12 cores not being used equally (cpu errors)

[JFreak]

Originally Posted by stevedresser83

Nope the plugins work better at higher sample rates. Sample rate is basically the amount of snapshots per second. So 88k is twice as many snapshots per second. This allows the plugins to be more accurate in how they draw an eq boost or cut, or compress a signal, or whatever it may be.

Yes and no.

Yes some plugins sound better at higher sampling rates, but it has nothing to do with the actual sampling rate of your files -- but at lower sampling rates some plugins oversample before processing and then downsample back, and if the plugin just processes at 96k without changing sampling rates internally then it sounds more open and actually might not even use so much cpu.

Forget the snapshots per second thing, because we are not processing snapshots. "Snapshots" a.k.a. samples are only happening on disk and the more samples you have the more disk space you waste and more storage performance you need to be able to use them.

Instead, we are processing a digital signal, and for that you need to google "Nyquist Theorem". That is one of the fundamental truths of digital processing and by understanding that first you will notice that plugins are not processing snapshots after snapshots, but instead the mixer is re-creating the actual signal from your audio files for plugin to process and after everything is said and done the plugin gives away a different signal for next plugin/buss. And the next time you will see "snapshots" again is when you bounce and have another audio file.

[JFreak]

Originally Posted by the.engineer

This is curious, if you recorded at 44.1 and then converted to 88k, surely nothing will change as the audio is still 44.1. The missing samples can’t be added in by upsampling.

Missing high-end cannot be added, but the plugins processing may change their behaviour -- read my last two posts about upsampling/downsampling. If that is the case and you are heavily using such plugins, it is generally best to do the upsampling once and just mix in 96k to allow plugins to work better

[JFreak]

Originally Posted by Eric Lambert

with all the VIs (...) there's essentially no reason to travel above 96K and many reasons to stay at or below it.

Sorry to abuse your post but if the session is VI heavy it makes most sense to stay 48k because most VI's are sampled 48k anyway. As per my previous posts, the possible benefits of 96k come from other types of plugins, but not VI's.

[Eric Lambert]

Originally Posted by JFreak

Sorry to abuse your post but if the session is VI heavy it makes most sense to stay 48k because most VI's are sampled 48k anyway. As per my previous posts, the possible benefits of 96k come from other types of plugins, but not VI's.

What I was attempting to say is that I'm already pushing the boundaries of what my computer can handle at 48K due to the CPU-hungry processing of VIs and other essential plugins. It's not about sound quality. At 96 I'd have no chance of even opening these sessions.

[Eric Lambert]

Originally Posted by stevedresser83

Nope the plugins work better at higher sample rates. Sample rate is basically the amount of snapshots per second. So 88k is twice as many snapshots per second. This allows the plugins to be more accurate in how they draw an eq boost or cut, or compress a signal, or whatever it may be. Even when i imported a drum loop (recorded 44k) into a 192k session and did 1 single plugin boosting high end, that alone created a difference compared to the same drum loop, same eq boost at 44k. The proof is I bounced both, flipped the phase on one and i hear high end content when played together. Conclusion is that plugins manipulate the audio better at higher samples rates.

This assumes that more is better. In these instances, more can exhibit no benefits whatsoever and in fact be worse, just as we've been attempting to explain. A phase cancellation test doesn't reveal sonic superiority, it reveals discrepancies. Discrepancies come in many forms.

[arche3]

Originally Posted by Eric Lambert

This assumes that more is better. In these instances, more can exhibit no benefits whatsoever and in fact be worse, just as we've been attempting to explain. A phase cancellation test doesn't reveal sonic superiority, it reveals discrepancies. Discrepancies come in many forms.

Exactly.

Sent from my SM-G975U using Tapatalk

[stevedresser83]

Originally Posted by JFreak

Missing high-end cannot be added, but the plugins processing may change their behaviour -- read my last two posts about upsampling/downsampling. If that is the case and you are heavily using such plugins, it is generally best to do the upsampling once and just mix in 96k to allow plugins to work better

Appreciate the info J. Gonna study more on this.

The other part of this thread was the uneven usage of cores. I feel like my machine has 5x the power that is being used and getting bogged down way too easy. Anyone with any ideas on what can be done to better optimize i'm all ears.

[musicman691]

Originally Posted by stevedresser83

Appreciate the info J. Gonna study more on this.

The other part of this thread was the uneven usage of cores. I feel like my machine has 5x the power that is being used and getting bogged down way too easy. Anyone with any ideas on what can be done to better optimize i'm all ears.

Already told you:
If you run vi's split the load amongst several instances of said vi instead of loading everything into one instance. Also try running stuff inside VEPro 6 (soon to be 7). That will help balance out cpu usage. Sadly it's normal for PT to do what you're seeing and it doesn't matter what computer & operating system you're running - Windoze and OSX share the same issue.

[killah_trakz]

Acustica and protools are **** together. I have begged and pleaded for them to fix the issue on BOTH sides. And all I’m being told is “ REAPER “.

With VEP it does even out things but the routing and setup becomes daunting.

Anyways it is what it is. Just freeze and move on. Hopefully the new update will fix all that BS.


Sent from my iPhone using Tapatalk

[stevedresser83]

Not talking about production or running VI's. Just purely mixing, sessions with lots of plugins. All audio with plugins. I wonder if anyone has ever done a DAW comparison in how each daw utilizes cpu usage, cores, ram, etc