Hi guys,

Could you tell us which sample rate you chose ?
Why this choice ?

Because a friend recommended it ?
Because you did tests and noticed a difference ?
Because of the style of music that you make ?
Aliasing problem ?
Another reason.....

Thanks again for sharing your knowledge

 

I have always been 44.1 but today I decided to switch to 96 Khz to try and see (and listen) if I find a difference.
I'll tell you if I notice a big, a small or no difference in the quality of the rendering, the stereo, the depth or the clarity of my mixes.

 

Last edited: Dec 2, 2019

96/24 for almost 8 years.

The reason - to push plug processing so far past the Nyquist limit that aliasing etc., isn't a problem. 32 bit floating point is 24 bit integral within a matissa - like 1.234 * 10^4. You only get 24 bits of resolution which you lose 1 bit from every 6dB above 0dBFS you go with 32bit float. 96/24 with EBU gainstaging.

 

Last edited by a moderator: Dec 2, 2019

I do not understand everything, what is Matissa ?

 

88.2 / 24

 

cause they know this in and out here....

what stuff do you produce...whats youre computer specs...

 

In the first place, 48kHz was introduced initially for the sake of synchronising with video only /for the audio parts of videos/.
But CDA standard is 44.1kHz. All seminal digitalisation of analog studio tapes (when they were more or less still 'fresh') in the past century was done in 16/44.1kHz standard. CDA standard still rules today. So in order to make a CDA you need to downsample files that are rendered in 48/96/192 standard. By that you get an extra quality loss because those sampling rates are not divisible by 44.1, so interpolation must be used.
mp3 are also just 16-bit. 90% of all recorded music ends up in mp3, especially on radio stations, but in internet exchange too.
So I record and master in 24/88.2. Whenever you go from 24 to 16 bit you should use dithering.

 

Last edited: Dec 2, 2019

48khz/24bit . 48 sounds nicer on top end than 44.1. try with clean electric guitars.
also, latency gets smaller with higher sample rates for a strange reason. 48/24 is the sweet spot.
48/24 sounds bigger and more detailed compared to 44/16, i used that for 10 years

 

Dither any audio that you are downsampling and then worry not. Read this cause I'm lazy.

https://www.waves.com/audio-dithering-what-you-need-to-know

 

unless I am recording Paul Simon solo with an acoustic only etc, I generally stick to 24bit/48. Have done projects at higher, but whatever. Working with picture a lot and with whatnot, it has served me fine with decent converters on input and whatnot.

 

48 for audio, 96 for video, 192 for design. Then the down-sampling happens and it's out of my control, unfortunately.

 

I only do 64bit/384khz because at this time of year i get distinguished clientele, like here in my studio:

 

Numbers in scientific format consist of a mantissa (e.g. 1.2345) and an exponent (e.g. 4 in ten to the power 4). Put the two together and you get 1234.5 as the true value. Floating point 32 bit audio files are the same you get 24bit of resolution and eight bits of exponent. If you fill up all 24 bits, the mantissa gets rolled along one bit and the exponent incresed by one. This causes the least significant bit of data to get rolled-off, when pushing a signal above 0dBFS. WHen the signal is mixed down top normal levels, the bits get rolled back the other way, but the fine detail is lost and the lower order bits are filled with 0's where there used to be detail. Pushing 32bit float past 0dBFS is equivalent in every respect to truncation without dither losing one bit for every 6dB. So if you push a 32bit float mix to +18dBFS, when you mix it down to 0dBFS you've truncated its bits to 21bits with no dither.

 

If youre asking then you probably dont need it or hear it.

 

@taskforce I've worked with the guy in middle. He always wants His stuff rendered all over 3.
I don't even know how to convert or render that but the sound of His words are always beautiful.

 

Last edited: Dec 3, 2019

48khz/24bit mainly because that's what the clients request and pretty much industry standard for broadcast audio.
I only record higher sample rates if I'm recording audio intended for sound effects/ manipulation.

 

I've used 44.1/16 for years,
then I moved to 44.1/24 for most recording (CDs are still 44.1/16),
thinking about 48/24 recording if I decide to do more video (because pretty much all video stuff is 48/16),

there is no reason to go above 48 for recording,
for mixing, it's more tricky because those fancy analog sims are crippling harmonics back to lower frequencies when not oversampled, I'm experimenting with 96/24 mixing and 44.1/16 render but I guess I'm deaf anyway

 

Even for the lazy ones, it's quite simple:

• Only dither when you render your audio to a lower bit-depth.
• Don’t dither before converting to MP3 or AAC.
• Always dither if you’re creating 16-bit files for a CD from a 24- or 32-bit mix.
• You don’t need to dither when going from 32-bit floating point to 24-bit (because 32-bit floating point doesn’t have a higher bit depth), but you do from 32-bit fixed point to any lower bit depth.
• If you’re preparing 24- or 32-bit data for a mastering engineer, don’t dither. Let your mastering engineer, who’s an expert, handle it.
• Make sure that dithering is the last step in your processing chain. Never insert a processor after it in your DAW or audio editor.
• Try to avoid dithering twice, because the effect is to add more noise to your audio. That said, if you go down in bit depth twice during a project (which is somewhat unlikely unless you start at 32-bit fixed point), you should dither during each of those conversions.

 

Last edited: Dec 2, 2019

Yes,when you are listening all to yourself. But if you try to sell it, you'll have to re-encode it into something else. No commercial sound carrier distribution format supports 24/48 standard.

 

24/96 on most of my recordings.