Hi! Thank you for your critical engagement with our new plug-in. We always appreciate honest feedback. It does make us sweat every time to release a new plug-in and having it eyed critically 😅, however, it always helps us to improve our plug-ins. For pure:unmask we needed to opt for a dedicated sidechain because of the limitations we are facing with most DAWs. Whenever cross-channel communication is used for time-critical processing like ducking, we can’t guarantee everything works as it is supposed to. That said, to ensure a correct timing of the audio processing of pure:unmask we had to use the dedicated sidechain of the DAWs. The cross-channel processing we are using with our spectral mixing approach in smart:EQ 4, uses a combination of static and time-critical processing which works well in all DAWs.
@@Whiteseastudio I don't agree. You questioned why they didn't use the same method like in smart-EQ and they basically responded that they don't use the same method like in smart-EQ. But I think we can all imagine the reason
I think the priority is correct. If you were doing this with sidechain compression you would put the compressor on the track you wanted to suppress. So thinking in this same way you put this plugin, or TrackSpacer for that matter, on the track you want pushed down.
Sidechain routing is necessary for this type of plugin. Inter-plugin communication (like what smartEQ uses) is not sample accurate because it doesn't know about any delay introduced by other plugins and does not have PDC (plugin delay compensation). Sidechain routing, at least for most daws including Reaper, does use PDC. So, smartEQ is not reacting to what's going on in other instances, but it trains its curve based on the overall spectrum. Does that help?
I have Focusrite Fast reveal, it uses Sonible algorithms but is really well designed and has an advanced mode giving you all the control you need. Its been around for a while but works nicely..
Surely, if they are designing these pure plugins to be simple to use, they probably already know that making the plugins communicate without the need for sidechaining would be the way to go.. So I wonder if there's a good reason for going with sidechaining, such as limitations with the plugin communication option?
Except it doesn't work the same way at all. This is a spectral operation. Compression is not. That is quite fundamental difference which you are ignoring altogether all while it makes enormous difference in the sound. It's like comparing a crayon drawing to oil painting and claiming them to be the same. No, they're not. Both pieces can be enormously good looking but they're still nothing alike.
@@anteshell I'm not saying that it works the same way. I'm saying that the thought process of where to put the plugin and how it is side-chained is the same. What is true that it is that it is going to affect the spectral balance of the "less priority" track (not the high priority track), so it makes sense to put it on that track. My comment was only to point out the similarity in which track to put it on. Like if you put the plugin on the "high priority" track how is it going to affect the low priority one? Voodoo? You can't change the signal on a sidechained track. Only the track on which the plugin sits.
@@kenvives9366 "I don’t know. If I want to duck the bass out for the kick, I put the compressor on the bass channel." To me this clearly implies that you are criticising the plugin for doing the same as compressor. I cannot see how that would imply the practical use method of the plugin, and as they are very similar I don't see any sense criticising that either. Even after your response, I don't see a way to interpret it in any other way. Maybe it's a language barrier?
Look, I am not saying anything about how the thing works. I have smart eq4 and understand how this is different than using a compressor. In the video, Whitsze was saying that he thought the placement of the plugin on the lower priority track and sidechaning the high priority track was not intuitive to him. All I’m trying to say is that it makes sense to me because it is similar to the way I would traditionally duck the bass when the kick hits using a compressor. Jeez. I guess I didn’t think it needed so much clarification when I wrote it right after watching the video.
The benefit of the sidechain approach is you only have the CPU overhead of one plugin being loaded instead of two. The downside with the approach is what if you want more than one source to cut the track, you need to run more instances or presumably have multiple sidechains as a source and the source type will be incorrect for one or more of the sources.
I've used Trackspacer by Wavesfactory (Similar idea to this) in Reaper with two instances of it reacting to two different tracks 3-4 for the bass end and 5-6 for the top end. It sounds like it works although I have no way of objectively proving this.
It is actually quite different than Smart EQ. You put smart eq on every channel you want to ballance and the EQ is applied to everything, in relation. This is instead a sidechain spectral compressor. It makes perfect sense that you put the compressor on the track you want to "duck" out of the way of another track, and it also makes 0 sense to have to put a plugin on the track you want to unmask even though you do not do any processing on it, just for the inter-plugin communication. It seems quite similar to what Smar:tComp 2 can do, would be curious is getting this on top is worth it if I already have Smart:Comp.
@@jocke1972 Well it has the sidechain profiles powered by machine learning, that smart:comp does not have. But how impactful are those to the overall result ... I do not know.
I don’t think that the two plug-in intercommunication version would be easier to understand than the side chain version. You have to insert to plug-ins into two tracks, and you have to understand which one is the sending one on which one is the receiving one. So from my point of view, there is no difference with understanding the side chain way, or am I missing something?
