Plug-in to raise noise floor / sidebands?

[Girolad]

Okay all,

As we all know digital being super pristine today - even more so now with 32 and 64 float, I can't help but wonder... is this out of touch?

I have noticed that alot of plugin's depsite being analog emulations, still don't 'dirty' the signal up enough imo, relatively speaking regarding levels,

What I mean is when I play a 'clean' tone and view through FFT
I get this: see attachment1.

Everything I've tried - tape emulators and upward compressors I struggle to get 'sidebands' I think their called any higher.
For example, see attachment2... This is considered a 'dirty' setting on a tape plug-in.

Another example, Red Phatt Pro by jeroen breebaart has an upward compression preset, it kind of works, but still not so much unless I crank one of the Ratio dials. (jattachment3)

Perhaps I'm wrong in thinking a universal plugin is required to simply raises the noise floor with great analog emulation here.

[Sendy]

There are a few plugins that will do an FFT and separate the sound into sine and noise components, then let you control the balance of them. AnarchyEffects have one such plugin.

There's a spectral gate called Spectral Monkeyage which does something similar. After the FFT you set a low and high threshold and then you can cut or boost any portions of the spectrum above or below a certain amplitude.

Thing with these FFT effects is they can change the sound even at neutral settings, so I prefer to use on single instruments, not an entire mix.

[BertKoor]

Why don't you just mix in some pink noise at -70dBfs to add some tape hiss?

[VariKusBrainZ]

If you been even more compression load another one (or more) in serial

[Girolad]

If you been even more compression load another one (or more) in serial

You mean have two plugins in series? Will give that a go.

The reason I focused on upward compression is because digital is obv perfected from 0dbfs top down, it makes more sense to use upward in the digital domain imo, any one else agree?

Thanks for the suggestions.

[mehum]

While I don't agree with your premise I can try and sort some things out...
First of all, you will have two types of noise, or non-linearities, in an older recording or reproduction system, such as tape.

The first is dormant noise from the random distribution of magnetic particles, in the case of tape. Ironically engineers would go through great lengths in order to get rid of this so to simulate this you could try to start with white noise and the shape it to copy the effect of a Dolby A system.

The other is distortion to the recorded signal in the form of slightly changed frequency response, compression and saturation. While often avoided on material that were supposed to be cleaned it was just as often used for its specific characteristics, especially for drums.
There are countless plugins simulating this, with more or less subtle effect. PSP vintage warmer seems popular. I've found the free IVGI work very well for giving drums a bit of punch. Variety of Sound has a multitude of plugs that simulate non-linearities in various recording systems.

Adding noise and distortion has never been a problem in the recording world...

Compression of any kind won't give you what I perceive as the desired result since it only will affect the noise inbetween louder parts.

I did not understand the part you wrote regarding upward compression and 0dbfs. I think you're misinformed regarding this but perhaps you could explain better what you mean?

[Girolad]

I did not understand the part you wrote regarding upward compression and 0dbfs. I think you're misinformed regarding this but perhaps you could explain better what you mean?

Hi, thanks for reply.
Well with digital and more-so today, is much easier to measure distortion, and if given the means to detect, fixate over the little amounts of distortions remaining (whats left of it). This design methodology reflects on how everything is marketed today including converters.
Why we would want to pursue the ability of create and hear signal clean to - sometimes - 120db down (regarding converters; achieved using 'SNR enhancement' tricks a la oversampling *cough cough*) to preserve all this 'low level detail' when I doubt ears find it pleasant.
Its unlikely our hearing enjoys listening past the top level and I think thats the main difference today.

If you wanted to emulate older digital hardware for example, wouldn't one use upward compression? Unlike typical compression, I feel upward still makes use of all the available dynamic range offered by converters, while still adding bottom.

For any newbie starting out it is not obvious the fact that there was a time when music came with a noise floor as a perquisite of using analog mediums (perquisite because of how the ears like to listen) and that this is not enabled by default when creating.

When I compare older and newer digital recordings, I think sometimes there is too much low level detail being preserved, or that it resides far too far down in the noise floor, where as older equipment would ameliorate anything that far down automatically.

You could argue; pipe digital signal through a tube pre-amp to round everything off at playback, (I'm thinking about doing this), but again this is a discretionary step.

Hopefully I'm more clear here.

As for upward compression specifically, If you wanted to preserve the peaks but increase the signal at the base, as was the case with older converters.
I just wondered if there was an alternative to buying a tube pre-amp.
IVGI looks good I'll give it a shot.

PS: I think exterior and interior are good terms to describe signals present in / outwith digital domain.

