I have the feeling that some here can't tell when you're kidding. You?

Only on the Scratch Tracks...
Chris

Lot’s of perception in here, Time to recycle https://youtu.be/-jCwIsT0X8M

Great video! Illustrates a point that I tried to make that got derailed and turned into bickering and sourness.

^^ cool video. I recall watching it in the past within a different context.

I got bad ears but even I could hear a difference between 44.1 without oversampling and 48 with over sampling. i hear it in the distortion and low end.

I am starting to wonder, maybe it makes sense using higher sample rates for particular types of electronic music being made with certain VSTs instruments and FX which utilize frequencies and harmonics that are normally not used in more traditional styles of music like rock, gospel, R&B, e.t.c..

one thing this video made me think is -- the fab filter guy is focused on high quality sound and clarity -- but the actual music itself he is demoing is not very good. higher clarity and modern tools does not always equate to better artistic sound?

sort of like comparing this modern portrait done with modern tools to this older portrait done with older tools? painting.

they are not comparable art forms though they portray similar objects and have similar goals. the same thing could be said of the imperfect older tools artists like Apex Twin and NIN used 20+ year ago to today's more modern and perfect tools? The new tools and higher clarity does not necessarily add higher artistic quality.

dont know that made sense.

Have you ever used dither or noise shaping? Not only does dithered 16 bit not sound identical to the 24 bit source, but every version of these processes sounds different. It's not the magic bullet you think it is.

Upsampling to 384khz is actually a real thing that does change the sound.

Yes that certainly makes sense to me. I’m not a guitar guy so his music choice does not really resonate with me, however the principles remain the same.
There are plenty of reasons why something that doesn’t measure better, might sound better.

Toneboosters even have got a special plugin for this https://www.toneboosters.com/tb_bitjuggler_v1.html

thanks for the link. interesting plugins.

In the FF video, at around 26:00 he mentions good quality transparent over sampling. How does one know or measure that there is a good quality oversampling filter in a plugin? Do you look at the signal before checking off oversampling and compare it to after checking off oversampling (or HQ) and look for strange frequencies or artifacts?

Good question! My honest answer? I trust my monitoring, if it sounds good it is good. I guess I'm old skool, applying some principles of bits getting a bit funky dancing around 0dB levels and not doing anything to drastically in the digital domain, but I'm not putting SPAN or any other metering plugin after every processing plugin to see wat the sound looks like. There are plugins I like and plugins that don't gell with me. For me the end result is what counts. Maybe not so scientific, but life is to short to become autistic on these topics. IMHO we're in the game of creation, not metering. But metering what you don't hear, for the sake of it, is something I trust developers to do.
And if a plugin measures like sh#t but sounds great? I couldn't give a toss and still use it.

I accept that you believe you hear a change by upsampling to 384khz, but the general consensus in research says it can't be.

Here's one example:
https://www.realhd-audio.com/?p=7143

As this is supposed to be a thread about perceptual audio preference, and because these threads often progress in the same direction, I'm not willing to go deep into the weeds. However, I trust this comprehensive meta analysis more than any results from Waldrep who is an ideologue. https://www.aes.org/tmpFiles/elib/20201116/18296.pdf

As for upsampling, the simple logic is that converters operate differently at different sample rates because of their filter bandwidth and the design specific to that sample rate. Many use much slower rolloff at 192khz and above because less alias suppression is required.

Waldrep's paper is entirely about perception. I don't know what your problem is with Waldrep, other than he doesn't agree with you. He's a guy who recorded at high rates and depths for a very long time, built a business on the concept, because he believed it mattered. Then he researched it. Something wrong with that?

I will agree to look at the paper you referenced, however.
I understand the logic, it makes sense to a point, then fails. But since your thread is about perception, this part of discussion seems to no long qualify.

I think 24/48 sounds a bit more 'relaxed' than 24/44.1. But I've listened to plenty of cd's at 44.1 that sounds frankly astounding so who knows.

It could be my imagination, but 96kHz cymbal crashes sound cleaner played back at 96, despite the Nyquiest theorem. Could just be that my system seems to output slightly louder at 48/96/192 vs 44.1.

In any case, of course most outlets downsample to 44.1 anyway, and plugins seem to all have oversampling/hi-quality mode. And for youtube uploads, somebody ran a test and got better results with 44.1 vs 48 which youtube downsamples to 44.1.

It would be nice if there was standardization btw the video and audio worlds, on 44.1 vs. 48.

For long time 24bit/44.1k. Since I get better PC and interface I use 24bit/48k. For best results it should be around 60k. But we don't have that sample rate as a standard in many hardware drivers, so the "logical" should be 88.2k. And probably it's an optimal "standard" sample rate you should use. Same with 96k (not mention higher sample rates but it's overkill for any computer today I guess). BUT...
1. Hardware drastically reduce number of inputs on preamps on higher sample rates.
2. Your computer works harder because of that.
3. Limited number of tracks. With 44.1 or 48 you can sometimes put more plugins and tracks on your project.
I would really use 88.2 or 96k and higher but it's impossible right now. Some projects just crash because my CPU dry out of power quick on higher sample rates and I can't finish my mixes. And it's still pointless to make records higher than 48k until most of streaming services, YT, etc. start using higher sample rates audio on their platforms. Maybe 10-15 tracks with a little number of plugins would be ok on higher than 48k but most of today projects are 50-100 or even more tracks so it's impossible to work on high sample rates with that amount of tracks.