YouRei HP/LP filters, Nebula

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
RELATED
PRODUCTS

Post

A set of HP/LP filters sampled from a unique filter unit. These aren't synth style filters (no resonance), but they are like the type you find on a console that are useful for mixing/production purposes. The cutoff frequency on this unit is fully variable, however, instead of just giving you a few fixed positions like most console modules. I've captured that ability here well. You have smooth control over the cutoff and the filter shape is maintained.

There's a combo filter with both programs in one, and it is mostly just capable of a clean sound which could be used for anything. Then there are separate programs for both the LP and HP filter, which can get grittier (if driven hotter), which could be perfect for sample based stuff if you want that lofi sound. This unit was released in 2 versions, and this is the more desired one with the output transformer, which adds a bit more to the sound.

The HP can go from fairly flat (no filtering), up to around 300-400Hz. The LP can go down to around 2kHz. The unit is known for it's color, but it can be fairly transparent also. These were sampled with dynamics and as far as I know they are the only 'channel-style' filters that have been so far. So for every position of the filter, you get a nice dynamic representation of the unit. You can send in a louder sound for a more 'lo-fi' effect, or a lower level in for a cleaner sound. A trim control allows you to quickly transition through those possibilities while keeping a constant output level.

They were sampled with Lynx Aurora and I think they sound great! If you want some nice filters for basic mixing duties, you should check these out!
http://www.cupwise.com/cup/yourei-hplp-filters/

I actually released these a few days ago and you can see what some people have said about them here, at the acustica forum, plus also some more details I give, particularly about an upcoming future update:
http://www.acustica-audio.com/forum/ind ... =viewtopic

Post

They sound really nice, I'm a big fan of characterful nonresonant cutoff filters. Pity I've no idea what Nebula is :D
http://sendy.bandcamp.com/releases < My new album at Bandcamp! Now pay what you like!

Post

it's a plugin made by Acustica Audio and it's been around for a few years now. if you know about 'convolution' which has been around for several years now, which is usually used for reverbs, well nebula took that concept to the next level or two.

you can 'sample' the sound of something by passing test tones through it, and then those tones are analyzed and a convolution process can recreate that effect. there are limits. standard convolution couldn't sample dynamic behavior, or distortion. nebula can (although it can only sample up to 9 orders of harmonic distortion). nebula has been used to sample equalizers, preamps, consoles, tapes/tape decks, compressors, reverbs, and i myself first started by sampling a bunch of old tube radios i bought. i actually sent the signal through the air with an FM transmitter, to be picked up by the radios, and in other cases took them apart to directly send the signal into them, to get various different sounds from them. the end result could be used as a variety of really colored filter effects that give stuff a really old sound. anyway, in my opinion it's a powerful tool that still has uses not yet realized.

a lot of people think it sounds more realistic than the more typical and very common algorithm based stuff trying to recreate the sound of 'vintage' hardware. but there are drawbacks, it takes a fair amount of cpu and it can take a little time to get used to it's eccentricities. lots of people swear by it though.

if you looked around in the acustica forum i linked to, in the 3rd party developers section, you might get some kind of idea about it.

Post

i've just added an update that adds stepped versions of the filters because some people seem to have an aversion towards interpolated filter programs. personally i think the interpolated ones are fine because the sampled positions are close together, but the stepped ones are now there as an option.
then there are non-dynamic versions which are light on cpu. there are also stepped/non-dynamic ones.
the combo filter couldn't get any of those because of a special way its samples are organized, so a patch wouldn't work on it. same is true for the 10k LP only program.

Post

Very, very cool! I'm probably going to buy this. Personally, I'll be using the stepped versions, simply because I care less about getting my filter frequencies spot-in than I do about possible negative artifacts.

Post

Uncle E wrote:Very, very cool! I'm probably going to buy this. Personally, I'll be using the stepped versions, simply because I care less about getting my filter frequencies spot-in than I do about possible negative artifacts.
i feel kind of a need to make my feelings on the matter known, since i am the developer who put this thing together, but i also don't want to come off as combative or anything towards anyone because of their feelings. obviously i added the update with stepped filters because i respect my customers' opinions and like to satisfy requests when i can.