You said it best when you said you basically just have to "trust your ears". It's great to have all this tech but just like vitamins, they should be considered as supplements and not substitutes for our human talents and abilities. I was just about to purchase this pluging and then came across your review. I asked myself, what did the pioneering guys do back in the day? Those engineers did incredible things with what they had. Now we've got all this stuff and I dare say that while some of it is indeed helpful, most of it just stifles our creativity and we're too obsessed with trying to make everything perfect and pretty. I'm going to go listen to some Led Zeppelin now and recalibrate my ears, to hell with A.I. Lol. PS Thank you for this excellent and unbiased review.
The issue with Trackspacer is that it uses a very low FFT length (64 spls/32 bins). It makes the plugin somewhat low latency and more accurate in the time domain, but also very rough sounding and very inaccurate in the frequency domain. This is especially noticeable in the high mids and highs, since a 32 bins FFT length makes each band 1.5kHz wide @48kHz sample rate (48000/32=1500)! TBProAudio DSEQ3 is IMO the perfect middle ground between Trackspacer, Soothe and Gullfoss. It's super flexible, very versatile, great sounding and comparatively cheap. However, now there's also a completely free one available: nih-plugs Spectral Compressor. It's not a dynamic EQ with spectral sidechain like Soothe and DSEQ, "only" a wideband spectral compressor with a threshold that you can tilt and curve (basically making it either a band-pass or band-reject filter depending on the curve), but it can do 99% of all the important stuff that Soothe and DSEQ do. It's also surprisingly CPU efficient! If you want to spectrally duck something like a synth beneath some vocals, I would definitely recommend trying out Spectral Compressor instead of Trackspacer.
I don’t see how this is a mind fuck, in daws like Ableton this is how we sidechain everything. There’s a small little drop down next to the plugin and we select the input and it’s done, its second nature to Ableton users at least
Hi Love your videos! Two points- 1. The unmask side chain must be the same as in your daw. If not, you can side chain two channels against each other (one in the DAW and one on sonibile separate system) 2. You put on the “masking” channel the name of the other instrument, so the unmask vst will know the “ important” frequencies of the instrument you ducking for. I hope that’s clear ❤
What you describing as far as inter-plugin communication sounds like what iZotope does with Neutron and particularly the Visual Mixer module. You have the Relay plugin in each track and then you put the Visual Mixer in the group/stereo bus and then go from there. But I think with the pure line, the idea is to have something that is simple to use but can do the job really well. I myself like using the pure line if I want something "quick and dirty" where I'm not spending a lot of time tweaking parameters. And similarly, this is appealing to those who are not engineering savvy. Because of this, I can see why the thought of setting up a sidechain when using a plug-in can be daunting. But then again, it's one of those where once you get it, you get it.
Well this has extra the Sidechain profile, which decides using machine learning what parts of the spectrum to cut, mased on the side-chain input type. Also, maybe it has better spectral processing (more transparent)? I don't know if it is the case, but it's a possibility.
@@AlexLapugean definitely much higher definition FFT. Trackspacer only uses 32 bins, while Sonible usually uses at least 2048 bins (at least they do in Smart:Comp2), maybe even more. Each "band" in Trackspacer is in other words 1.5kHz wide on a 48kHz samplerate, while a 2048 bins FFT size makes each "band" 23Hz wide at 48kHz samplerate.
@@nj1255 Can you link to something that confirms that Sonible:Unmask works at the resolution you're saying here? I've not seen anything that makes this claim
_"It's €99 euros until May 6th after that it's €149 euros"_ Not even close . . . pure:unmask is just £15/$19 if you already have any of the pure plugins - and £25/€29 for everyone else. The price you quoted is for the 'pure bundle' (all 5 pure plugins).
Sidechaining should not be hard to setup. At least in Cubase it’s just a button on top of every plugin. I find that definitely easier than putting an instance of a plugin on both tracks for „inter plugin communication“. After watching I’m now curious about how you actually do get mixing feedback of bands?
Interplugin communication is better for static stuff. Not good for things that are in the time domain (compression or multi-band). What you are asking is not possible without access to the back end infrastructure of all DAWs. I already asked them and it’s not feasible.
Not entirely sure, it usually works for me but it can be a little buggy if i try to move it while playing the track and having the eqs on dynamic mode!
Izotope Neutron 4 does the unmasking in the same fashion. You put an instance on the forward and background tracks and set up the sidechain on the background track so it'll quiet down (EQ-wise) dynamically as needed to let the forward track be heard. Simple and very effective (at least in Ableton).
