PlanetZ forums - scopeusers.com

Anyone in here use 96 KHz as your standard sample rate?
How many tracks can you get with plugins?
What CPU are you using?

Yes. I've been using 96KHz since Jan 2015. I run an i7-4790 (not overlocked) and my biggest mix is over 80 tracks. Its not the number of tracks but the number and type of plug-ins (Rack Extensions) that would eventually overload the system. Running XITE-1D, Pulsar 2 and 2 x Luna 2 - all at 96KHz on same system.

Only on 1 or 2 occasions have I maxed out the host DAW (Reason). I don't have any need pre-render RE synths to audio.

More details of what I run on Scope here : http://www.hitfoundry.com/issue_27/rst_mast.htm

It depends entirely on your plugin load, whether native or VST. With over sampled plugins the increased load will be less, assuming they respond to the higher sample rate appropriately (4X oversampling @ 48Khz =2X @ 96 for instance). The reduced latency can be nice for the live recording side of things in the native realm, but also comes at a cost as the whole system has to be that much more responsive under heavy loads.

The most important question would be to ask those of you using 96 KHz, if it is worth the effort... Do you see an important difference, which is it and how do you take advantage of it at the end of the process once you get the material back to the leastening stations?

I would say its definitely worth it. Especially if you can rationalise your workflow in reasonable ways such that your system can handle it without bouncing tracks. Even though reduced to MP3 theres a higher resolution in the mixing stage which seems to carry though.

Also if you maintain a stems archive @ 96khz, when you have future systems having abundance of capacity you may regret that all your back catalog is only at 48Khz.

That's cool, thank you for your feedback. For the moment, I cannot do it without some problems. The bottleneck, I think, it is in my old Pulsars I cards, not the PC.

Yes I would say if you can stretch your wallet to get at least a second hand pulsar II it would make a big difference

That is good to know. I'm switching over to Dante with the Focusrite Rednet PCIe Dante card. I believe this will give me the lowest latency. It should give me better performance at 96 KHz if I'm not mistaken. I might even be able to keep my old computer with that. I'll test it out with a bunch of high end plugins. It's an old i7 computer but has 4 cores. I have a friend who is an engineer and he does everything at 96KHz at home.

I would certainly be happy to get into 96 KHz if I could, I love great sound, but it is not only Scope you need to upgrade then, you need more Ram too and a faster HHD as well, it’s a whole move I guess.

To be honest, I am happy with what I can do. Nevertheless, lately, I have been spending quite a lot of money in new instruments to cover new jobs, so this one will have to wait for me

Not faster HDD or more RAM, less bottlenecks and more responsive system. DPC latency calls start to have more impact, you might buy an SSD only to find that it has certain latency issues on writing more than a certain amount of data, RAM can help if you’re for some reason swapping to disk...

Doesn't it increase the size of the audio file data significantly too?

yayajohn wrote: ↑Mon Apr 23, 2018 8:16 am Doesn't it increase the size of the audio file data significantly too?

Yeah but you know, storage is cheap these days.

valis wrote: ↑Mon Apr 23, 2018 8:14 am Not faster HDD or more RAM, less bottlenecks and more responsive system.

But..., how do you achieve it without upgrading? I don't see it. Less bottlenecks means upgrading, doesn't it?

I would think if you're using a virtual sampler with custom samples it would nee more RAM.

Nestor wrote: ↑Mon Apr 23, 2018 10:06 am

valis wrote: ↑Mon Apr 23, 2018 8:14 am Not faster HDD or more RAM, less bottlenecks and more responsive system.

But..., how do you achieve it without upgrading? I don't see it. Less bottlenecks means upgrading, doesn't it?

Well that would depend on the system you currently have, wouldn't it? Bottlenecks may come in the form of system drivers or even the system BIOS, different boards based off the same exact Intel chipsets can have wildly different DPC latencies even with a modern board. It may also be the result of choosing a cpu that offers less lanes with your specific board configuration than another cpu, and so that combined with your choice of additional peripherals may cause resource sharing. These things can be improved and/or mitigated by looking at the system as it currently is and attempting to change drivers, try different BIOS revisions, consider perhaps *only* a different CPU and or motherboard and so on. Or perhaps that early Core2 era system really does need to go...

braincell wrote: ↑Mon Apr 23, 2018 10:22 am I would think if you're using a virtual sampler with custom samples it would nee more RAM.

The sampler's programs may or may not load different samples if you change samplerates.

This whole discussion has been done here before, but it's important to look first at why samplerates were a thing to consider in the early days of the digital workflow. The first digital recording devices worked at 48khz so that they could shift the combined artifacts of the decimation filter (which removes frequencies above what the ADC can handle) above the intended destination format's ability to reproduce. In those days that was primary consumer cassette and CD formats. This avoided all of the same ADC's from adding a 'ring' to the top end of the spectrum of a recording, especially since in that era most if not all of your sources came from an analog source or microphone.

Then digital processes and mixers were added to the workflow, starting with the introduction of things like the first Sony digital mixer (which the developer of the Sony Oxford/Sonnox plugins has discussed at length on Gearslutz, if you care to search). In that case, again higher samplerates pushed artifacts up higher into the spectrum where they were either less audible, or wouldn't make it into destination formats at all.

