What's new

Sampling Rates for Orchestral Mockups: 48kHz vs 96kHz

As far as i'm concerned, the only reason for 96k is the aliasing issue with some plugins. What's your opinion on the aliasing topic?
If aliasing due to (some of your) plugins is your only concern, check out DDMF's Metaplugin. You can demo it, and see if it works for you.
 
Well, looks like that settles it then, I will now be exporting at 64 bit float 192kHz!!
The only two sensible reasons for any professional to record at a super high resolution are:
1. Archiving;
2. Reproduce material on audiophile-grade gear. (EDIT: Either Analog synths or Acoustically-recorded instruments)

When I used to work in mastering studios, I kept seeing some people treating/seeing digital in the same way as they see analog (where you can upsample information by a ridiculous amount without producing artifacts) -- completely disregarding about what's to come. I personally don't think average consumers will ever care for audiophile gear in the same they do for high-resolution TV sets, but who knows?

But yes, the difference is noticeable. Even the average consumer can distinguish a 48Khz and a DSD recording if he/she is in a great listening room full of audiophile gear. The top end is very distinct..not because they can hear the high-freq content well, but because of all the "e.g:airiness" and overall spread information across the range in which they can hear. The low end is also far better, but I somehow haven't seen many people distinguishing the difference on that range.
 
Last edited:
Even the average consumer can distinguish a 48Khz and a DSD recording if he/she is in a great listening room full of audiophile gear. The top end is very distinct..not because they can hear the high-freq content well, but because of all the "e.g:airiness" and overall spread information across the range in which they can hear. The low end is also far better, but I somehow haven't seen many people distinguishing the difference on that range.
All very subjective.
 
All very subjective.
It's mostly contextual... There probably never was any point to a high-resolution audio print of the Dead Kennedys, but there probably is to a high-resolution audio print of Saariaho's chamber music (properly recorded, of course.)
 
... I especially appreciate the last paragraph:


Also, the common misconception is that somehow it increases the accuracy of the waveform by adding more points to guide it in the right direction.

But it ain't that way. The waveform is a bunch of sine waves. I always explain this by pointing out that a speaker can only go in and out.

(Obviously this is not directed to Dietz, who it's safe to say would know that. :) )
 
Also, the common misconception is that somehow it increases the accuracy of the waveform by adding more points to guide it in the right direction.

But it ain't that way. The waveform is a bunch of sine waves.
The only thing it will change is that any additional processing on the audio with higher sample rate will have more points to work with -- but technically at that point, any additional processing is distorting the source audio (for good or ill).
 
Also, the common misconception is that somehow it increases the accuracy of the waveform by adding more points to guide it in the right direction.

But it ain't that way. The waveform is a bunch of sine waves. I always explain this by pointing out that a speaker can only go in and out.

(Obviously this is not directed to Dietz, who it's safe to say would know that. :) )

Here's what the International Association of Sound and Audiovisual Archives (IASA) has to say about it:

Although the higher sampling rates encode audio outside of the human hearing range, the net effect of higher sampling rate and conversion technology improves the audio quality within the ideal range of human hearing.

As a side note, Fourier Analysis offers a very comprehensive representation of wave-like phenomena, but it's not the full picture. You should check out Wavelet Analysis.
 
Given the pricing differences between Tidal and Apple Music, I don't know why anyone would opt into a Tidal subscription.
Yeah, I dropped it as soon as my last online class ended.

Yeah, Dan Worrall's video is great... A major a caveat to his whole argument is the "temporal resolution" of the sounds being recorded and/or processed. I have posted this earlier, but here's what the International Association of Sound and Audiovisual Archives (IASA) has to say about it:
You know, it's a lot like RAW files in Photography, I always urge people to shoot RAW of course but don't discard the RAW files after you process. The future might bring a way to go back and re-develop the images utilizing all of the 14-bit data in ways we can't now. Same thing with audio, and I keep hopping back and forth, but your IASA quote made me realize that if we have the PC/Mac power, do higher resolution projects when available. 24/48 may be the standard deliverable to a film studio in 2021 but what about the future? Methinks I need to heed my own photography advice, pertaining to audio.

