Not “what do you record at” - everyone has their own reasons for recording and mixing anywhere from 44.1 to 192kHz, I trust we all record at 24 bit and mix at either 32- or 64-bit float, or 48 bit fixed (I think Pro-Tools did that for a bit?)
What I mean is, when you mix down, when you master, when you’re left with a folder with your final, ready for distribution audio files, what’s your target?
I ask because the project I’m working on is pretty much finished - I decided to do DIY mastering for my 12-track album, and I’ve got 4 tracks of those that I’ll also get properly mastered for a promo EP, sending to radio etc.
The project was done at 88.2kHz. I don’t intend to do a run of CDs, it’ll all be online distribution. So, is there actually any point in downsampling to 44.1kHz? And is there any point going down to 16 bit? Why not just stick at 88.2, 24 bit wavs for those who want to get full quality files, since the vast majority (of the 3 people who will ever chose to listen to my album… ) will be perfectly happy with mp3 streaming/ downloads? And I can make the mp3s straight from the full resolution wavs…
Or, would you go down to 44.1/16 bit anyway, because that’s the standard?
What happens when you try playing an 88kHz mp3 from different web/software players? What is doing the converting down to whatever the hardware wants to hear? Is it the server (I doubt it)? Is it the software? Is it the sound card?
The main downside (aside from size, which may not matter) is whether or not the end user can even play it back. And if they can play it back, is it getting screwed up by the conversion more than you would have screwed it up by converting it?
It would depend upon the online distro requirements. I believe most take 24 bit wavs at either 44.1 or 48 and do conversion on their end. I also think that there are a few specific “hi-res” venues available that could take advantage of 88.2.
But I would worry about getting the best sounding mix and file you can and hope for the best. It is out of your hands at the end-user level.
If you are going to have something professionally mastered, ask the engineer what he requires and what his final product recommendations are…
Spotify I don’t really know, I think it has its own conversion system going on, but I can’t find details of how to directly submit files to them - you need to go through an aggregator.
Itunes expressly say that the higher the quality the better in their own guidance;
“An ideal master will have 24-bit 96kHz resolution. These files contain more detail from which our encoders can create more accurate encodes. However, any resolution above 16-bit 44.1kHz, including sample rates of 48kHz, 88.2kHz, 96kHz, and 192kHz, will benefit from our encoding process.”
That seems quite definitive - and suggests that if I happen to be producing at 88.2, they don’t feel that I need to downsample before submitting wavs to them.
Soundcloud just sounds bad to me no matter what resolution wav I upload - I get better results uploading mp3s I’ve converted myself, and that seems to hold true when I read what others complain about online.
For my own personal distribution, I’ll send 320kbs MP3s encoded with lame at the highest quality options to people, so the question there is, does my particular encoder work as well with 88.2 as it does with 44.1 wav files fed into it? So far, I’ve not been able to tell a difference. It’s outputting 44.1khz mp3s, btw…
That just leaves the possibility of distributing the high quality wav files. I don’t really know any environment that can’t play 88.2/24 bit wavs, though I’d be open to learning of them! Android can, Apple’s IOS can, Windows can… and on the subect of Windows, Windows 10 re-samples everything through whatever user defined sample rate/ bit depth you choose in the control panel. And it’s defaulted to 48kHz as far as I can tell. So even if you’re distributing 44.1kHz files via CD or otherwise, if it’s playing back in windows 10 it’s probably being resampled on the fly regardless by the audio engine!
I agree, really, and thanks for posting. I doubt anyone could tell the difference since these tracks aren’t all about audio above 20kHz and levels around the -100dB mark. But the question behind my post, really, is, in the current environment does it still matter enough to bother with SRC or bit depth reduction at all, unless you’re burning to CD? Or might I as well stick with the sample rate I started at, since it’s going to be mangled by audio drivers, MP3 encoding, various streaming algorithms etc anyway.
And if in the future I do want to make it a CD, it’s just a question of finding SRC I’m happy with (does anyone have any opinions in that regard? Anyone not happy with what their DAW does?) and dithering to 16 bit?
Just food for thought, really. I suppose I do agree that 16/44 is a fine spec and there are legacy reasons to stick with it.