12 cores not being used equally (cpu errors)

[Darryl Ramm]

Originally Posted by metalmixer

I am working on a project now all @ 192K I am even using a 2018 MBP i9 on Mojave (not even supported) - and the CPU is killing it for me no spike issues and have not even had to freeze any track (yet). I did a blind test with one of the band members and he picked the 192 bounce and I even just sent him MP3's he had no idea what the difference was and he picked it right away.

Cool You night want to let folks know what plugins you are using and how much CPU use you are seeing just for a comparison. My 2017 i7 MBP rocks as well (but I don't work at 192 KHz), I am pretty amazed how much that laptop delivers.

[Eric Lambert]

Originally Posted by killah_trakz

I use VEP to as a plugin host for acustica stuff mainly. And it’s NOT only for instruments.

True, it does more than just load instruments, which is why I'd said that it's best suited for VIs; Vienna VIs in particular -- the products it was written for.

[Eric Lambert]

Originally Posted by metalmixer

I am working on a project now all @ 192K I am even using a 2018 MBP i9 on Mojave (not even supported) - and the CPU is killing it for me no spike issues and have not even had to freeze any track (yet). I did a blind test with one of the band members and he picked the 192 bounce and I even just sent him MP3's he had no idea what the difference was and he picked it right away.

He picked the 192 version as it was being compared to what other sample rate?

There are many factors that would affect the comparison so it's tough, based on what's mentioned above, to know what was being heard.

[Darryl Ramm]

Originally Posted by Emi

I’ve read those papers as they came out...

You guys are right, but you also keep staying at the same problem (nequyst filter and antialiasing filter) and it’s true, very good converter at 44 can sound even better than other at higher sample rate.

But you miss the point that at higher sample rates there is more resolution in the 20-20k range, not above (I’m not a bat). Think of it like a 30 FPS vídeo compared to 60 FPS. 60 FPS has double the slices. Higher sample rates is the same.

If you look for a reason not to go high is noise. Above 20k there is noise generated from the AD’s. That is because the anti-aliasing filter is positioned higher. I would pair a good antialiasing filter started right above maybe 22k and high sample rate and it would a really be a great improvement in the digital audio world. All this last statement is a theory.

And also, as I already said, the cpu power suffers and also the stability, but hack, we have lots of power from cpu now. Let’s fix this isssue of how to use it at maximum within pro tools.

You really need to spend some time understanding digital signal processing.

A higher clock rate does *not* provide more "resolution" within the 20-20 KHz region. This is basic Nyquist theory stuff, the very foundation of digital audio signal processing. Explained in the video I linked to above. If there is more "resolution" in the signal that you get by sampling at higher frequencies, it's above the 20kHz range (keeping with your example).

This is absolutely fundamental, it's the very definition of conversion of an analog signal into digital when looked at in the frequency domain, extra "resolution" can't get where you are trying to claim it can, just mathematically not how it works. It's just mathematics: https://en.wikipedia.org/wiki/Nyquis...mpling_theorem

Sampling at a higher sample rate and then moving the antialias filter down to a lower frequency makes no sense to me. One of the main benefits of higher sampling rates is being able to run the antialiasing filter (in front of the DAC on input and after the ADC on playback) at higher frequencies and so avoid phase and other effects encroaching down into the audio band. That is arguably *the* main technical benefit claimed for 88kHz and above playback sample rates, and one that at least seems reasonable to me, and what I am assuming some 96KHz systems playing back sound *great* to me. Why would you want to undo that exact benefit?

Reducing ultrasonic issues can be mitigated by systems components designed to handle the ultrasonics and gentler filtering that needed by antialiasing filters. But as many experts point out the ultrasonic issues may make very high sample rate playback systems worse than low sample rates ones.

[arche3]

Originally Posted by Emi

I’ve read those papers as they came out...



You guys are right, but you also keep staying at the same problem (nequyst filter and antialiasing filter) and it’s true, very good converter at 44 can sound even better than other at higher sample rate.



But you miss the point that at higher sample rates there is more resolution in the 20-20k range, not above (I’m not a bat). Think of it like a 30 FPS vídeo compared to 60 FPS. 60 FPS has double the slices. Higher sample rates is the same.



If you look for a reason not to go high is noise. Above 20k there is noise generated from the AD’s. That is because the anti-aliasing filter is positioned higher. I would pair a good antialiasing filter started right above maybe 22k and high sample rate and it would a really be a great improvement in the digital audio world. All this last statement is a theory.



And also, as I already said, the cpu power suffers and also the stability, but hack, we have lots of power from cpu now. Let’s fix this isssue of how to use it at maximum within pro tools.

Leave the actual science to people who understand it. Lol..

Sent from my SM-G975U using Tapatalk

[JFreak]

Originally Posted by Emi

at higher sample rates there is more resolution in the 20-20k range, not above (I’m not a bat). Think of it like a 30 FPS vídeo compared to 60 FPS. 60 FPS has double the slices. Higher sample rates is the same.

No, again -- audio is not the same as video and you cannot compare the two like this. More information (fps) in video is compared to more information (Hz) in audio, not apples to oranges. Meaning if you store more information in video you get smoother playback, and if you store more information in audio you get higher frequencies. More information has nothing to do with resolution (as in the video, all of those extra frames can very well be just 50% gray and our eyes would not notice; and as in audio, all of those extra samples only enable representing higher frequencies which our ears would not notice).

But as far as the human hearing (20Hz to 20kHz) there is absolutely no more extra resolution no matter how fast you sample it -- EVERYONE in this planet (except you) have understood that you only need 40kHz sampling rate and "perfect AD/DA conversion" and you have a perfect digital representation of the signal on the human hearing frequency range. The reason for higher than 40kHz sampling rate lies in less than perfect AD/DA conversion and especially the less than perfect filters on the analog domain. When those imperfections are outside human hearing range, the signal sounds cleaner but there is no extra resolution even if you sampled at 1GHz rate.

Unless you're thinking there's something better than perfect, that is...

[JFreak]

Originally Posted by Emi

Ok, time for me to re-read everything. Thanks for taking the time and explain it to me.

No prob :)

[7Seas]

Originally Posted by killah_trakz

I use VEP to as a plugin host for acustica stuff mainly. And it’s NOT only for instruments.


Sent from my iPhone using Tapatalk

Hi.

Does it introduce any latency/delay that PT has trouble compensating? or is it perfectly transparent in that effect, just giving you the opportunity to run more acoustica plugs, much like the waves HW options?

THX.

[musicman691]

Originally Posted by 7Seas

Hi.

Does it introduce any latency/delay that PT has trouble compensating? or is it perfectly transparent in that effect, just giving you the opportunity to run more acoustica plugs, much like the waves HW options?

THX.

VEPro has it's own delay comp for plugins inside it. As far as PT is concerned I've never had any delay comp issues.

[7Seas]

Originally Posted by musicman691

VEPro has it's own delay comp for plugins inside it. As far as PT is concerned I've never had any delay comp issues.

Thanks