I really appreciate your expertise, demand for greatness from these companies, AND huge compassion for beginner engineers... your community loves you ❤thank you so much 🙏
Ok I bit and bought it. Messing around with it I can see the value. Could you sidechain a compressor? Yes but I can see where if you blend it in with a subtle unmasking & speed setting, it does let the lead track stick out a little bit more. For the price, it's worth having it in my arsenal.
@@EpilepsyGod I get it. Trackspacer has a nicer gui and has a couple more adjustments. On sale, this one is not bad and it does work. I might still check out Trackspacer too for sure.
I do a similar thing with Soothe 2 , the perk of doing it with soothe 2 is that you can really hone in on the frequencies that are being cut from your sidechain “focus” . I actually don’t mind this as a quick sort of solution but I think I’ll stick to soothe for doing something like this for more control. Sometimes the way sounds combine or “mask” I don’t mind so much. Just depends on where in the fq spectrum and given scenario.
@@SoupOfSessions Sure, that makes sense. I use soothe for other applications as well so I guess that’s why it makes more sense for me personally, and why maybe I wouldn’t spend the money on this right now.
I respectfully disagree. Grabbing a sidechain input in Ableton Live for example is as simple as clicking to pull up the menu where all of the tracks are listed, and then clicking on that track. If that's more complicated on other DAWs I guess I'd understand, but from my perspective where its so easy to do in Ableton, adding 2 plugins on multiple tracks instead of just one plugin on one track is way better for saving computer resources. And that's especially relevant since I have pretty big projects. So I'd support an option for both if its really difficult to do in other DAWs but only if it was optional.
yeah i was going to ask what would be the different between this and just doing it with soothe .. well except the part where the "AI" choses where in the spectrum is the clashing occuring.
My guess is that the fancy interplugin communication stuff in Smart:EQ was still being developed when they started developing Pure:Unmask. It probably just wasn't ready for P:U to use during the development cycle.
I feel like it's Sonible's take on the idea behind Wavefactory's Trackspacer? Am I right, or am i missing something? I was also exactly hoping that they didnt use sidechain routing and used their own inter plugin talking, since Trackspacer already does sidechaining. It works fine once you get the hang of it but it can be a ressource hog (especially the way I use it) due to the nature of sidechain routing in most DAWs. On a more positive note, it does look like it has a higher resolution than Trackspacer with its limited amounts of bands.
I think they've got issues with their inter-plugin communication technology. They used it in smart:eq4 and that still doesn't work properly in some DAWs, even months after release and with no updated version being published.
@@tez4792 ah! Thanks for pointing that out this is a complaint I wasn't aware of because I frankly don't use that feature in smart eq4 since I'd rather do that part of the mixing myself
Works a lot like Focusrite Reveal, (Sonible AI) which I own, and which was more expensive, so this might be a cheaper alternative, doing basically the same thing with a bit less control.
Focusrite Fast: Reveal. I'm surprised this wasn't brought up. Focusrite partnered with Sonible to make the Fast stuff, and this Pure: Unmask looks like Fast:Reveal in "Pure" format... And I'm willing to bet it is exactly that, with an updated algorithm. I would love to see a comparison.
the instrument player mentality is so true, it took me years to "retrain" my ears from when I was a player and basically I had a monitoring mix for years in my ears and not a release mix
For the record, machine learning is a field of AI. What people are calling AI nowadays are just new applications of machine learning. Kinda weird to explain it by text in a comment box but given today's technology the two terms are interchangeable since there's no "modern" AI without ML and ML is AI by definition.
@@EpilepsyGod appreciate it, mate. It's what I do for a living actually. :) I'm pretty sure that most of the AI plugins are not AI but marketing. It's questionable that AI plugins pop up everywhere when the technology needed for this exists for years. Sonible looks legit though. They've been pioneering this before it was a fad.
I have a feeling that a smart:unmask version (with full visuals and more controls) may be in the works, but they can't get the inter-plugin communication to work with stability yet, as seen in their smart:eq4 release which STILL doesn't work properly in some DAWs for some people, including myself.
@@KYTHERAOfficial same here I couldn't get what the problem was with the routing. Like the plugin should be on the track being ducked and not the one ducking
I totally get why sonible did it that way. The other way around would mean that you boost frequencies to compete with the background track which could mess up gain staging. Manual unmasking would also most likely be done with a EQ cut on the track that occupies too much spectral space. I get it though that this might feel unusual when you are used to work with smart:eq that uses „inter-plugin communication“😊.
Ya The need to rethink this.. It like having to climb behind a rack and patch something in. Some times it just a bit more work then Im up to doing. If they have a way to make it easy! Then why not.
It's fine if you take it as a little baby plugin like the "pure" series is in general. But I have similar plugins that are very advanced and work very very well...