Maddening. Between sample rate decisions and if I should upgrade EWHO to Opus while it's on sale has me totally sidelined. :roflmao:
 
It does sound good but for $24(?) a month it’s not worth it at all, Apple Music just integrated full lossless and that’s only $10 a month, much bigger catalog of music and all of it is lossless, they really put the nail in the coffin for Tidal
Bingo. Our Berklee student price was $9.99/month for Tidal and it was only for the 9-month certificate program I enrolled in, plus the song analyses (these were critical listening and advanced mixing classes) were from a specific list picked by the instructors, and everything was on Tidal, so I went for it for the time. And you had better like hip-hop because the landing page always promoted all that first. I get it, the fans will more go for that than Alan Silvestri's greatest hits.
 
Yeah, I dropped it as soon as my last online class ended.


You know, it's a lot like RAW files in Photography, I always urge people to shoot RAW of course but don't discard the RAW files after you process. The future might bring a way to go back and re-develop the images utilizing all of the 14-bit data in ways we can't now. Same thing with audio, and I keep hopping back and forth, but your IASA quote made me realize that if we have the PC/Mac power, do higher resolution projects when available. 24/48 may be the standard deliverable to a film studio in 2021 but what about the future? Methinks I need to heed my own photography advice, pertaining to audio.

Maddening. Between sample rate decisions and if I should upgrade EWHO to Opus while it's on sale has me totally sidelined. :roflmao:
Not living in technotopia has its own downsides, but at least we are not living in a cyberpunk hellscape. Seriously though, for musicians and sound artists I think questions like these are mostly contextual... There's no universal standard to be applied. As Dan Worrall is keen on saying: "If it sounds good, it is good."
 
The only thing it will change is that any additional processing on the audio with higher sample rate will have more points to work with -- but technically at that point, any additional processing is distorting the source audio (for good or ill).
Again - to be absolutely precise (and totally obnoxious) - there are no "points," just higher sine waves.

As a side note, Fourier Analysis offers a very comprehensive representation of wave-like phenomena, but it's not the full picture. You should check out Wavelet Analysis.
Again again, picture a speaker cone. It moves out and in, back and forth, no "points" in between to tell it where to move.

And I'm not talking about the waveform representation, I'm talking about how sampling works.

Also, I just searched for "wavelet analysis"... well, unfortunately I was an idiot when I was a teenager (and well beyond). Instead of going to algebra and geometry classes, I ditched and smoked joints with other bad boys.
 
Using 96k only makes sense for internal processing, not as final output for a mockup.

For example, you want to slow down something you've recorded. Yeah having double (or triple) the recorded samples per second will make a huge difference in detail here.

An argument could be made that 96k is better for the final mastering but that doesn't make much sense for a mockup IMO. Honestly, nobody will be able to hear the difference between 48k and 96k. I seriously doubt anyone will be able to even hear the difference between 44.1k and 96k.
 
Using 96k only makes sense for internal processing, not as final output for a mockup.

For example, you want to slow down something you've recorded. Yeah having double (or triple) the recorded samples per second will make a huge difference in detail here.

An argument could be made that 96k is better for the final mastering but that doesn't make much sense for a mockup IMO. Honestly, nobody will be able to hear the difference between 48k and 96k. I seriously doubt anyone will be able to even hear the difference between 44.1k and 96k.
Yeah, I agree. Given that most professional sample libraries are at 48kHz or 44.1kHz, there's no point in bouncing an orchestral mockup at 96kHz or higher.
 
As Dan Worrall is keen on saying: "If it sounds good, it is good."
Honestly, we need to print that on a P-Touch and stick it along the bottom edge of our monitor, above the keys on our controller and every box of Fruit Loops in the pantry.
 
Hello,

Not to bore you but I am an electronic music producer since the 90's with various releases under my belt. Just to give you an idea of my musical experience.

Recently, I am working with like minded individuals and we are stepping into the composition field.

We have noticed remarkable changes in the sound of our exports at 192khz compared to 44.1 using mostly synths and plugins, with the occasional library added.

Give it a try and see what you think. We also understand the industry works at 48khz and we appreciate the tone of that sample rate as well.

Lastly, look for a site called Qobuz and take a listen to their 192khz classical collection. You may be pleasantly surprised.

Enjoy!
 
Last edited:
Alan Meyerson said there’s basically no audible difference between 48 and anything above that, and I trust Alan Meyerson above pretty much everyone. Only reason to ever go above 48 is so you can be of the cool kids with a “Hi Res Lossless” sticker on your Apple Music releases
 
Top Bottom