Cherry Audio's CA2600 is available NOW!

[abernathy]

Complaining about the sound quality of a plugin when you're running at 44.1 is like bitching about an amp because it sounds like shit on your $5 gas station earbuds.

Record in 96 or 48 (192 if you're feeling foxy), master down to 44.1 with a decent resampling algorithm afterwards. The 48 > 44.1 (or 96 > 88.1) resample is a slightly mathematically intense procedure (when it's done properly) that can eat up a fair amount of CPU cycles. If Cherry had gone that route people would be screaming because it was a CPU pig.

My bad sounding example is at 88.2 ... sorry, is that also stupid of me to complain about? And gosh thanks for pointing out that CDs sound terrible.

[WatchTheGuitar]

As said before sounds like they’re wrapping the Voltage Modular engine which runs internally at 48 kHz so 88.2, 176.4 whatever is probably going to cause rounding errors unless they recode to permit the engine to run at the same rate as the interface.

[gentleclockdivider]

Complaining about the sound quality of a plugin when you're running at 44.1 is like bitching about an amp because it sounds like shit on your $5 gas station earbuds.

Record in 96 or 48 (192 if you're feeling foxy), master down to 44.1 with a decent resampling algorithm afterwards. The 48 > 44.1 (or 96 > 88.1) resample is a slightly mathematically intense procedure (when it's done properly) that can eat up a fair amount of CPU cycles. If Cherry had gone that route people would be screaming because it was a CPU pig.

Do you run al your plugs at 192k ?:party: , you’de be surprised how many hardware synths run at 44k ( except my nord modular which runs at 96 )
I run all my vst’s at 44khz , reaktor 2* internally oversampled if needed but most stuff @44khz
If cherry has artifacts at 44k they should have mentioned that .
Reaktor blocks sounds just great at 44
This is imho a weak excuse ..it’s a S.R. conversion issue ,nothing more nothing less.

[Dallas_Cashin]

Complaining about the sound quality of a plugin when you're running at 44.1 is like bitching about an amp because it sounds like shit on your $5 gas station earbuds.

Record in 96 or 48 (192 if you're feeling foxy), master down to 44.1 with a decent resampling algorithm afterwards. The 48 > 44.1 (or 96 > 88.1) resample is a slightly mathematically intense procedure (when it's done properly) that can eat up a fair amount of CPU cycles. If Cherry had gone that route people would be screaming because it was a CPU pig.

My bad sounding example is at 88.2 ... sorry, is that also stupid of me to complain about?

There's really no point in going at 88.2k, so yeah, it's stupid, either go up to 96 or drop down to 48. Literally the only reason 88.2 is a thing is to make it easier to drop down to 44.1 on old hardware from twenty years ago. The math behind the 96/48 > 88.2 conversion is the same as 96/48 > 44.1, as you'd know if you read my post.

And gosh thanks for pointing out that CDs sound terrible.

CDs don't sound terrible. Going from 48 > 44.1, however, sounds terrible unless you invest a bit of CPU elbow-grease in proper dithering.

Do you run al your plugs at 192k ?:party: , you’de be surprised how many hardware synths run at 44k ( except my nord modular which runs at 96 )
I run all my vst’s at 44khz , reaktor 2* internally oversampled if needed but most stuff @44khz
If cherry has artifacts at 44k they should have mentioned that .
Reaktor blocks sounds just great at 44

192k is just stupid imo, 96k offers enough headroom, and is easily copacetic with the 192 sampling rate of my audio interface.

Sure, some hardware runs at 44.1, but running your computer at 44.1 to capture it is not just pointless, but doubly pointless - pointless because you're not capturing individual bits, you're capturing the audio stream, where sample rate is no longer rigid as soon as it hits those TRS jacks (you're taking ice pellets from the fridge dispenser, melting them into a glass of water, and then pouring that glass of water into your computer's icemaker); and pointless again because going up is much easier than going down.

It's easy to take things in a small box and put them in a bigger box (the nature of digital sampling provides a built-in analogy to those foam inserts that keep things from sliding around, too), it's going from bigger > smaller where things get tricky. With 192 > 96 > 48, or 88.2 > 44.1, it's mathematically simple, since (oversimplifying a little here) you're just cutting things in half, but "crossing the streams" between those two groups requires some fancier precision mathematics.