I saw that it is part of the pure bundle and one-knob plugins are not my favorite but maybe some people will be able to fit it in their projects. I dunno 🤷
might be useful if used in conjunction with blue cat audio mb-7 as there's not enough control for me in the unit itself. That said.....I doubt it'll do a better job than trackspacer.
another reason I love you channel is the music you use in your demonstrations, ( ironically in this particular video the music you used was terrible imo, but usually its super awesome ) . you should like make a playlist of all the songs you used in your videos, because all of those sandstorms are amazing.
I don't see much difference between "inter plugin communication" (IPC) and sidechain routing. For the first you need two instances of the plugin on both channels (one on the channel you want to hear in front and one you want to have in the back), for the latter you need only one instance (less CPU maybe?) on the channel you want to have in the background and select the channel you want to have in the front as sidechain, or am I overseeing something? In addition, for the first (IPC) mathod, when you do processing after the instance of the first plugin you have to move the plugin to the last position of the effect chain, to maintain the correct signal passed to the second instance. Seems cumbersome. But yeah, there are much other plugins who do the same thing with less CPU usage.
Interesting about the mix getting a tendency to the musicians instrument. I wonder what you would say about my music, since I play all the instruments and sing also... 🤔
@@tez4792 What DAW are you using? If you're using a DAW that can isolate plugin processes such as Bitwig (which is what I use) then you might need to make sure it's configured to create an instance per plugin type, rather than for each individual plugin, as the cross-instance communication won't work if the instances are separated into different processes. With this configured it works fine for me in Bitwig.
@@sburton84 _"Doesn't smart:EQ already do this?"_ Yeah, it does, so there would be some overlap, but smart:EQ focuses on EQ rather than spectral shaping (but then I suppose you could argue that spectral shaping is just EQ) . . . Hmmmm, who knows, maybe you're right.
Sidechain routing perhaps keeps the cost down? And fulfills a marketing need for users to step up to Smart EQ. The pure bundle are just baby steps into the jungle.
Soothe is expensive… but since you can do much more with it and it is in many producer/engineer’s arsenal already, this plugin only makes sense if you don’t already own soothe/trackspacer/even ozone or neutron
@White Sea Studio It would have been helpful on a video like this one if you had explained what result you were hoping to achieve. In other words, you mentioned there was mud, but what did you want to be happening with the ducking. I really couldn't hear any change with the drums, but you never said what we should be listening for.
Oh, and this is doing pretty similar things to what TrackSpacer and iZotope Neutron is doing. Haven't tried this to see if it sounds better, but not sure there's too much in the code that you could do differently.
You gave the price for the whole Pure bundle - not the Pure:Unmask plug-in on it's own - which is about 15-20 Euros depending on the other plug-ins you own.
I think you are trying to fix problems that dont exist. Pure:unmix only works with ONE signal. Do you know how complicated it is to explain the plug-in which instance it should react to? Also Trackspacer, soothe2, etc. are working with the same system. Sidechain is one of the easiest things to understand. In deutsch würden wir sagen.: Du hast aus einer Mücke einen Elefanten gemacht.
It doesn’t make any sense? What do you mean? Lol you usually do sidechain a track into unmasking plugins, unless it’s like neutron which will do that on it’s own (using 2 instances), like, you can do it with soothe 2 just like that too. You send the signal of the track that you want ‘in front’ and let it spectral correct a different track.
Although on a second thought, it sounds like you did it backwards somehow? It sounds like the guitar is now being unmasked from the drums instead of the other way around? Unless that’s what you intended… to actually mask the drums lol.
The difference I believe is that Trackspacer doesn’t know what type of sound or instrument the sidechain signal is and just uses the signal’s frequency response. Whereas Unmask has some nuance to be able to react differently depending on what profile you select for the sidechain signal. Exactly the lengths this difference goes to or the effectiveness of it, I am unsure at this point.
How would you do it the opposite way around?... it logically would need to be on the channel it is creating space for... otherwise it would be adding things to the channel you want forward to be more forward and it aims to make space... so it need to be on the channel space is to be made... I'm puzzled how you think it's better the other way around.. it will need the side channel to also hear the track it is making space to otherwise it's just a general setting for drums or bass or vocal or whatever... that setting is only so it can better act to what it is reading... if there is no side chain, it can not use any form of machine learning
It is not bad as a quick fixer, but not really a professional tool... maybe just to get a idea if you are a bit lost... then mimic it and work from there
Trackspacer has m/s control, I don’t like ai deciding for me on what frequencies to cut more, I tried this and couldn’t get same results as trackspacer…
1:51 just how trackspacer works. The sidechain ducks out the masked frequencies out of the track. Its just unfortunate that they called it unmask because it actually masks the track (ducks it out). You can also sidechain from a group track or side chain a group of instruments (all guitars] and let one vocal through…. Hiw else would you do it, have 2 instances of the plugin and call it receiver and sender? And then deal with different send groups and basically reinvent the wheel…
So if you had a kick and bass clashing, the kick would be “masked” by the bass. So, you use the plugin on the bass to “unmask” the kick. I think that’s why it’s named that way.
