NAM and 96k ?

[BenIfin]

Hey all !

Am putting together a nice speedy PC Laptop with Gigperformer 5 and a F/Rite Solo 4th Gen Audio i/o and Win 11 Pro

All running super good - NDSP Nolly, Scuffham, Genome and NAM.

I've emailed NDSP, Scuffham and T/N-Genome and they have confirmed that their software is "coded" to work at full 96k [and up] "resolution" - please pardon my non-techo language.

The hardware is running rock-solid at 96k / 32 Samples / RTL of 3.4ms under full load and plenty of stress testing with all Audio apps loaded,

My Question ?

Unlike NDSP Nolly, Scuffham, Genome at 96k .... NAM sounds exactly the same when the system is running at 48k or 96k ? - and I mean exactly the same

Is NAM coded only to internally run at a max. 48k resolution ?

Thanks :)

 

[2dor]

Hey all !

Am putting together a nice speedy PC Laptop with Gigperformer 5 and a F/Rite Solo 4th Gen Audio i/o and Win 11 Pro

All running super good - NDSP Nolly, Scuffham, Genome and NAM.

I've emailed NDSP, Scuffham and T/N-Genome and they have confirmed that their software is "coded" to work at full 96k [and up] "resolution" - please pardon my non-techo language.

The hardware is running rock-solid at 96k / 32 Samples / RTL of 3.4ms under full load and plenty of stress testing with all Audio apps loaded,

My Question ?

Unlike NDSP Nolly, Scuffham, Genome at 96k .... NAM sounds exactly the same when the system is running at 48k or 96k ? - and I mean exactly the same

Is NAM coded only to internally run at a max. 48k resolution ?

Thanks :)

NAM profiles can be trained at 96KHz as well. I did a few of those but training takes considerably more time.

The NAM plugin can adjust to the sample rate of the host (even if the profile is created / trained at another sample rate).

 

[BenIfin]

The NAM plugin can adjust to the sample rate of the host (even if the profile is created / trained at another sample rate).

Thanks :)

Am using some Amalgam Packs which are all done at 48k.

So .... just to clarify as this is a bit beyond my techy level :)

a) the NAM plugin, when hosted in a 96k host / environment will "play" these 48k Amalgam Captures but they will sound no different to if they were being "played" in a 48k host / environment ?

b) whereas when running NDSP Nolly, Scuffham and Genome in a 96k host / environment, they will "play" at their full 96k resolution and therefore (i) use a boatload more CPU but (ii) sound "marginally better" than if they were being "played" in a 48k host / environment ?

Got this right ?

Thanks !

 

[2dor]

Thanks :)

Am using some Amalgam Packs which are all done at 48k.

So .... just to clarify as this is a bit beyond my techy level :)

a) the NAM plugin, when hosted in a 96k host / environment will "play" these 48k Amalgam Captures but they will sound no different to if they were being "played" in a 48k host / environment ?

b) whereas when running NDSP Nolly, Scuffham and Genome in a 96k host / environment, they will "play" at their full 96k resolution and therefore (i) use a boatload more CPU but (ii) sound "marginally better" than if they were being "played" in a 48k host / environment ?

Got this right ?

Thanks !

a) correct - they will sound the same. I have not been able to make out a difference when trying out with my system set for 96K

b) not sure what the ask here is to be honest; think point a) sums it up. Now, if you load up NAM profiles trained at 96KHz then you may notice some differences but it depends on a lot of other things too (ESR etc).

 

[MirrorProfiles]

There’s no real guarantee for things sounding better just because they’re at 96k. If you like it at 96k crack on. Plugins coded at 44.1 or 48k usually upsample if there is a benefit to it and the main downside would be how the filtering and resampling processes are implemented. You can just decide where you’re happy.

IIRC Amplitube at high sample rates sounds quite different but generally speaking, most amp sims sound about the same no matter what sample rate you’re at (because they oversample anyway)

 

[Chocol8]

What are you expecting to get out of 96k? Other than not needing as much oversampling/anti-aliasing I am not sure you would gain anything.

 

[Sascha Franck]

What are you expecting to get out of 96k? Other than not needing as much oversampling/anti-aliasing I am not sure you would gain anything.

That'd be my question as well.

Even with these days fast computers, running a complexed setup demanding plenty of "live thread" (as opposed to playback tasks which some DAWs manage to use higher buffersizes for) performance at 96kHz is a good receipe to bring pretty much any machine down to its knees.

I could understand it if lowest latencies would be the main concern - but I can get as low as 3.5ms under 44.1 already.

But then, in case you're not running into issues, why not?

 

[MirrorProfiles]

hard drive space vanishes rapidly at 96k too, and most playback mediums are going to downsample anyway.

I actually enjoy working at 96k but it just uses so much more disk space that I don’t find the slight trade off to be worth it. I do recommend everyone to try it for themselves though, I think having a personal preference for any sample rate is valid. But it’s personal rather than an objective better/worse.

 

[Sascha Franck]

Fwiw, if doubling the samplerate would actually result in half the latency, I might consider using 96k for a live rig (file sizes are irrelevant for that as nothing signigicantly gets recorded) in case the system provides enough *oomph*. Such low latencies could then allow for, say, setting up additional analog loops without latency getting too high.
But then, usually latency isn't getting that much lower as many interfaces raise their internal safety buffers for being able to work properly at such low samplerates. Add to this, that in case you're not running the top of the crop interfaces (in terms of latency) already, it'd possibly be a better trade off to grab one of those interfaces providing very low latencies (such as RME when money is no objection and Motu in case you're on a tighter budget).

Also, I have been fooling around with some testing scenarios regarding what I'd need for a live rig. And I managed to almost max my Macbook M3 out with all the "nice to have just in case" things - which, for ideal performance (gapless switching and delay/reverb spillover), would always have to be loaded in the background. Granted, I didn't try with Gig Performer (which is said to be extremely efficient and great at multithreading whenever possible), but at one point even the most efficient host doesn't help anymore.
Anyhow, my (admittedly rather brief) testruns didn't scream "oh yes, let's run everything at 96k now!"

And then, as far as anything regarding sound quality goes - if this is really meant to be a live setup (is it?), I absolutely doubt anyone would be able to notice the differences between, say, 48 and 96 at all.

 

[JiveTurkey]

Disable the cab block and get your tone clarity back