I don’t think you need dither in all seriousness when you’re bouncing from your DAW to a wave file tho, unless you have bit depth reductions in samples somewhere.
I see why you would think that, but it’s not quite the case.
The reason is that your DAW’s mixing engine is either 32 or 64 bit, so their will be quantize errors when bouncing to 16 bit.
I mean, chances are it’s not going to be noticeable anyways.
I also think uploading to soundcloud as 16/44k would probably result in the least artifcact… And that’s because they are certainly not streaming at a higher dit depth or sample rate, so you might as well give them one less thing to convert.
I’ll record/render to 32bit 48 kHz wave. For Soundcloud uploads I’ll reformat in Soundforge, first the bit-depth converter (with dither and noiseshaping) to 16 bit, and then save that as a 16 bit 44.1 kHz wave file.
About human-made climate change:
I try to limit my footprint as much as possible, eating local organic produce, not flying my jet so much and all that, - but I don’t really think it makes much of a difference (I don’t actually have a jet).
In my opinion, the main cause of the problem is global over-population.
There are simply too many of us, and people keep making more…
Telling them that they can’t have babies anyomore would be political suicide, - which means the problem can’t be solved democratically.
At least not the way things work today.
The “kill the babies” party is simply not going to get enough votes to make a difference.
I honestly don’t see a solution anytime soon. I’m sorry.
Perhaps we’ll get lucky and some new kind of plague will do the job for us. Who knows?
Who would have thought we’d have a debate on the best bitrate for soundcloud compression? I’m going full mp3 96kbps on my next beat battle
dude people say cattle exploitation is an evironmental disaster because cows produce methane or whatnot well you know what else produces methane in great quantites? humans. especially humans eating fast food. So, we need to first ban fast food, and then skim some of that human population. Here, that’s my campaign argument summed up for you
ahhhhh, so, in practice then, for the least quantization errors, should one use dither in a DAW when bouncing to 16 bit? Or even 24 bit?
Yes, always bounce 44.1kH 16 bit with dither. That’s standard for redbook audio CDs (a.k.a audio CDs), and you don’t really need anything more than that to enjoy a song anyway. Only reason to do 24 bit 48k is for video, because that audio lines up properly with the frame rate of tv if I’m remembering properly. In a studio setting, you’re gonna record all your audio in 48/24 as a minimum just in case you need to later put the stuff to video, but when you wanna listen to the song in your car you better bounce/render the actual mix in 44.1/16.
There’s even dither to 32-bit floating point now for the crazy plugins that work at 64 and 128-bit depths.
I don’t dither to 32-bit renders, which are what I render my mixes at. I do use dither for 24 and 16-bit renders of my master. I just put the 24-bit wav into soundcloud and honestly, their conversion has gotten pretty good. I have no complaints except for the occasional hint of aliasing, so maybe they could filter a bit better?
For other platforms though, they are nowhere near as good. Bandcamp specifically asks for your lossless masters, hearthis can stream losslessly so it also wants lossless masters. Pretty much everywhere else I’ve heard problems. I was just checking out Audius last night and I uploaded the same audio file to there and soundcloud called “The 909 Way” which makes extensive use of 909 samples including the closed hat. Now, this is best in lossless wav, there is a little added crunch on Soundcloud, but on Audius there is an occasional loud screech coming from the hi-hat. When I heard it I thought it might have been a live layer in the hi-hat sample I had forgotten about, but no, it was nothing but conversion artefacts.
When they say lossless do they mean uncompressed? Because a broadcast wave file is not compressed at any bit rate whereas an mp3 is always compressed. From my knowledge, shouldn’t you always do 16 bit anywhere?
Lossless = uncompressed I.e. wav, aiff, flac
Lossy = mp3, m4a, wma…
16-bit is good, indeed it’s the standard for red book audio as you said, but I often do 24 just for the extra bit depth.
Unless someone specifically asks me for 16-bit or I’m doing an MP3 (which is always 16-bit no matter what), I do 24-bit. Nothing crazy, it just means that quantization error is down around -130 db instead of -90 db, which means you can get away with that much less dither noise. Is it ever going to matter 99.99% of the time? No. But that one day that someone decides to play your track back on a PA running 115db peak loudness, you’ll sleep a little sounder knowing there’s just no possible way they’re getting noise floor from your recording. At that point it has to come from their signal chain, not yours (because your noise floor was 130 db because you rendered to 24-bit like a good boy).
actually low bitrate mp3 would be a fabulous rule.
lossless is generally referring to a compressed format in which the audio is somehow magically identical to the uncompressed wave. it’s black magic for sure.
Well, lossless doesn’t HAVE to be compressed. WAV is lossless, but not compressed. FLAC tho, is a lossless compressed codec. Pretty cool.
it’s all about 192khz/64bit music with 432hz tuning. everything lower than that creates weird artifacts on my beats pro.
The method I use was primarilly gotten to by a massive amount of trial and error we did several years ago - but to be honest, I have no scientific facts to justify it. We used to do CD mastering for local freaks here and we just used the method we thought sounded best and stuck with that.
I’ve used Soundforge since before SONY got involved, and it’s still my goto software for detail editing.
I’ve used Sound Forge forever as well, since back when Sonic Foundry owned it. Awesome software, loved the plug ins and FM synth.
less than a week left, how’s everyone doing on their submissions?
I got a Microgranny and an Octatrack this week, so I’m using these samples to explore both machines. Have a TON of material recorded, which was the fun part. Now to actually make something out of this mess…
i just started lol… we’ll see what happens. i’ll upload something
I honestly tried to make this glitchy - but, as you can hear, failed miserably in that. Perhaps the lack of glitch can be considered a glitch itself? Hope you can dig it:O)