[xoxos]

i couldn't read this thread, because i am used to most kvr users being filthy masons and spreading filth all over the place. i'm absolutely confident i know what has been stated and what hasn't regardless jedi skills bro.

adze (or the free version stoooner) have this function (demonstrated in the video) but it's a percussion sampler.



i suppose i could port this to an effect, but i make plugins with synthedit, 32 bit, and 2.4, so since so many people are outspoken about their adversion to these things, i don't see the point in bothering. even if you begged me and i did it, the masons would cover it up and no one else would find out.

[mehum]

Hi, thanks for reply.
Well with digital and more-so today, is much easier to measure distortion, and if given the means to detect, fixate over the little amounts of distortions remaining (whats left of it). This design methodology reflects on how everything is marketed today including converters.
Why we would want to pursue the ability of create and hear signal clean to - sometimes - 120db down (regarding converters; achieved using 'SNR enhancement' tricks a la oversampling *cough cough*) to preserve all this 'low level detail' when I doubt ears find it pleasant.
Its unlikely our hearing enjoys listening past the top level and I think thats the main difference today.

Well, having a good dynamic range has its perks, especially when recording since you don't have to push any piece of your equipment to get a clean recording. For creating sounds ITB it could certainly be considered overkill but I don't see any downsides except perhaps spending CPU cycles.

If you wanted to emulate older digital hardware for example, wouldn't one use upward compression? Unlike typical compression, I feel upward still makes use of all the available dynamic range offered by converters, while still adding bottom.

For any newbie starting out it is not obvious the fact that there was a time when music came with a noise floor as a perquisite of using analog mediums (perquisite because of how the ears like to listen) and that this is not enabled by default when creating.

When I compare older and newer digital recordings, I think sometimes there is too much low level detail being preserved, or that it resides far too far down in the noise floor, where as older equipment would ameliorate anything that far down automatically.

Do you have an example? I believe most people actually would argue the opposite, that modern recordings don't use all this available dynamic range due to the extreme settings used when mastering.

There could be a multitude of other explanations to why you prefer the sound an older recording to a newer. It's very hard to compare...
Say we'd compare recordings of symphonic music. A recording to analog tape and played back on vinyl would never be as convincingly close to what you hear sitting in a concert hall as a digital recording.

So our approach to noise should be dependent on the experience we want to create...
For electronic music this could of course go anywhere and there's every possibility to both use the 100 or so dB of dynamic range or choose not to.

You could argue; pipe digital signal through a tube pre-amp to round everything off at playback, (I'm thinking about doing this), but again this is a discretionary step.

I know of electronic artists who run the final mix through tube equipped master tape machines to get that certain flavor of compression and saturation, so this is certainly not unheard of. A guy I used to work with built stereo machines by slaughtering three different old tape machines (two mono Studers and one stereo Revox). "Best compressor ever" he said

Hopefully I'm more clear here.

As for upward compression specifically, If you wanted to preserve the peaks but increase the signal at the base, as was the case with older converters.
I just wondered if there was an alternative to buying a tube pre-amp.
IVGI looks good I'll give it a shot.

PS: I think exterior and interior are good terms to describe signals present in / outwith digital domain.

It sounds on your description that you're more after the compression and saturation than actually making the noise floor higher. Try out a few saturation plugins and see if they do the trick. I'd be a bit careful with cheaper tube equipment since they're sometimes just using the tube as a marketing ploy and they sound like shit if overdriven even slightly

And anyway, your compression technique probably won't give you any desired result. I think you are correct in that an upward compressor would make noise louder while preserving the rest of the content better than classic downward compression. However, setting the threshold as low as you'd need to go you would probably not hear any difference.

I think any kind of compression to achieve noisier result is wrong for two other reasons:
First, the only noise you'd make louder is the noise in between. Whenever your recorded or played content starts, the compression will stop (or opposite if using downward) and will again push down the noise. This is actually opposite what you'd expect from an analog recording where expanders were usually used together with filters in order to keep the noise as low as possible.

Second, think about what noise you're actually making louder. Ideally, if you're doing everything in the box, it will be nothing. There's simply no bits of data to make louder. At this stage there won't even be quantization noise. If you have some tape emulation you might be able to make this louder but a much easier way is to just add another tape emulation in series. If you're just interested in the tape hiss, add it to a channel with nothing else on and boost it to taste.

If it really is noise floor you want, the easiest way would be to add noise. Any synth will be able to generate noise and then you can use EQ to shape it to something pleasant. This will give you much more control than any strange compressor trick. You could even run the silent parts your favorite recordings through an analyzer to see how the noise is shaped and try to emulate it.