When it comes to the digital processes, this is arguably of more concern to us now. Ie, even the cheaper soundcards we have available are markedly better than what was on offer in the mod to late 80's and early 90's. So unless you're summing 192 channels of orchestral microphones in the digital realm the combined ring of multiple tracks of ADC's are of little concern. So, the artifacts of our digital processes come from a variety of things: filters and eq's and the particular implementation you face with a specific design, compressors and limiters, saturation and waveshaping plugins and so on. In the case of many algorithms, internal upsampling to work at a higher samplerate can push the aliasing and other artifacts out of the audible band range without needing to work at 96khz for your whole workflow, and so the benefits of higher samplerates can be had with normal working processes.

Other plugins may benefit highly from being used in a higher samplerate, case in point for our tools we have Vectron. I haven't checked this in Scope v7, but in 4.5 & 5.1 there is a marked difference in audio quality at 96khz. This may not always be what you want however! Sometimes thickening, especially for basses and sounds with the top end rolled off, may actually be more pleasing.

And then we have our modern workflow of mostly plugins, individually recorded passes of instruments, and braincell's example of samplers. Here we find the same thing, some of the onboard functions of a sampler or virtual instrument (oscillators, filtering, dynamics, waveshaping/distortion etc) may or may not be implemented in such a way that immediate gains come from higher samplerates. Also in the case of simply resampling the pitch of a sample, an algorithm may or may not benefit. If a sampleset comes with a variety of source samples to match your working samplerate, at what point are you playing enough notes to make a difference? Etc...

In no way am I saying I think higher samplerates aren't worthwhile by the way, but there are specific points in the mixing and sound design workflow where the gains may or may not be had, and that's how these gains may or may not be had overall (since the digital artifacts, or lack thereof, will be cumulative).

It was a great explanation, thank you Valis! It is a complex area indeed all the pro and cons to take into consideration. What I like about your approach is that instate of just saying higher is better, you go to investigate all the elements in the interaction game within the system, that makes sense to me.

Folks, you there using 96 KHz, if you were to give a percentage about the subjective perception of the sound achieved, what would you say? I will be clearer with a precise question: How much better to your ears do you think the jump between 48 and 96 is? Is it 10%, 20%, 40%, 60%? Try to be as precise as you can please.

Nestor wrote: ↑Mon Apr 23, 2018 5:47 pm
Folks, you there using 96 KHz, if you were to give a percentage about the subjective perception of the sound achieved, what would you say? I will be clearer with a precise question: How much better to your ears do you think the jump between 48 and 96 is? Is it 10%, 20%, 40%, 60%? Try to be as precise as you can please.

For some (not all) plugins, 96KHz makes a clearly audible difference. I don't know this but that is what I have been told. For pitch shifting, the higher the better. It's nice to have that option if you are a weird sound designer as I am.

Nestor wrote: ↑Mon Apr 23, 2018 5:47 pmFolks, you there using 96 KHz, if you were to give a percentage about the subjective perception of the sound achieved, what would you say? I will be clearer with a precise question: How much better to your ears do you think the jump between 48 and 96 is? Is it 10%, 20%, 40%, 60%? Try to be as precise as you can please.

I would say 5% in what I do. There really isn't a lot in it. It took a lot of mixes before I got accustomed to that difference to the point where I didn't want to go back to 48Khz. It probably wont be obvious at first, depending on your age. I'm old but I still like it (96Khz).

Thank you guys, your answers are VERY relevant to solve this mystery, if you allow me to describe a KHz upgrade like this

Yes, you have given me the answer I was looking for. Sometimes the right question gives you the right answer, but you have to be lucky enough to find the right question in the first place, which is not always easy. And yes, a perfect question it is SOOO important in life, so you end up in good land!

I think I have poured onto you THE key question here, which is about “the percentage of quality you can achieve going from 48 to 96 KHz”.

My answer: I will stay just where I am, definitely, for two reasons:

First, I really like the sound achieved already, very, very much so, and clients like it too, which is very important for me of course, so I’m fine, it is not “a real need”, it is rather a pleasure let say the search for that perfect sharpness, but not a need. It is like going from an Epiphone Les Paul to a Gibson Les Paul with a few thousand dollars in price difference to, "perhaps", notice now and then a small nuance in the tone change for better... I would definetely go for an Epiphone in that case, be happy and satisfied with that.

Second, for only a 5% of improvement in sound quality when the sound achieved is already totally satisfying for me, I don’t think it is worth it to spend a chunk of money and what scares me, to have to go through the process of messing up with my stable system with the stress of possibly screwing things up without return, having to reinstall everything.

Well, I am in a way happy with your answer, because this means you can most probably achieve similar results either ways, and in the other hand I have to say I was expecting, actually, at its very least a 20% of improvement, and then I could start thinking about it. If you tell me there is about 50 to 60% improvement, the market would push you to upgrade, willingly or unwillingly.

You have it? That’s great, keep it! I wish I could have it too. I mean, I don’t want to discourage you guys for having this possibility, it is just not for me, have fun with it please. I prefer to go with more RAM, for instance, being able to load even more Kontakt instruments, or spend my hard earned money in important instruments that will give me more versatility.

Thank you for your answers brothers