So, I am an amateur when it comes to mixing, and I plan on sending the stems of this project to a professional once I get everything together. Nevertheless, I still feel as if I should ask this in consideration for whoever that engineer may be. Anyways, I am working on an album that is attempting to blend synths and orchestral kontakt libraries together (As well as some live recordings I plan on getting recorded of solo violin). Upon looking at the specs of some of the kontakt libraries, I noticed that most were recorded in 48khz and 96khz quality (Spitfire Symphonic Strings, Orchestral tools Metropolis Ark 4, etc. ). I am unsure of what to do, because I do notice a difference in quality when I switch to a higher rate, but I know that 44.1k is still standard for streaming services. I have heard that if an engineer converts a 96k project to 44.1K the depth/brilliance is still maintained...but I have also heard the opposite. I should mention that I am using a multitude of software synths such as Serum, Zebra, Sublab, and Omnisphere. I am also curious as to what my bit rate should be. I have been suggested to export at 32bit due to the variety of instruments being used, but I am not hesitant to change that if I am convinced otherwise.

Thank you!

I hope my question was clear. I would be happy to try and clarify if it is not.

 

What does this mean?

 

I can hear intricate details that were muddled before, and it seems to just be clearer altogether.

 

if your computer can handle 96khz ... just do it in 96khz.. downsample to 44.1khz will be at the last process during mastering print... let mastering engineer concern about it. Or the mixing engineer that mixing your song downsample it if he/she want to work with a lower sample rate.

but here the thing that needs to be cautious about.... of coz doing in 96khz would be beautiful, u can here the better resolution of the plugin processing, such as distortion plugin, harmonic plugin, reverb, delay... but if only the plugin capable to works with 96khz...

the thing that needs to be concern is for the plugin that not made to work with 96khz sample.. especially dynamic processing plugins.. it could introduce ringing and anti-aliasing issue ... there is a workaround like inserting TDR Ultrasonic filter before and after plugin that not support 96khz.

I won't be worry so much about if the concern is the sound library samples... so far I'm using variety sample rate samples with no problem. But I have this habit to keep the equal adds on the sample rate... like if it 44.1 with 88.2 and 48 with 96.

but if it is so much hassle like u just wanna focus on writing music.. just do it in 48khz.. it still wonderful. No problem at all.

you won't hear much or maybe none noticeable of critical differences if comparing 2 samples rate files, but u will hear or feel the differences when you start working and processing it with plugins.

 

Why I'am asking is because if it gets more brilliant it is because the sound gets pitched if you change the samplerate. And more brilliant usually means better. The reason for this is that the samples get scanned faster and therefore the speed and pitch changes because Kontakt corresponds completely with the host in terms of time. The samplerate however do not get translated to the samples so it begins to simply play the samples faster.
You should never change the samplerate while using Kontakt in a daw and it is generally not necessary either. If you want to test this you have to make two projects in two different samplerates and then you will not notice this anymore.

So, the general rule here is: Use at least the samplerate of the samples recorded with the highest samplerate if you are able to. The reason for this is that samples which are recorded in high samplerates do have information above 22kHz and those informations can create aliasing if you have chosen a lower samplerate.

Why do recording studios chose a higher samplerate than 48kHz? Because they mostly use analogue equipment to modify the sound and therefore they also get harmonic distortion above 22kHz. So if you decide to use a sample library with recorded orchestral samples at for example 96kHz you should ask yourself "how natural are those sounds anyway if they need to record with this samplerate?"
Because changing the sound with analogue equipment is really mostly the reason.

 

Last edited: Feb 25, 2020

Thank you, this is very insightful. So, just to clarify, I should set my project to match the library/sample with the highest sample rate as to prevent aliasing? Do you think that if a sound is expendable, is at 96k, and can be replaced by a recording set at 48khz I should do it? Like if that one sound is the reason I have to put the project at 96k, because one thing I did noticed was that one of my libraries stated that it was recorded at 48khz, but it sounded better when I set my daw's sample rate to 96khz? Do you know why this might be? Will that quality be most likely tarnished once it is processed with various plugins?

 

Not the kind of answer you're looking for, but did you compare your tests side by side?
I mean rendering in both 48kHz and 96kHz to audio, level match them and then hearing the results? cause if you just change the sample rate in your DAW settings and hit play, what you hear is biased and if you reload your DAW and wait for the whole project to load again, your ears need time to adjust.

 

Again...

What means better?

You can only decide what sounds better if you make one project with for example 44.1kHz and another project with for example 96kHz. Render both and compare the two wave files. Just take one violins sample let it play one bar with the note C4 in both samplerates in different projects and then compare. No big difference will be noticed and if a difference is noted it should be strong because aliasing creates a harsh sound.

 

Upper values over 48 Khz are only good for acoustic instruments and voices recorded with high end microphones.

Only if you use real samples recorded on high end equipment that came at origin in 48, 96 or more high sampling rate, you can set up your projects at the high sample rate.

 

I would say it mostly only matters if you also record live instruments/vocals.
Otherwise you can use whichever, and also switch it on the fly for redering when you want.

 

Pitch=frequency. Do you mean tone (harmonic balance, upper harmonics accentuated rel. lower ones)?

Meaning what? moar high end?

If that was true, a 1-sec-long sample recorded @44k would be .5-sec long when transcoded to 88k. This doesn't happen.

For the same reason graphic artists work with large (as in high pixel count) raster images -- to minimize aliasing when those are resized to use in future ~~mixes~~ projects.

Think of audio bitrate as image size, audio bit depth as image color depth (number of possible colors), transcoding (changing sample rate) as stretching/shrinking that image, and you won't be far off.

 

Interesting video

 

Last edited: Feb 25, 2020

What would you recommend in my situation? I wil be releasing mainly through youtube, spotify, apple music.

 

So I did a couple sound tests and I actually got my brother to listen too. He has a decent ear for music. His response was that the 96k sample sounded clearer than the 48k, but he still preferred the 48k as opposed to the 96k. He stated that it sounded more authentic when compared to the 96k. I still personally prefer the 96k sound due to its clarity. I can make it sound more authentic with proper dynamic changes and maybe using something that sparsely changes the pitch (iZotope Vinyl), but I thought it was very interesting to hear him say that nevertheless. I am still on the fence, but I might just stick with 48k for my computer's sake. I am sure a good sound engineer can mask minor aliasing...unless I am wrong.

 

Yes, I did this and the the 96k sample was much clearer. I noticed that a bit of top end of the violin sample, which was recorded in 96k, sounded cut off in the 48k sample. I honestly probably would not care that much if it was a synth, but since it is an orchestral sample I think it is worth considering, because I am trying to pick up the human interaction with the instrument as clear as possible.

 

Then @Jeffriezal gave the best solution answer for your situation. Work in 96k, finish the song and let a mastering engineer take care of the final stage. Even if you mix your own songs and you're good at it, I still suggest to leave the mastering to an expert with a set of fresh ears.
I'm not sure if your instrument remains at 96k when you load iZotope Vinyl.

 

Digital domain in audio is not the same as it is in visuals and design. I study graphic design as a hobby and you can trust me when I say that. What you said about images and digital canvas is true but comparing it to audio is just wrong in many ways.

 

Actually, it makes quite a big difference in VST synths, too. Especially those that don't have built-in oversampling.

 

Here is a great article on the topic by Ryan Schwabe.
Mixing and Mastering at 96kHz
http://www.ryanschwabe.com/blog/96k

 

To answer all your doubts, questions and everything else:

All thanks to Dan as usual.