Synth Quality in Scope @96khz - especially ModularIII

hifiboom · Post by **hifiboom** » Sun Apr 01, 2007 4:21 pm

just checked the differences in scope at 96khz....

I know that the improvement in mixing with a higher sampling rate is minimal, I was interested, if there is much improvement with the synth....

So I checked the modular stuff, build a Detuned SAW patch, and I have to say that the improvment was very large.
Especially the many different CW filters reached a complete new level !
They do no more sound that aggressive in the high end area.

The oscillators also improved slightly and get a more creamy sound.

Generally its really worth it with the synth....
Scope sounds superior compared to VST stuff at 44,1 also,
but @96 khz modularIII sound is very close to analog....
sampling down the result after recording seems to not affect that better sound...

But you really get very low voices amount out of it . So its not really worth it to work at that sampling rate.
The next problem is that all the high class synths ( minimax and co) do not work anymore.

So interpolating antialiasing filters would improve the sound of mod enormiously ! At the moment the more aggressive a filter is affecting the sound (heavy resonance) the more its degrading the sound.

BTW the CW filters at 96khz are enormiously good sounding.... They really grab into the sound.

At least a new Scope should run internally always at 96khz, if not even higher...

valis · Post by **valis** » Sun Apr 01, 2007 4:40 pm

The high class synths are upsamping internally for their filters & Oscs, hence the 'better' sound quality. This is fairly common in Vsti's these days too. What would be nice is if newer CW devices had a setting so u could control the amount, that way if you prefer to work in 96k for using other devices at the same time you could switch Minimax back down to 1x sampling internally to allow it to function. Of course 2x internal samplerate at 96k (192khz) would be even better, but I suspect we'd need more than just a few bug fixes to make that worthwhile (next gen hardware for the dsp power to do more than 1 voice).

hifiboom · Post by **hifiboom** » Sun Apr 01, 2007 4:46 pm

true valis,

one voice isn`t what we want at all

But I`m sure 2 TigerSharcs or something like that handle that pretty good.

Q:
do the new upcoming Adern filters do some kind of upsampling, to improve the sound... like minimax/prodyssey filters? ?

Immanuel · Post by **Immanuel** » Sun Apr 01, 2007 5:03 pm

The Masterverbs Pro (and maybe other masterverbs too) also sound better at 96KHz.

Shayne White · Post by **Shayne White** » Sun Apr 01, 2007 9:40 pm

Vectron sounds WAY better at 96K.

Shayne

FrancisHarmany · Post by **FrancisHarmany** » Mon Apr 02, 2007 7:03 am

OK... I never tried it even.... Does it make a big difference when (vdat) recording also ?
I could do all my sampling in 96khz!

astroman · Post by **astroman** » Mon Apr 02, 2007 10:16 am

the aliasing of a 44.1k signal falls into the audible range of the spectrum, simplyfied there are additional frequencies (supposed to be supressed by a special filter) that interact with the 'correct' tone - as there's no such thing as the 'perfect' filter.
The tone gets a bit rougher and slightly blurred by the interaction.

In 96k the same applies, but beyond the range of human listening, so the tone is perceived as more transparent with (particulary) improved highs.

For my personal taste a perfect (synthetic) 96k track would be entirely boring and fatiguing with all it's 'beauty' sounds - it's like eating all icing without a cake

The roughness of an original Waldorf Oscillator or the DX7's 10bit converter add a lot to the sound - to 'improve' the Vectron by stealing it's aliasing is not at all my cup of tea. But it's a matter of taste.

What counts is the final result - I know more than enough perfect mixes that do just one thing - they s*ck and spread boredom

cheers, Tom

ARCADIOS · Post by **ARCADIOS** » Mon Apr 02, 2007 1:15 pm

i agree with astroman about synths.
i used to try to put them all to 96khz.
now i use 44.1
it is easier for scope to operate and also it has that kind of taste that i like listening to.

there is an issue though if we're talking about mixing and mastering.
this, i do not know yet what is better. it depends offcourse on what it the recording machine you use. and more.......

a yes the company's name: SCREAMWARE.

voidar · Post by **voidar** » Thu Apr 10, 2008 4:57 pm

Record everything at 96K (even from the ADC), then r8brain Pro them (MinPhase/Ultra Steep) down to 48K (or 44.1K if you prefere).

You won't hear a difference. Possibly only the pleasing touch of the minimum phase distortion filtering.

This is a valid point when using the A16U (or similar/lower quality device). r8brains AA-filtering will surely be much better.

At 96K, your ADC won't need such a steep filter, which are hard and expensive to implement in hardware.

astroman · Post by **astroman** » Fri Apr 11, 2008 3:50 am

just to get this straight ...
you record at 96k and for whatever reason it sounds better than at a lower rate
then you resample down via the Voxengo thingy and save the improvement into the low rate domain - due to the careful filter design of the software
correctly received ?

cheers, Tom

voidar · Post by **voidar** » Fri Apr 11, 2008 4:47 am

Yup.

That is, if you feel your converters work better at 96K.

astroman · Post by **astroman** » Fri Apr 11, 2008 7:10 am

hmmm - I dunno - the alias filtering might apply, tho

thanks anyway, Tom
and congrats to the first kilo-post

voidar · Post by **voidar** » Fri Apr 11, 2008 9:05 am

Point is, r8brain has very good filtering .. ! Close to 150dB S/N

astroman · Post by **astroman** » Fri Apr 11, 2008 1:11 pm

maybe in that case Stephen Hummel aka Wavelength can return for a short visit and make the Vectron oscillators harsh again