatically? Is that right ? Why if you have just downsampled ?

1 Like

Classic mode and down sampling are the same thing. As you said classic mode just ignores the 4 least significant bits. It’s a 16 bit wav file but that’s only because there is no standard for wav files that uses 12bit data. You end up losing some dynamic range but you don’t really lose much fidelity at 12 bit as long as it is recorded correctly. The real mojo comes from the down sampling from 44.1/48k to 26k. The aliasing becomes more apparent. Also keep in mind that while the s2400 assumes 44.1 or 48k in hi fi mode, it will play back any sample rate that you have and this can create some cool effects. I’ve a bunch of sound font samples that are recorded at 11.25, 22050 and 32khz. These files end up getting sped up to match 48khz so you get some different aliasing when you move the pitch around.

2 Likes

Thanks for that. So I take it once you resample, it’s put into classic 12bit mode but doesn’t reprocess it on the fly , just sees it as a 12bit file to play.
Then if you have a 12bit file playing at high quality mode it just sounds like it has more fidelity because it’s playing it at a higher bitrate even though the source is 12bit.
It’s just in my head I thought a 12bit file processed at higher bit rates couldn’t improve quality.

I’ve kind of just reiterated what you explained on a more detailed read , thank you, I feel like I have it straight in my head now , I shall experiment like you suggest.

1 Like

Well let me clarify on bit depth. Once you drop it to 12 bit the files 4 least significant bits are set to 0000. There are still 16 bits in the file but the quietest parts of the file are now 0. That is really not very noticeable unless it was recorded quietly to begin with and you aren’t filling the higher order bits of the file. This can be a cool effect of its own though. The dusty sound comes from changing the sample rate since you start bumping into nyquest frequency and generating aliasing. I would bet that just about anyone here would have trouble detecting a file that was 12/26khz or 16/26khz that was aliasing. I think it is difficult to detect a 12/48khz and 16/48khz file as long as it didn’t have a sustained section that decreased in volume slowly to 0db. Anyway hope this clears it up slightly. I guess the TLDR is that sample rate makes things sound dirty way more quickly than bit depth which are more sensitive to volume changes and dynamic range.

2 Likes

Got it , frequency is more significant here