that said, i really feel like the interpolation with the interpolated filters (at least THESE filters, i can't speak for any others) is a non-issue. this is from my experience doing this for almost 3yrs. i think the whole reason some people started thinking negatively towards interpolated filters is because of:
a) synth filters with resonance where more sampled positions are required (vs. a filter like this without resonance) to get a smooth sweeping transition. some of these that have been done for Nebula, especially the earlier ones from years back, didn't have quite the required # needed to get that smooth transition. people noticed this and remembered.
b) even with a non-resonant filter, if you interpolate between two sampled cutoff frequency positions that are relatively far apart, you will get an 'odd' result. i wouldn't call it an artifact at all. from my experience i don't think you would ever get anything i would refer to as a 'bad artifact' from less-than-ideal interpolation between filter cutoff points. a strangely shaped filter that isn't characteristic of the unit, yes. bad artifacts, no. at worse, you might get some unwanted phase cancellation, but it would be easy to see on a freq response analysis and i've never seen it from filter interpolation.
c) console filters usually have just a few fixed freq positions for the cutoff, and there is a comparatively large distance between those points. here, i have a sample about every 10hz with the HP filter. compare that to a console module where you have a selection of maybe 50hz, 120hz, 320hz, something like that. now i've never sampled a console module filter, but i get the impression that maybe someone who has, offered an 'interpolated' version of that, which would give you kind of wacky filter shapes. it's not that it would be 'bad' in a certain sense, just a strange shape that really isn't like the unit. (what i mean by this, is that instead of the 'knee' shape you get with a typical HP/LP filter at the cutoff point, you will get something a bit more like a two-stage knee, for example a high-pass filter just before a cutting low-shelf. it will sound odd compared to the actual filter, but it isn't doing anything inherently 'bad' to the audio. it's just an unusual filter shape, because it's really a mix of the two high passes)

you don't get that with this filter. there isn't any other voodoo to it other than that. you can see on graphs what is going on and nothing out of the ordinary happens, if you put the filter between sampled points, where nebula has to interpolate.

basically- if a filter is sampled properly, there isn't going to be any issue with interpolation. and i feel like i sampled these about as proper as you can get, with the addition of dynamics. the combo program could use a bit more resolution with the HP filter, but consider that there is a separate program for both the HP and LP, AND that all three of these were done with dynamics. that means the 6k combo program has over 4000 samples in it. look at your nebula collection and you'll see you have nothing else that comes close to that. there is a ton of stuff crammed into that. nobody else is doing filters with dynamics. so, the tradeoff is that i had to skimp on resolution with the HP filter for the combo program, just a bit. i STILL think it sounds great, and the only way anyone would notice it is by looking at an analyzer.

long story short, these programs, interpolated or not, absolutely just DO NOT have any kind of bad artifacts in them. i'm saying that as the guy who put countless hours into them, and who knows what he is doing. it's not even a question or even debatable. the only thing that's debatable is, would more sampled positions give an even more smoother transition. the answer will always be yes. but the real question becomes at what point can a human notice the difference. i think that at least the solo LP and HP programs are pretty well approaching or past that point.

i made the stepped versions but i don't think i'll do it again. i feel like i'm pandering to unwarranted paranoia. for exampled, the main person who requested this, in the same post said these were his favorite Nebula filters (based off of the interpolated version). if something sounds good, it IS. like i said, if it's sampled properly, there's no problem. with this many sampled points, there isn't going to be any difference in the interpolation here, from what you get between dynamic steps in any dynamic program. nobody worries about interpolation there. this includes every tape, console, compressor, and preamp program.

again, i REALLY hope i didn't seem combative here. but i hope anyone who actually for whatever reason punished themselves by reading this rant understands that, i said it because the whole issue of filters w/interpolation comes down to how well is it sampled, so i'm basically backing up my work. if you can look at an analyzer (like vstanalyser.exe) and see that the shape of the filter is maintained as you sweep across, then it was sampled properly, and i can't fathom how anything could be wrong with it.

-end rant

Post

ok, your post seems reasonable, but this is the thing...

last night i had this dream. i was in a futuristic recording studio at some point far away in the future. my assistant engineer was a beautiful young woman who had surgically altered space ears (among other interesting surgical alterations). basically she had quantum computer assisted hearing.

we were mid way into a mix session and every time i made a change to the mix she began to get more and more excited.

"oh, thermal, your judgement is so wise!" she moaned.

she pulled off her high heels and started to gyrate by the console.

"oh yeah! eq that bassline... yeah, use nebula for best transient response... yeah!"

"oh my space ears are stating to tingle, this mix is starting to go cosmic"

THEN i pulled up your nebula HP filter program on the screen and started to cut the low frequencies using your *interpolated* version.

suddenly alarm bells started sounding from the console. a big red flashing light was blazing on and off. my space babe started shrieking in horror clutching at her quantum computer assisted ears.

"NO... WHAT HAVE YOU DONE... INTERPOLATION ERROR INTERPOLATION ERROR!!"

the computer spontaneously crashed and deleted its hard drive in protest at my crime, the band came running through the door shouting and pointing lasers at me.