The FFT tricks that has been mentioned could work if you have an already saturated signal, such as an emulated analog synth, and want to add just the noise components. But you'd probably have to work harder than necessary to get a good result compared to just add saturation and/or noise afterwards.

As I said in my first reply, I'm not really in line with the premise that noise-free is bad, but I still hope I'm of some help in finding the best solution.

[highkoo]

Not sure I completely understand the question, and I am sure I do not completely understand this tool, but Neodynium is extremely good for raising a noise floor and bringing out the subtle dirt;

http://www.kvraudio.com/forum/archive/e ... /56174.php

[Girolad]

I should have said I wasn't after a real solution but rather some opinions. I think I underestimated the cumulative distortion that would result if I wanted anything higher than what I presented in OP... which was a single tone. Really I wanted to focus on upward compression and not FFT editing...

For creating sounds ITB (in the box?) it (clean dynamic range) could certainly be considered overkill but I don't see any downsides except perhaps spending CPU cycles.

It's preserved until its played off a CD back through a converter with similar specs and that depth is maintained, I believe there's too much depth today if you take into account how the ears work.

When I compare older and newer digital recordings, I think sometimes there is too much low level detail being preserved [...] where as older equipment would ameliorate anything that far down automatically.

Do you have an example?

51 minutes in lot's of oomph in that track.

I believe most people actually would argue the opposite, that modern recordings don't use all this available dynamic range due to the extreme settings used when mastering.

No, I meant old practices made use of all available DR and I am in agreement with what you say about dynamic range.
The irony is scientists have labored and perfected the digital path and it's disregarded in final instance anyway - but this is down to competence really, my point is; it's not exploited by default! where as during the magical days a la tape, people never even though twice about; damn that hiss.
The only exception was those who's job it was to sell a product with less hiss, who's taste for classical music happened to appeal to that requirement. E.g. flutes playing in silent orchestral pit at -72dB
As for example, most music in my collection, some paid for (I'm glad I didn't pay for all of it) but typically early 90s, the electronic stuff, e.g. early house music was converted to digital and pressed to vinyl simultaneously, most likely from tape, it's all still very dynamic (measured using DR meter). Crosstalk, wow, flutter and noise is all there, lots if going by todays standard and yet it still sounds fairly loud n great.

Compared recently, particularly 2000 onwards - overall - mastering practices and tech changed, now theres no tape in between, recordings exhibit minimal crosstalk (maximum stereo separation) that digital offers. Do ears enjoy listening to subtle detail at depths of -192db? (-96db plus from either speaker)? I think not.

A recording to analog tape and played back on vinyl would never be as convincingly close to what you hear sitting in a concert hall as a digital recording.

I think this largely depends on how it's mastered for vinyl.

You could argue; pipe digital signal through a tube pre-amp to round everything off at playback, (I'm thinking about doing this), but again this is a discretionary step.

I know of electronic artists who run the final mix through tube equipped master tape machines to get that certain flavor of compression and saturation, so this is certainly not unheard of. [...] "Best compressor ever" he said

I whole fully agree.

I think you are correct in that an upward compressor would make noise louder while preserving the rest of the content better than classic downward compression.

I'm glad, so I'm wondering if there is a dedicated upward compressor out there for this role, there probably is and hoping or preaching for an expert on this to chime in.
There does appear to be a 'bottomizer' preset on decimort (digital 'grime box'.

As I said in my first reply, I'm not really in line with the premise that noise-free is bad, but I still hope I'm of some help in finding the best solution.

It's all good.

[mehum]

I should have said I wasn't after a real solution but rather some opinions. I think I underestimated the cumulative distortion that would result if I wanted anything higher than what I presented in OP... which was a single tone. Really I wanted to focus on upward compression and not FFT editing...

I think you are correct in that an upward compressor would make noise louder while preserving the rest of the content better than classic downward compression.

I'm glad, so I'm wondering if there is a dedicated upward compressor out there for this role, there probably is and hoping or preaching for an expert on this to chime in.
There does appear to be a 'bottomizer' preset on decimort (digital 'grime box'.

Well, as I stated, a compressor still won't give you the result you're after.
Single-most easy way is to just add noise and distortion (saturation).
The compressor method (wether upward or uplift) will only give you the noise between the music, not the noise going on at the same time. In a purely digital ITB recording this will be quantization noise (and perhaps a bit of reverb tail if you've set the threshold too high).
If this is what you want to add you should just try a bit reducer (like the Decimort you're referring to). Set to something like 12-14 bits you'll get an audible noise and reduced dynamics.

But if you'd rather have something like vinyl noise or tape hiss you're much better off just adding a white noise source and filter it to taste.

