Some SD3 sample rate conversion and SD2 kit questions?

[Nip]

Hi
I am very satisfied with SD2 – but wonder if anything might be worth upgrading to SD3 engine due to SRC:

#1. Does SD3 allow selecting SRC algorithms to use for playback and rendering?
Thinking one algo for each, to decide for yourself about performance/cpu.

Some sample based engines allow that.

#2. Can you install or preconvert all samples to like 48k – so no rrealtime SRC is needed when running projects at 48k?
Thinking either conversions made by SD itself – or I can do it using a batch conversion of what I know is really high quality conversion.

#3. Can SD3 use SD2 stuff as they are?
I will not consider using 200 GB just for drum samples, very happy with SD2 stuff and reasonable size on it all preloaded in memory and all.

I have the Avatar, Custom & Vintage and Roots sticks.

Thank you for any response on this.

Topic

[John]

Hi,

I’ll answer as best I can.

#1 No, there’s no real-time sample rate conversion going on while you playback. The SRC happens when loading the samples into RAM, then they are held there until unloaded.

#2 Short answer: No.

#3 Yes and IMHO better. From my own experience I think SD3 loads the old SDX:s faster and I simply love all the new features I get in SD3.

BR,
John

John Rammelt - Toontrack
Technical Advisor

[Nip]

Thanks for a extensive and good response.

I got impression from Olle two years ago, when I reported a crash bug(2.4.1) – which was to do with init of sample rate conversion when loading a user preset. He said it takes a little extra cpu – but not so much.

Thereby ability to choose algo – but seems I got it wrong then.
https://www.toontrack.com/forum/superior-drummer-2-help/sd-2-4-1-instability-issue-loading-new-instance/#p150583

“When you run a sample rate other than 44.1kHz there are ‘resamplers’
at the outputs that convert from 44.1 to whatever is used in your
project. “

https://www.toontrack.com/forum/superior-drummer-2-daw/what-resampling-algorithms-are-used-by-sd-and-where-and-when/#p151864

“Regarding your questions, the adaptation to a different sample rate
in the host is done by resamplers at the outputs of S2. They
are always high quality and eat up a sizeable portion of a CPU
core if you use all outputs and a sample rate other than 44.1kHz.”

“We offer two version when
you pitch sounds, though, ‘Fix:ed and un-Fix:ed’. The Fix:ed
sounds are resampled in memory using that same high quality
algorithm while the un-Fix:ed sounds are interpolated on the
fly.”

Or is it different in SD2 and SD3?

[John]

Hi,

it was a sloppy answer I gave you, sorry about that. Even though I am labeled as “Technical Advisor” I’m not that technical, obviously. 😉
The principle would be the same but I imagine improved, since it has been several years in between developing the two plugins.
Tuning that is Processed (formerly Fixed Pitch), is done in RAM, though.

I hope Olle chimes in if there’s more to add (or to correct me).

BR,
John

John Rammelt - Toontrack
Technical Advisor

[Nip]

No worries. Thanks for your response.

It sounds like smart move to do like you suggested, if regular kit staying on original or some fixed tunings – are made at load time.
But maybe dynamic articulations are done in realtime, to make every hit a little different than previous, and need to be made in realtime or use a lot of memory duplicating with variations or similar.

[Olof Westman]

>Or is it different in SD2 and SD3?

SD3 and SD2 work the same in this regard.

SD3 has an alternative high quality pitch option for tuning toms, though.

Olof Westman - Toontrack
Coder

[Nip]

OK, thank you for response on this.

Running 7 stereo outs normally at 48, I have not had any problems or anything – with SD2.
Just collecting points for possible upgrade down the line.

[Author]

Replies