Hello, im using not many plugins. What are you using and set in your DAW? What i have to consider to use 48 khz? Thank you. I just want test it because i read somewhere that 48 has less latency than 44.1 and it seems 48khz is more (recommended) standard on platforms like youtube, soundcloud.. thanks
It depends what you're doing for Youtube.. 48Khz is usually used for video production (A single session can only have one sample rate or you risk morphed audio). 16 and 24 Bit Audio you can hear the difference (somewhat) 24 bit/44.1Khz and 24/48Khz who knows? Youtube used to resample all audio to 44.1Khz, your session sizes are going to increase if everything is at 48Khz, Youtube upload requirements As for plugins, i'm not sure but some VST's have the oversampling button? As for Soundcloud, I think what could change your quality game is the bit depth 16 or 24 should be good for WAV files (ONLY FOR DOWNLOADS) All Soundcloud audio is streamed at 128kb/s. At the end of the day, use what works for you as the ' quality ' really only caters for those who consume content in HD. Read these links: http://www.keyboardmag.com/how-to/1255/should-you-record-at-96khz/48364 https://www.gearslutz.com/board/rap-hip-hop-engineering-production/690225-96khz-sample-rate.html
48K isn't necessary/advisable in non-video audio unless you're recording 24-bit audio, for which 48K is more amenable (following the 2X rule, with 2X24 = 48).
Since 2013, I'm exclusively using 48kHz for multitude of reasons, regardless of target media: Many Kontakt libraries defaults to 48k, so when working native, less SRC have to be performed, leading to slightly better SQ and lower CPU usage Opus codec, used by Youtube and most modern internet media works only at 48k. 48k provides slightly more headroom during band-pass SRC process, and thus allows the use of shallower filter, leading to less ringing artifacts. 48k is the default sample rate for Windows MME and ALSA. No resampling is necessary, leading to potentially better SQ. 48k is easily scalable to 96k or 192k, while 88.2 or 176.4 might not be supported on many devices also, when doing SRC from 48k to 44.1k (as opposed to 44.1>48), it's potentially less "lossy" as aliasing artifacts don't occupy the last portion of the spectrum As for 16bit vs. 24bit - for final renders, 24bit is overkill. Properly dithered 16bit has enough headroom for every possible genre and every possible listening scenario. Soundcloud (to my knowledge) compresses everything to dinosaur 128kbit MP3 at 44.1kHz. Last edited: Feb 20, 2017
48k will give you such a minor amount of latency differential it isn't even worth considering in that degree... I work a lot with picture and whatnot, so now use 48k, i certainly couldn't hear the difference between that and 44.1.... when you go to 96/192 etc, yes, latency will improve,.... but don't worry about it too much... 24bit 48k works for me in most circumstances, 96k if recording acoustic or something, but probably overthinking it... but try and stick with one... definitely a nice improvement from 16 to 24bit (especially as a beginner and learning headroom/gainstaging etc....)... and hard drives are cheap...
I work in 48/24. For my "mastering" sessions, I bump the bitrate up to 32. Then dither it back down when I do my final renders. If I was recording a lot of live stuff I would probably try to work at 96k.
@Spacely @Nimbuss @stevitch @Andrew @sisyphus @subGENRE Thanks for that answer but.. is that really true? i mean Soundcloud is more dedicated for music as Youtube, but Youtube has millions of 4K-Videos. I would not understand when this low quality of Soundcloud is just to finance their servers
I'm slightly confused by your answers .. DVD = 24bit 44.1 CD = 16bit 44.1 so when you go to 96/192 etc, yes, latency will improve, ????????? Since when did more drain on the cpu sound card and pc mean less latency I'm slightly confused here .. I've always worked at 24bit - 44.1 for every thing be it film radio tv ect and i just cant wait to hear your reply !
Assuming you keep the same size sample buffer, the higher your sample rate goes, the lower the latency becomes. Whether your CPU can handle the load is a completely different story. e.g with buffer at 512 samples and increasing sample rate at 512/44100 = 11.6 ms latency at 512/48000 = 10.6 ms latency at 512/88200 = 5.8 ms latency at 512/96000 = 5.3 ms latency at 512/192000 = 2.6 ms latency Just to clarify: 96/192 doesn't mean 96bit 192kHz
Hello. If you allow me a joke, with today's wonderful mastering using heavy limiting (especially the beloved L2), the audio quality on many commercial cd's is equivalent to 10 bit / 30khz. My advice is to use the highest available quality for recording (24/96 is a good compromise) and go from there, according to specific project requirements.
Since CDs as media are mostly dead, there's no need to do anything in 44.1/16 any more. I'm all for making that format go away as soon as possible. If you decide to make a CD one day, you can easily convert it with a really good SRC like Voxengo R8Brain [better quality than Weiss hardware SRC which still costs a couple of thousand $]. Generally, I'm recording at 96/24 or even more often 48/24, depending on a project. Acoustic music sounds nicer at 96/24, and when I'm doing electronica and using my samplers I use 48/24 because I don't feel like there's need to use anything better. Besides, 48/24, and 96/24 to some extent, has become more of a standard than 44.1/16, due to Youtube and others. And really, what type2002n said, these days, even though we could make such great sounding songs/tracks, we destroy them in the name of "loudness". Loudness with quotes because one day you'll witness how limp it sounds when compared to a quality dynamic recording at the same level. It's also best to preserve/archive the dynamic masters in full resolution for such case, so you can remaster as needed.
My format of choice for projects would be 48/20. Enough for mastering (24dB extra compared to 16bit) and still practical, as the last 4 bits in 24bit container are rarely used.
