Waves, please add at least 2x oversampling option to your plugins if nothing else

[poshook]

I still really like some Waves plugins. However 2x oversampling would help to most of them to stay awesome even in current competition. The most I would see this option in API bundle (which I still really like)

[MogwaiBoy]

I agree - even their dedicated saturators like Aphex or Vitamin don't oversample.

However... with their recent plugins, Waves seem to be struggling to keep CPU use under control as it is. Oversampling would make some of those newer plugins even less usable.

Still, there are ways around that like realtime/render-only settings.

[simon.a.billington]

2x and 4x, that would be a nice option.

I’m sure some of their later plugins are oversampling, but I think it should be standard for any dev to offer a 1x, 2x and 4x as user option.

What I would really like is an intelligent way of telling whether the user is playing, mixing, or bouncing/rendering and automatically set the plugin accordingly or allow us to set the options up in the preferences.

[acYm]

it does indeed feel quite modern to have adaptive quality settings in some plugins. Waves boast of their strongly forward-compatible and adaptable platform, imo they should flex on that a little more and give the old classics some love.

[simon.a.billington]

Agreed.

I’m hoping a new fix for everything old is on the way, like we have seen in the recent past with a few other plugins.

They need to d something to earn back a bit of faith from the angry mob running around with torches and pitchforks after the lacklustre release of V10. Perhaps something shiny to distract them with?!? Haha!!

[Rockatansky]

I’m sure some of their later plugins are oversampling

Insert a white noise generator before the plugin and a spectrum analyzer after it.
If there's a steep high cut somewhere around or above 18 kHz, it's (probably) oversampling.
That way you won't have to rely on conjecture.

Quick test...

No downsampling dip:
CLA-2A, CLA-3A, H-Comp, C1, C4, SSL Bus, eMo Q4+D5, AudioTrack, Renaissance Channel+Comp+Axx+Bass, API 2500, V-Comp, PuigChild, Linear Multiband, Linear Phase EQ, L1+L2+L3 in all their various flavors, MaxxVolume, MV2, Vitamin, MaxxBass, both SSL channels, API 550B,

Potential downsampling dip, some could just be high cuts for "character", some ranging down to around 15 kHz:
CLA-76, dbx 160, BSS DPR, Kramer PIE+Tape+HLS, H-EQ, API 550A, Scheps 73, both PuigTec EQs

Renaissance EQ seems to have compensated filters (that don't squash near Nyquist), Q10 doesn't.
V-EQ3/V-EQ4 seem to have a fixed "character high shelf".

[simon.a.billington]

I’m sure some of their later plugins are oversampling

Insert a white noise generator before the plugin and a spectrum analyzer after it.
If there's a steep high cut somewhere around or above 18 kHz, it's (probably) oversampling.
That way you won't have to rely on conjecture.

Quick test...

No downsampling dip:
CLA-2A, CLA-3A, H-Comp, C1, C4, SSL Bus, eMo Q4+D5, AudioTrack, Renaissance Channel+Comp+Axx+Bass, API 2500, V-Comp, PuigChild, Linear Multiband, Linear Phase EQ, L1+L2+L3 in all their various flavors, MaxxVolume, MV2, Vitamin, MaxxBass, both SSL channels, API 550B,

Potential downsampling dip, some could just be high cuts for "character", some ranging down to around 15 kHz:
CLA-76, dbx 160, BSS DPR, Kramer PIE+Tape+HLS, H-EQ, API 550A, Scheps 73, both PuigTec EQs

Renaissance EQ seems to have compensated filters (that don't squash near Nyquist), Q10 doesn't.
V-EQ3/V-EQ4 seem to have a fixed "character high shelf".

Well in all fairness I tend to prefer to work at 96k. So it makes me wonder what happens to plugins running audio at that sample rate.

Also consider that the original units may have filters built into their design as well, or simply and more realistically, they were just built to respond to a specific group or frequencies within four hearing range and the manufactures weren’t interested in making devices that can respond to frequencies we can’t here.

Although, that would be slightly faulty logic on their behalf.

[The Noodlist]

GUI resizing, sleek interfaces and oversampling, features often seen in some modern plugins. Especially size, for modern hi-def monitors.

AntiAlias Free VST Audio Filter For Windows, seems to bridge to 64bits.
https://www.kvraudio.com/product/antial ... ne/details

VST Oversampler
Working download link

[pixel85]

Why not simply work in higher SR projects instead? Every modern computer can handle it easily. Beside that it's much more efficient than oversampling as plugins doesn't need to do: upsample-processing-downsample (which not only demand more CPU power but also affect latency).

[plexuss]

My feeling about Waves is that they've invested so much in their status quo that now they basically can't catch up. Waves is Yesterday.

I still use their analogue modelled tools but I do so knowing I am not using anything respresentative of current technology. I'd unload them all but now with their WUP+transfer fee it's not worth it. That's fine because when a company plays games like this I cut them out of my purchase decisions.

I will no longer buy Waves products unless they change their ways.

[Grizzellda]

My feeling about Waves is that they've invested so much in their status quo that now they basically can't catch up. Waves is Yesterday.

Gonna sort of agree & disagree with you, plexuss. Sure, they do have alot of legacy plugs to uphold & maintain, but they have also been talking about a new approach called OrganicReSynthesis, which has appeared in their latest products, Torque & SmackAttack.