For creating sounds ITB (in the box?) it (clean dynamic range) could certainly be considered overkill but I don't see any downsides except perhaps spending CPU cycles.

It's preserved until its played off a CD back through a converter with similar specs and that depth is maintained, I believe there's too much depth today if you take into account how the ears work.

Why? I honestly don't understand... You will usually compress and effectively not use all bits. Noise is then added at every part in the reproduction chain, including inside your body.
The dynamic range still adds options for the artists and engineers to use this. This range is rarely used but in experimental electronic music or symphonies it's certainly possible to use these ~90dB to some effect.

If you're making music where the noise (for example from the vinyl) is part of the sound, just add noise and distortion to taste.

Crosstalk is a bit of a different beast... It would be nice to see a DAW implementing this, feeding a bit of the signal over a certain threshold to the closest tracks
Putting something like IVGI on the master bus can at least simulate crosstalk between left and right channels.
You could also want to simulate different signal paths for left and right channel by having separated EQ:s with slightly different settings.
Both vinyl and tape machines need to be meticulously aligned for phase correlation both when recording and reproducing, and will never be absolutely perfect, so some slight time offset for one of the channels will also help if you're after that kind of sound.

Wow and flutter is a bit harder in the digital domain but should be possible with a good pitch shifter, automating play speed.

Most of these things should probably be used very very subtly...

Do you have an example?

51 minutes in lot's of oomph in that track.

Yeah, that's nice

Again, this is very much a style where the noise is part of the genre (and virtually impossible to get rid of if you're sampling from vinyl). The ooomph could certainly still be there without the noise though.
What modern music are you comparing this to? The comparison is virtually impossible in my opinion. Go to a dubstep club with massive sub bass and you'll hear serious OOOMPH not possible in the vinyl era since it would make the needle skip

Compared recently, particularly 2000 onwards - overall - mastering practices and tech changed, now theres no tape in between, recordings exhibit minimal crosstalk (maximum stereo separation) that digital offers. Do ears enjoy listening to subtle detail at depths of -192db? (-96db plus from either speaker)? I think not.

Well, then just don't make what you think the ears don't like. Of course you're not alternating between a sine at -90dBFS and one at -0.02dBFS. The dynamic range gives you options to make quiet stuff noise free, that's all there is to it.

A recording to analog tape and played back on vinyl would never be as convincingly close to what you hear sitting in a concert hall as a digital recording.

I think this largely depends on how it's mastered for vinyl.

Haha, make it directly from tape then. Sure, 1" master tape at 30ips is close to PCM when it comes to dynamic range and frequency response. Played off a nice Studer machine with good engines the flutter would be negligible as well.

As a last point, I'd like to emphasize that I believe the compressor technique is the worst possible to achieve what I'm guessing is your end result. Instead, adding noise and distortion in a controlled fashion is my recommended method.


EDIT:
I just saw this one: http://www.u-he.com/cms/satin
Looks great for getting some tape vibe on all or selected tracks!

[simon.a.billington]

Tell ya what, try this...

Waves NLS


It emulates the noise, the harmonics and the saturation of three different analogue desks. Even the noise has been modelled.

Also try any of their analogue device emulations. Waves has incorporated a switch or a dial called Analog with the design, specifically designed to introduce the noise of the original unit back in.

http://www.waves.com/plugins/analog-models#

[xoxos]

you know,

i posted a video above and a link to a free plugin

that does exactly what you want


except on one-shot samples.


i can turn this into an effect for continuous audio in a matter of hours.

but i don't see the point of making exactly what you want, if you are gonig to ignore it.

[BertKoor]

'sidebands' I think their called

What shows on your spectrograph pictures is simply called harmonic distortion. Frequency content that was not there is added. You get that with clipping, overdrive, saturation, extreme compression & limiting, any part in the signal chain that is non-linear.

Why we would want to pursue the ability of create and hear signal clean to - sometimes - 120db down (regarding converters; achieved using 'SNR enhancement' tricks a la oversampling *cough cough*) to preserve all this 'low level detail' when I doubt ears find it pleasant.

Today in this thread I found a link to this excellent article that that discusses this subject in depth. Do yourself a favour and get educated. Take the time to read it and understand it fully.

You are right that for playback 44kHz/16bits is sufficient. Thanks to dithering you can record/play signals with an amplitude well below the theoretical limit of -96dBfs of 16bit audio. And AD/DA converters have improved their frequency response in the highest octave significantly thanks to internal oversampling! So there are very good use cases for higher sampling rates and higher bit depths. But not for simple recording & playback indeed. Not all the marketing propaganda is bullshit, there is some truth in it.