To do the smart EQ balancing you need to put plugin on every channel you want to balance, and the plugins are talking to each other (one ring to rule them all). this you put it only in the channel you want to effect. like dynamic eq with side chain on channel you want to make room for, but AI handling the frequencies being compressed/ducked. I get the issue with side channel understanding, it's an engineers tool, not every guitar player gets it.😜
@@JohanBuurma I’ve not had an email thus far. I am eagerly awaiting the update because I can’t use the groups function on my DAW at all. I saw on a forum that Sonible’s support team told a couple of people an update was coming very soon. The version on the website is still 1.0.1. Hopefully soon!
wouldn't you want to duck out the low end of guitar and let the bass handle low end? seems backwards at 4 min. i wonder if you can do both ways, limit bass ducking to mids and guitar ducking allowing bass to handle lows.
When playing live it is important for the player to hear themselves just a bit louder than every other player on stage. It always caries over to recording. Even when you play all of the instruments yourself it just gets worse, then you can't determine which instruments should be louder. 😜
I followed your link and you got the price way wrong somehow? Maybe only if you already own other Sonible plugin bundles as I do, but I just got it (and a buy one get one free €50 saturator) for just €28. Also, I demand sympathy. I have an ear infection in my right ear. I can barely hear a thing, and nothing over 250hz, on my heads right stereo in. Worst nightmare.
the best one for doing this is mspectraldynamics. still the best after all those years and super deep. Uses around the same amount of cpu and is linear phase as opposed to this. Sonible's AI isn't great, it's better to have manual control over it
Anyone talking about Fuser by Mastering The Mix? I really like it, and have the Sonible, Fuser and Trackspacer. Just wondering what people think of Fuser in the conversation
I like Fuser’s loudness match meter and phase rotation feature, and will definitely continue using it for that, but I wasn’t so pleased with its unmasking ability. From my brief testing pure:unmask seems to be more effective so far. I’d be tempted to run it through both, particular on a bass.
Hi! Thank you for your critical engagement with our new plug-in. We always appreciate honest feedback. It does make us sweat every time to release a new plug-in and having it eyed critically 😅, however, it always helps us to improve our plug-ins. For pure:unmask we needed to opt for a dedicated sidechain because of the limitations we are facing with most DAWs. Whenever cross-channel communication is used for time-critical processing like ducking, we can’t guarantee everything works as it is supposed to. That said, to ensure a correct timing of the audio processing of pure:unmask we had to use the dedicated sidechain of the DAWs. The cross-channel processing we are using with our spectral mixing approach in smart:EQ 4, uses a combination of static and time-critical processing which works well in all DAWs.
Fair enough, I feel like an idiot now 😅
@@Whiteseastudio I don't agree. You questioned why they didn't use the same method like in smart-EQ and they basically responded that they don't use the same method like in smart-EQ.
But I think we can all imagine the reason
@@Jaburu can we? What are we imagining?
@@Whiteseastudio maybe you could ask these questions of someone who knows the answers, before making a video that emphasises that opinion?
@@iainmcguire7190 I actually got confused with him comparing it to smart:eq. it is a light version of smart:comp. There is your reason
I think the priority is correct. If you were doing this with sidechain compression you would put the compressor on the track you wanted to suppress. So thinking in this same way you put this plugin, or TrackSpacer for that matter, on the track you want pushed down.
I agree. My verbiage for remembering is “put the compressor or dynamic EQ on the OFFENDER.”
I find sidechain routing much quicker and more simple than placing another instance on the other channel o.O
EXACTLY… as a professional with a highly viewed TH-cam channel it’s weird that he’s giving such a backwards take on this
Sidechain routing is necessary for this type of plugin. Inter-plugin communication (like what smartEQ uses) is not sample accurate because it doesn't know about any delay introduced by other plugins and does not have PDC (plugin delay compensation). Sidechain routing, at least for most daws including Reaper, does use PDC. So, smartEQ is not reacting to what's going on in other instances, but it trains its curve based on the overall spectrum. Does that help?
Interesting :)
thanks a lot for that insight
I’m perfectly happy with trackspacer.