"what the hell are you doing with our mix mate!! is this some kind of joke!!"

i tried to run but they grabbed hold of me and started to beat the living daylights out of me. my sexy space babe was laughing in my face as i fell to the floor.

well... i guess interpolation could be a "non issue"... for some people :D

Post

OK.

...
I STAND CORRECTED.












;)

Post

Cupwise wrote:i made the stepped versions but i don't think i'll do it again. i feel like i'm pandering to unwarranted paranoia.
If the frequencies jumped from 70Hz to 160Hz, like on some Neve's, I'd definitely be interested in trying an interpolated version because the jump is so large and there are some useful frequencies in between there. In this case, however, the filter steps are only every 10Hz and there frankly isn't any need to interpolate.

Post

Uncle E wrote:
Cupwise wrote:i made the stepped versions but i don't think i'll do it again. i feel like i'm pandering to unwarranted paranoia.
If the frequencies jumped from 70Hz to 160Hz, like on some Neve's, I'd definitely be interested in trying an interpolated version because the jump is so large and there are some useful frequencies in between there. In this case, however, the filter steps are only every 10Hz and there frankly isn't any need to interpolate.
ok but:
a) like i said, with a jump from a distance of 70hz to 160hz, you aren't really going to get any 'interesting frequencies between there', instead you get half of a high pass at 70hz, and half of one at 160hz. if you looked at the result on an analyzer it would look like a low shelf cut of about 6db (if you went right between the two, at 115hz) starting at 160hz, then a high pass at 70hz. i could even quickly put together a sample program showing this, sampling a vst eq.
b) to say 'isn't any need to interpolate' suggests that interpolation is inherently bad in and of itself and should be avoided altogether. it isn't. if it were 'bad' somehow between 10hz steps, it would be unusably bad between 90hz steps, negating any use you might get out of interpolating between those 2 steps (and the only use would be if you wanted a mix of a low shelf with a highpass). further, any dynamic programs you have would also be bad and unusable.

a preamp, console, tape sim, or any other dynamic program uses interpolation. usually a distance of 1db between sampled levels. i would say this is about comparable to my frequency distance of about 10hz. so if interpolation is bad between my filtered steps, it should also be bad between sampled dynamic steps. take out the dynamic steps and now you are pretty much left with standard convolution. the ability to interpolate is one of a few of the most important things Nebula improved over standard convolution. i mean, if you really want to debate how interpolation is bad then we can, but you have to actually address these points i'm making. if interpolation were bad then nebula would be bad. i don't think it is but that's me.

Post

not only do the dynamic programs use interpolation, but if you think about it, what you hear from them is pretty much always the product of interpolation. if you think of each sampled dynamic step as a point, which are all being connected to form a line, then the audio level going in tells nebula which sampled level to use from the line, which pretty much always is going to be between any two sampled points. so what you hear is always interpolated. why is that ok but interpolation anywhere else is bad?

Post

Cupwise wrote:a) like i said, with a jump from a distance of 70hz to 160hz, you aren't really going to get any 'interesting frequencies between there', instead you get half of a high pass at 70hz, and half of one at 160hz. if you looked at the result on an analyzer it would look like a low shelf cut of about 6db (if you went right between the two, at 115hz) starting at 160hz, then a high pass at 70hz. i could even quickly put together a sample program showing this, sampling a vst eq.
This pushes me further toward the idea of stepped versions. Again, I haven't bought your impulses yet and haven't heard for myself what you're talking about, this is all just conceptual for me right now. The concept of using interpolated frequencies when I don't need them just doesn't make sense to me.
a preamp, console, tape sim, or any other dynamic program uses interpolation. usually a distance of 1db between sampled levels. i would say this is about comparable to my frequency distance of about 10hz. so if interpolation is bad between my filtered steps, it should also be bad between sampled dynamic steps.
Right, I get that. The difference is interpolation is impossible to avoid if we want dynamic convolution, whereas it's very easy to avoid with no major downside with your filters.
i mean, if you really want to debate how interpolation is bad then we can, but you have to actually address these points i'm making.
I don't really want to debate it. In fact, I don't think that I've once said I think you're wrong about it, I'm merely stating that I'd rather have stepped versions because I have no need for non-stepped versions. You're talking about lack of harm, I'm talking about lack of benefit.