This is imho a weak excuse ..it’s a S.R. conversion issue ,nothing more nothing less.

Yes, it's a sample rate conversion issue, but not appreciating the math behind sample rate conversion doesn't make it a "weak excuse". Sonically-perfect-for-our-intents-and-purposes 96/48 > 44.1 can be achieved, but it's not automatic, and I don't think Cherry Audio deserves this amount of bile for figuring that most people were running their DAWs at 48/96. I'm sure the GUI probably has some issues at 256-color VGA, too.

Now, those stepped coarse-tune sliders? Those maybe deserve a little bit of bile

[Etienne1973]

Going from 48 > 44.1, however, sounds terrible unless you invest a bit of CPU elbow-grease in proper dithering.

*cough* ... dithering and sampling rate?

[chk071]

Complaining about the sound quality of a plugin when you're running at 44.1 is like bitching about an amp because it sounds like shit on your $5 gas station earbuds.

Record in 96 or 48 (192 if you're feeling foxy), master down to 44.1 with a decent resampling algorithm afterwards. The 48 > 44.1 (or 96 > 88.1) resample is a slightly mathematically intense procedure (when it's done properly) that can eat up a fair amount of CPU cycles. If Cherry had gone that route people would be screaming because it was a CPU pig.

Batman just entered the thread.

[gentleclockdivider]

I am trying to do simple amplitude modulation using the vca 2 in the voltage processing section .
The osc is modulated by an lfo , the lfo is routed to vca control , yet the osc always comes tr ue at full amplitude .
It seems that the lfo is bipolar and for amplitude modulation , the control signal needs to be unipolar ( positive energy only )
I tried different settings no luck
What am I doing wrong


Not having any luck either when using the ar envelope as default mod ( repeat trigger by clock out )

[WatchTheGuitar]

not at my computer right now, but if source is bipolar can you route it through that +10v line first and then into the vca to force the voltage to be only +ve?

into port labelled 6 then to 5/6 out

[dave dove]

could someone explain why i should run anything above 44.1KHz
if i'm going to master at 44.1???

[WatchTheGuitar]

could someone explain why i should run anything above 44.1KHz
if i'm going to master at 44.1???

because the 48 to 44.1 internal down conversion the CA2600 has to do currently causes more audible artifacts in real time than recording at 48 and then offline downconverting to 44.1 would. That’s why you should do it now while waiting to see if Cherry Audio come up with a better fix later on.

[gentleclockdivider]

not at my computer right now, but if source is bipolar can you route it through that +10v line first and then into the vca to force the voltage to be only +ve?

into port labelled 6 then to 5/6 out

Tried that , not working either


edit this is the solution

[Dallas_Cashin]

Going from 48 > 44.1, however, sounds terrible unless you invest a bit of CPU elbow-grease in proper dithering.

*cough* ... dithering and sampling rate?

Apologies - the editor I grew up on had a janktastic SRC, and the workaround was to force redither whenever switching sample rates (otherwise, as I understood it at the time, it'd basically just shove a bunch of dithered noise into, say, the 48000 band on a 96k file, then bring all that crap down to the audible spectrum when "half of 96000" became "half of 44100" - forced redither would calibrate for whatever sample rate you were saving as).

So there's always a little bit of print-thru between sample rate and sample depth, filter and dither in my mind

[WatchTheGuitar]

not at my computer right now, but if source is bipolar can you route it through that +10v line first and then into the vca to force the voltage to be only +ve?

into port labelled 6 then to 5/6 out

Tried that , not working either

In theory you have -5v to +5v, you add 10v to get 5v to 15v so you need to lower that by 5v somehow.
Try centring the slider on the +10v line

[gentleclockdivider]

I routed audio to the wrong input n the verry first example
It's pretty simple , just the input sockets were a tad confusing
This does it

[gentleclockdivider]

Windowed sync , formant syle

audio
https://app.box.com/s/6zkxg152kls5o2bqq9096nize9rv7jvg
https://app.box.com/s/ums4rapxmdh42yz579gc8jjn19mwwzzc