I Got this cheaper than trackspacer or soothe so im happy too 😂
Trackspacer is legendary
Trackspacer absolutely devours CPU
Trackspacer all the way! I've got an old computer too! No issues.
Trackspacer 👍
I have Focusrite Fast reveal, it uses Sonible algorithms but is really well designed and has an advanced mode giving you all the control you need.
Its been around for a while but works nicely..
Fast Reveal works well for me too.
Surely, if they are designing these pure plugins to be simple to use, they probably already know that making the plugins communicate without the need for sidechaining would be the way to go.. So I wonder if there's a good reason for going with sidechaining, such as limitations with the plugin communication option?
Yes, dear Wytse, I'd love to know about Sonible and their struggle with development limitations. Thanks for the great videos!
I don’t know. If I want to duck the bass out for the kick, I put the compressor on the bass channel. Seems to work the same way and makes sense to me.
Except it doesn't work the same way at all. This is a spectral operation. Compression is not. That is quite fundamental difference which you are ignoring altogether all while it makes enormous difference in the sound.
It's like comparing a crayon drawing to oil painting and claiming them to be the same. No, they're not. Both pieces can be enormously good looking but they're still nothing alike.
@@anteshell I'm not saying that it works the same way. I'm saying that the thought process of where to put the plugin and how it is side-chained is the same. What is true that it is that it is going to affect the spectral balance of the "less priority" track (not the high priority track), so it makes sense to put it on that track. My comment was only to point out the similarity in which track to put it on. Like if you put the plugin on the "high priority" track how is it going to affect the low priority one? Voodoo? You can't change the signal on a sidechained track. Only the track on which the plugin sits.
Then you’re ducking the whole channel, as opposed to just the frequencies that the kick uses. Both ways work but yield different results
@@kenvives9366 "I don’t know. If I want to duck the bass out for the kick, I put the compressor on the bass channel."
To me this clearly implies that you are criticising the plugin for doing the same as compressor. I cannot see how that would imply the practical use method of the plugin, and as they are very similar I don't see any sense criticising that either.
Even after your response, I don't see a way to interpret it in any other way. Maybe it's a language barrier?
Look, I am not saying anything about how the thing works. I have smart eq4 and understand how this is different than using a compressor. In the video, Whitsze was saying that he thought the placement of the plugin on the lower priority track and sidechaning the high priority track was not intuitive to him. All I’m trying to say is that it makes sense to me because it is similar to the way I would traditionally duck the bass when the kick hits using a compressor. Jeez. I guess I didn’t think it needed so much clarification when I wrote it right after watching the video.
Smart EQ 4 does not have a dynamic unmasking. Something Sonible will need to address.
dosent the adaptive (i.e : Dynamic) option pushed to a 100 make the EQ curves adaptive, and as a result the unmasking too?
The benefit of the sidechain approach is you only have the CPU overhead of one plugin being loaded instead of two. The downside with the approach is what if you want more than one source to cut the track, you need to run more instances or presumably have multiple sidechains as a source and the source type will be incorrect for one or more of the sources.
You could bus your sidechain
I've used Trackspacer by Wavesfactory (Similar idea to this) in Reaper with two instances of it reacting to two different tracks 3-4 for the bass end and 5-6 for the top end. It sounds like it works although I have no way of objectively proving this.
It is actually quite different than Smart EQ. You put smart eq on every channel you want to ballance and the EQ is applied to everything, in relation. This is instead a sidechain spectral compressor. It makes perfect sense that you put the compressor on the track you want to "duck" out of the way of another track, and it also makes 0 sense to have to put a plugin on the track you want to unmask even though you do not do any processing on it, just for the inter-plugin communication.
It seems quite similar to what Smar:tComp 2 can do, would be curious is getting this on top is worth it if I already have Smart:Comp.
Yes, this is more ore less smartcomp with spectral mode on and side chaining active
@@jocke1972 Well it has the sidechain profiles powered by machine learning, that smart:comp does not have. But how impactful are those to the overall result ... I do not know.
I don’t think that the two plug-in intercommunication version would be easier to understand than the side chain version. You have to insert to plug-ins into two tracks, and you have to understand which one is the sending one on which one is the receiving one. So from my point of view, there is no difference with understanding the side chain way, or am I missing something?
I tend to agree with you.
You said it best when you said you basically just have to "trust your ears". It's great to have all this tech but just like vitamins, they should be considered as supplements and not substitutes for our human talents and abilities. I was just about to purchase this pluging and then came across your review. I asked myself, what did the pioneering guys do back in the day? Those engineers did incredible things with what they had. Now we've got all this stuff and I dare say that while some of it is indeed helpful, most of it just stifles our creativity and we're too obsessed with trying to make everything perfect and pretty. I'm going to go listen to some Led Zeppelin now and recalibrate my ears, to hell with A.I. Lol. PS Thank you for this excellent and unbiased review.