Post

Uncle E wrote: interpolated frequencies when I don't need them just doesn't make sense to me.
the concept of being able to do a smooth filter sweep, on that occasion where you want one, doesn't make sense? what if we actually were talking about synth filters where filter sweeping is a big part of using them? again, the only reason i'm discussing this here, isn't to personally attack your opinion on this, it's because i develop for the Nebula platform and so i do have a personal interest invested in clearing up what i consider to be unfounded misconceptions about how it works. if people think filter interpolation is inherently bad for some reason, even if you can look at that filter with several different analyzers and see nothing unlike what you would see when sweeping a typical algorithm based filter, then that hurts a program i develop for. if i intend to one day do more more filters, it then becomes in my interest to clear this up.

my example of the interpolation between larger gaps SHOULD steer you towards stepped filters- for programs with larger gaps. that was the point. my program has smaller gaps. but your earlier statement was the exact opposite, which was that you want interpolation with larger gaps and stepped with smaller, which to anyone who develops anything with interpolation, this is exactly the opposite of how you want things to work. that's all i was trying to explain.

Uncle E wrote: Right, I get that. The difference is interpolation is impossible to avoid if we want dynamic convolution, whereas it's very easy to avoid with no major downside with your filters.
see, my problem is with your wording. your wording here is just taking for granted that interpolation IS something we 'want to avoid'. it doesn't address my point at all, which is that lots of people love the sound of those preamp, console, and tape emulations, which are all constantly giving them an effect that's using interpolation.
this should then therefore mean, that there is nothing wrong with the sound of that interpolation.
in fact, it means that it has been PROVEN, time and time again, that there ISN'T anything wrong with interpolation, when used correctly.
not only that, but it's been PROVEN that lot's of people think Nebula's interpolation sounds GREAT, whether they realize it or not. if someone says 'this Nebula preamp program sounds like the real thing', then what they are also saying is: 'interpolation in Nebula can sound like an analog preamp'
that would then mean that it isn't something we need to avoid, 'just because'. only in cases where it won't work properly (such as large gaps or differences between the two samples being interpolated).


Uncle E wrote:I'm merely stating that I'd rather have stepped versions because I have no need for non-stepped versions.
but lots of people do see a use for interpolated filters. the actual hardware unit i sampled here gives a smooth, fully variable filter sweep through its frequency range.

when i sampled it, i intended to do it so that the end result works like the unit. i did. i achieved that by using interpolation. a stepped program is less like the unit, and for no real demonstrable reason. if interpolation can't be proven by you or anyone else to have any negative impact on the sound, then why cripple my end result by using stepped which makes it less like the actual unit?
Uncle E wrote:You're talking about lack of harm, I'm talking about lack of benefit.
so a smoothly sweepable, fully variable filter is a 'lack of benefit'? ok, maybe you think so, but the problem is that when i make something i have to aim towards what i think the average person will think. let's say for comparison that you were arguing that having a 'dry/wet' mix control on every plug-in, wasn't needed because you saw it as a lack of benefit. but it doesn't hurt anything by being there, whether you use it or not. lots of other people would like it. as a developer should i be expected to not include the wet/dry slider because a handful of people start speaking out against it with claims that it hurts the sound, without giving any real reason or proof that it is?

at the end of the day, my feeling is that if people want me to cripple my filter programs because of some imaginary 'badness' being caused by interpolation, then the burden of proof should be on them, to at least explain how or why interpolation is bad. if it's just something you saw someone else say, well maybe they were talking about a filter that used a poor implementation of interpolation. that has nothing to do with my program here.

Post

Cupwise wrote:the concept of being able to do a smooth filter sweep, on that occasion where you want one, doesn't make sense?
If you are aiming this towards the EDM market, you'd of course dead on. However, that's not something we're looking for in the realm of mixing. What's great about your filters is you've provided your customers with both the stepped and the interpolated versions, so there's no possible compromise for either group. :)
if people think filter interpolation is inherently bad for some reason, even if you can look at that filter with several different analyzers and see nothing unlike what you would see when sweeping a typical algorithm based filter, then that hurts a program i develop for. if i intend to one day do more more filters, it then becomes in my interest to clear this up.
Personally, I think you could simply acknowledge whatever limitations exist and then develop the very best products possible within those boundaries. Imagine if someone did release a filter pack that interpolated between 75Hz and 160Hz, don't you agree that it would do more harm than good to people's perceptions of the platform?
my example of the interpolation between larger gaps SHOULD steer you towards stepped filters- for programs with larger gaps. that was the point. my program has smaller gaps. but your earlier statement was the exact opposite, which was that you want interpolation with larger gaps and stepped with smaller, which to anyone who develops anything with interpolation, this is exactly the opposite of how you want things to work.
Yes, I see where I was wrong on that. Thank you for getting me clear about it. :)
but lots of people do see a use for interpolated filters. the actual hardware unit i sampled here gives a smooth, fully variable filter sweep through its frequency range.
Again, what's great about your pack is you've provided both, so we're all happy.
if interpolation can't be proven by you or anyone else to have any negative impact on the sound, then why cripple my end result by using stepped which makes it less like the actual unit?
Right, and now we get to hear that for ourselves because you've provided both.
so a smoothly sweepable, fully variable filter is a 'lack of benefit'? ok, maybe you think so, but the problem is that when i make something i have to aim towards what i think the average person will think. let's say for comparison that you were arguing that having a 'dry/wet' mix control on every plug-in, wasn't needed because you saw it as a lack of benefit. but it doesn't hurt anything by being there, whether you use it or not. lots of other people would like it. as a developer should i be expected to not include the wet/dry slider because a handful of people start speaking out against it with claims that it hurts the sound, without giving any real reason or proof that it is?
A dry/wet control does not have the possibility of causing harm, at least as far as I'm aware. By your own assertion regarding the 75Hz to 160Hz example, an interpolating filter does have the possibility of causing harm. Anyway, you've provided both so everybody's happy. :)