There is a nice article about it on their website blog. I think they will release a synth, or some new stuff, utilizing this technology.

I don't think they are stagnant, rather still to be reckoned with. Lately I have been playing with Element 2...on the surface it is a straightforward , pretty simple synth, but man it sounds the dope...a few tweaks here & there and the sounds are just bitch'n. It produces sounds of deep quality, and that takes experience in the whole DSP field.

And experience is corelated to "know-how". And know-how can be gotten by learning from mistakes over time. (experience). And they have been around for quite awhile.

So..."nope". Not quite ready to say they are "yesterday" has-been's. Actually, I'm looking forward to new stuff from them!

[simon.a.billington]

My feeling about Waves is that they've invested so much in their status quo that now they basically can't catch up. Waves is Yesterday.

Gonna sort of agree & disagree with you, plexuss. Sure, they do have alot of legacy plugs to uphold & maintain, but they have also been talking about a new approach called OrganicReSynthesis, which has appeared in their latest products, Torque & SmackAttack.

There is a nice article about it on their website blog. I think they will release a synth, or some new stuff, utilizing this technology.

I don't think they are stagnant, rather still to be reckoned with. Lately I have been playing with Element 2...on the surface it is a straightforward , pretty simple synth, but man it sounds the dope...a few tweaks here & there and the sounds are just bitch'n. It produces sounds of deep quality, and that takes experience in the whole DSP field.

And experience is corelated to "know-how". And know-how can be gotten by learning from mistakes over time. (experience). And they have been around for quite awhile.

So..."nope". Not quite ready to say they are "yesterday" has-been's. Actually, I'm looking forward to new stuff from them!

Mind you, Organic Resynthesis isn’t some rehash of the Fourier Transform, it’s new tech entirely. Then there is the FIR tech that went into H-Reverb. It’s not your regular IIR convolution or algorithmic reverb, it’s something different again. You can find stuff dotted around the net on FIR. In short, it produces better quality, but at the expense of a larger CPU hit.

Personally, I’m quite sure this is the same engine that is going into their Abbey Road Plates and Chambers. That also sounds like no convolution or algorithmic design. Plus their CPU consumption is a bit of a give away. Consider also that you wouldn’t put that much time and energy into new tech only to use it once. They definitely had other plans for it, H-Reverb was just the start.

Their analogue synth engine is quite lush. On any other synth you’d have to use multiple oscillators and detuning to even kind of get close. Element gets their on just two oscillators!! This is because their tech isn’t just some vanilla approach where they just recreate the different synth parts digitally and throw in some detuning and saturation on top. Their tech actually models what happens to electricity as it travels down a wire. As you can hear, it produces very good results.

Then there are the other tools Waves has, that you can’t get anywhere, Cobalt Saphira, Vitamin, some of those Abbey Road plugins, Torque. Then there are those plugins that offer more flexibility than your usual, like Smack Attack and the Scheps Omni. Very intuitive design, time saving design and special features.

And there will be more on the way. Waves certainly isn’t yesterday, they’re doing new things that other competition aren’t doing yet.

[3243421342]

Waves plugins are trash for not oversampling. Even on higher sampling rates you'd still get aliasing unless you cutoff those frequencies manually. But that defeats the purpose of working in high sampling rates anyway. To prevent aliasing you have to cutoff audio at the nyquist frequency before it gets transformed to digital, which is what any decent DAC can do. But since you never leave the digital realm you'd have to oversample the audio so the plugin can see what frequencies to cut out. You'd argue that working in a higher sampling rate would fix that problem right? Wrong! Fabfilter illustrated why this isn't true and why oversampling is a much better solution. https://youtu.be/-jCwIsT0X8M

[jochicago]

I understand the principles but I'm fuzzy on the specifics of the math. Regardless, I know that ~200khz (or about 4x oversample at 48k) is still not enough to get rid of aliasing completely in all situations. I remember TDR said that their "insane" precision mode is effectively getting rid of aliasing, and they go to ~450khz for that (or 9x oversample at 48khz). Many plugins offer 8x and 16x options for this reason as well.

It's not enough to just run your sessions at 96khz, although that would help some for plugins that don't have internal oversample (assuming the plugins can handle it, not every old plugin does well at 96k). But also keep in mind not every algorithm needs oversampling. Typically, a saturator like Saturn is going to need the most oversample love because of the character of what it's doing. But many algorithms can probably cover a lot of healthy sonic ground without quite as much oversample.

I remember Hornet uses different internal oversamples for his LA2A and his 1176 compressors (I think). I think one is at 2x and the other at 4x, and the logic is that that's all they needed so why abuse the CPU. You wouldn't see the option to oversample, it just happens internally.

I know that some Waves plugins are overampling for sure, particularly the more recent ones. They certainly eat enough CPU. I think they just try to make a decision on what's appropriate, and not expose the feature.

I don't see them adding a 2x oversample option to old plugins. It might be seen as a confession that their 20 y/o DSP could use updating. Would open a pandora's box of people asking questions about what else in that historic DSP could use refurbishing. It's a lot easier to just let the marketing guys keep calling Ren EQ the most revolutionary "analog" EQ money can buy in 2020.

[krabbencutter]

---