Like trackspacer. But maybe better because it was trained to cut "the right" frequencies depending on the source you want to "shine".
In theory yes, in practice doesn’t come close to track spacer!
@@Mrfpsara Totally disagree - both are great
The issue with Trackspacer is that it uses a very low FFT length (64 spls/32 bins). It makes the plugin somewhat low latency and more accurate in the time domain, but also very rough sounding and very inaccurate in the frequency domain. This is especially noticeable in the high mids and highs, since a 32 bins FFT length makes each band 1.5kHz wide @48kHz sample rate (48000/32=1500)! TBProAudio DSEQ3 is IMO the perfect middle ground between Trackspacer, Soothe and Gullfoss. It's super flexible, very versatile, great sounding and comparatively cheap.
However, now there's also a completely free one available: nih-plugs Spectral Compressor. It's not a dynamic EQ with spectral sidechain like Soothe and DSEQ, "only" a wideband spectral compressor with a threshold that you can tilt and curve (basically making it either a band-pass or band-reject filter depending on the curve), but it can do 99% of all the important stuff that Soothe and DSEQ do. It's also surprisingly CPU efficient! If you want to spectrally duck something like a synth beneath some vocals, I would definitely recommend trying out Spectral Compressor instead of Trackspacer.
@@nj1255Nice insights. Have you tried Hornet’s Sleek? Curious how that stacks up in this realm but it isn’t being talked about much.
@@nj1255 It's always impressive when someone have facts and knowledge behind their opinions!
I don’t see how this is a mind fuck, in daws like Ableton this is how we sidechain everything. There’s a small little drop down next to the plugin and we select the input and it’s done, its second nature to Ableton users at least
Soothe also does this prolly just as good or better when selected sidechain mode as well
Hi
Love your videos!
Two points-
1. The unmask side chain must be the same as in your daw. If not, you can side chain two channels against each other (one in the DAW and one on sonibile separate system)
2. You put on the “masking” channel the name of the other instrument, so the unmask vst will know the “ important” frequencies of the instrument you ducking for.
I hope that’s clear ❤
What you describing as far as inter-plugin communication sounds like what iZotope does with Neutron and particularly the Visual Mixer module. You have the Relay plugin in each track and then you put the Visual Mixer in the group/stereo bus and then go from there.
But I think with the pure line, the idea is to have something that is simple to use but can do the job really well. I myself like using the pure line if I want something "quick and dirty" where I'm not spending a lot of time tweaking parameters. And similarly, this is appealing to those who are not engineering savvy. Because of this, I can see why the thought of setting up a sidechain when using a plug-in can be daunting. But then again, it's one of those where once you get it, you get it.
So it’s basically Trackspacer from Waves Factory? 🤷🏽♂️
Well this has extra the Sidechain profile, which decides using machine learning what parts of the spectrum to cut, mased on the side-chain input type. Also, maybe it has better spectral processing (more transparent)? I don't know if it is the case, but it's a possibility.
@@AlexLapugean definitely much higher definition FFT. Trackspacer only uses 32 bins, while Sonible usually uses at least 2048 bins (at least they do in Smart:Comp2), maybe even more. Each "band" in Trackspacer is in other words 1.5kHz wide on a 48kHz samplerate, while a 2048 bins FFT size makes each "band" 23Hz wide at 48kHz samplerate.
@@nj1255 Can you link to something that confirms that Sonible:Unmask works at the resolution you're saying here? I've not seen anything that makes this claim
At the time Intro Sale its cheaper as track spacer.
No, it's basically Trackspacer from Sonible. 😂
How else would it compare one track to another without using a sidechain input??
_"It's €99 euros until May 6th after that it's €149 euros"_
Not even close . . . pure:unmask is just £15/$19 if you already have any of the pure plugins - and £25/€29 for everyone else.
The price you quoted is for the 'pure bundle' (all 5 pure plugins).
There's a written statement in the video at 12:56 that the price is for the whole pure bundle.
@@Fastvoice Sure, he's quoting - _and showing_ - the wrong price.
Sidechaining should not be hard to setup. At least in Cubase it’s just a button on top of every plugin. I find that definitely easier than putting an instance of a plugin on both tracks for „inter plugin communication“.
After watching I’m now curious about how you actually do get mixing feedback of bands?
Interplugin communication is better for static stuff. Not good for things that are in the time domain (compression or multi-band). What you are asking is not possible without access to the back end infrastructure of all DAWs. I already asked them and it’s not feasible.
Ah, could that be why their grouping function on smart:eq4 has so many issues?