Post

Uncle E wrote: Personally, I think you could simply acknowledge whatever limitations exist and...
i will,
the moment that you prove to me that they exist.
i'll be right here, waiting. because so far you haven't offered one shred of evidence. i don't expect this to change at this point. you are only using your 'feelings' as your argument that there 'might' be something wrong with interpolation. i'm using facts. as a developer of the platform for going on 3 yrs now, excuse me if i happen to think i know a bit about the subject. ;)
Uncle E wrote: Imagine if someone did release a filter pack that interpolated between 75Hz and 160Hz, don't you agree that it would do more harm than good to people's perceptions of the platform?
it sure would, and i've already explained that that's what i'm fighting against here. other people making sh*tty filters, or deciding to give an interpolated version of a stepped hardware filter just as a fun bonus, shouldn't affect anyone's perception of my recreation of a smoothly sweepable fully variable filter. consider what you are suggesting to me: that i limit/change what i do because other people might provide bad filters. how does that make sense? if i were a robot that kind of unlogic would cause my circuitry to fry.

but while we are on the subject of 'what if' scenarios, why don't you imagine if i were an algorithmic vst plug-in developer, and i were to release a HP or LP filter that were stepped at 10Hz steps. what would the general community here at KVR have to say about that after they tested it? they would be like:
"wtf?" -KVR community

and for good reason. it wouldn't make any sense to do that. it makes even less sense here where i am trying to accurately recreate a hardware unit. my original release did that pretty closely. the stepped version doesn't.

also in an earlier post you said that at least here with filters the interpolation is avoidable. well, not with these ones it isn't. not if you are using the dynamic versions which were the main set and a big part of what i set out to do. so even if you use the stepped versions with dynamics, you still are getting a continuously interpolated result.

Uncle E wrote: Again, what's great about your pack is you've provided both, so we're all happy.
yeah and that's why i did it, just to make people happy. :) but now i regret it because i feel like i just played along with completely unfounded paranoia, instead of taking the chance to dispel that paranoia. i'm trying to correct that now. why do i care so much? maybe it's because i'm also working on sampling the peak/notch filters on this unit. those could be used in a primarily mixing environment as you mentioned, just as any type of eq. but, the main reason i got this unit is because of the FX ability of those notches/peaks, when swept. it can be almost phaser-like.

now, if i were to submit to your request that i 'admit that there are limitations' with interpolation between frequency positions of a filter, then i would be basically admitting that i am developing a sweepable notch/peak filter that is bad at it's core because Nebula can't handle this kind of thing.

no.


anyone who wants to go solely based on something they heard (which was probably someone talking about a poorly sampled filter with interpolation between large gaps), or just their emotional belief that it's wrong because 'interpolation' is a bad, dirty word (even though they love the sound of interpolation and use Nebula with interpolation all the time), or whatever it is, well, they can miss out. i don't need those customers. not if it means limiting myself because of an incorrect belief. that's the only limitation i see here. and this is coming from a guy who has gone out of his way to offer plenty of free updates, and listened to lots of (reasonable) requests. and will continue doing so.
the only thing i will not continue doing is providing stepped versions where it's unnecessary, and after i've clearly demonstrated that interpolation doesn't hurt anything if done correctly. it's more work for me, it adds more clutter to the program selection menu, it can possibly confuse some users who now have to learn what the difference is and probably don't care anyway (and for good reason), it adds more files to deal with, and it's entirely pointless.

people were praising this filter BEFORE i gave the stepped versions.
Last edited by Cupwise on Thu Dec 13, 2012 5:33 pm, edited 1 time in total.

Post Reply

Return to “Effects”