Not entirely sure, it usually works for me but it can be a little buggy if i try to move it while playing the track and having the eqs on dynamic mode!
Izotope Neutron 4 does the unmasking in the same fashion. You put an instance on the forward and background tracks and set up the sidechain on the background track so it'll quiet down (EQ-wise) dynamically as needed to let the forward track be heard. Simple and very effective (at least in Ableton).
I really appreciate your expertise, demand for greatness from these companies, AND huge compassion for beginner engineers... your community loves you ❤thank you so much 🙏
Ok I bit and bought it. Messing around with it I can see the value. Could you sidechain a compressor? Yes but I can see where if you blend it in with a subtle unmasking & speed setting, it does let the lead track stick out a little bit more. For the price, it's worth having it in my arsenal.
Trackspacer and soothe2...
@@EpilepsyGod I get it. Trackspacer has a nicer gui and has a couple more adjustments. On sale, this one is not bad and it does work. I might still check out Trackspacer too for sure.
I do a similar thing with Soothe 2 , the perk of doing it with soothe 2 is that you can really hone in on the frequencies that are being cut from your sidechain “focus” . I actually don’t mind this as a quick sort of solution but I think I’ll stick to soothe for doing something like this for more control. Sometimes the way sounds combine or “mask” I don’t mind so much. Just depends on where in the fq spectrum and given scenario.
my initial reaction is: i would rather spend 30 on this than 200 on soothe
@@SoupOfSessions Sure, that makes sense. I use soothe for other applications as well so I guess that’s why it makes more sense for me personally, and why maybe I wouldn’t spend the money on this right now.
I respectfully disagree. Grabbing a sidechain input in Ableton Live for example is as simple as clicking to pull up the menu where all of the tracks are listed, and then clicking on that track. If that's more complicated on other DAWs I guess I'd understand, but from my perspective where its so easy to do in Ableton, adding 2 plugins on multiple tracks instead of just one plugin on one track is way better for saving computer resources. And that's especially relevant since I have pretty big projects. So I'd support an option for both if its really difficult to do in other DAWs but only if it was optional.
Just always used sidechained soothe for this
yea but the latency is always the biggest issue with this method :(
yeah i was going to ask what would be the different between this and just doing it with soothe .. well except the part where the "AI" choses where in the spectrum is the clashing occuring.
I play all the instruments and sing which is why everything is turned up 👍
🤣
My guess is that the fancy interplugin communication stuff in Smart:EQ was still being developed when they started developing Pure:Unmask. It probably just wasn't ready for P:U to use during the development cycle.
I feel like it's Sonible's take on the idea behind Wavefactory's Trackspacer? Am I right, or am i missing something?
I was also exactly hoping that they didnt use sidechain routing and used their own inter plugin talking, since Trackspacer already does sidechaining. It works fine once you get the hang of it but it can be a ressource hog (especially the way I use it) due to the nature of sidechain routing in most DAWs.
On a more positive note, it does look like it has a higher resolution than Trackspacer with its limited amounts of bands.
I think they've got issues with their inter-plugin communication technology. They used it in smart:eq4 and that still doesn't work properly in some DAWs, even months after release and with no updated version being published.
@@tez4792 ah! Thanks for pointing that out this is a complaint I wasn't aware of because I frankly don't use that feature in smart eq4 since I'd rather do that part of the mixing myself
@Wytse @ 1:30 nice desk! is it custom made? would love to see a video tour of your hardware gear. peace
No problem with their side chaining choices here, and I like the amount of visual feedback. Trying to justify the price though. How is it on CPU?
I really liked the song in the video, maybe you could link the artist?
Works a lot like Focusrite Reveal, (Sonible AI) which I own, and which was more expensive, so this might be a cheaper alternative, doing basically the same thing with a bit less control.
I noticed the same
I like the FAST UI better, but it looks functionally the same to me.
@@gutterg0d I think sonible’s pure plugins are just their own branded versions of the Fast plugins they did for Focusrite.
They should have made a "listening" part of this plugin to put on the track that you wan't to sidechain to. (Tvo plugins)
Then interplugin communication would be a breeze too 😊
What about using test tracks it’s possible to unmask without messing up the target track? Just asking.
Funny, in my mind it’s the perfect way to route it 😂
Me too. Is this not exactly the same way trackspacer is routed?
Best video you ever made, this was a really honest video, a real eye opener (and I’m not talking about the plugin)
Just curious: how many peeps using some type of spectral unmasking on the regular?
Focusrite Fast: Reveal. I'm surprised this wasn't brought up. Focusrite partnered with Sonible to make the Fast stuff, and this Pure: Unmask looks like Fast:Reveal in "Pure" format... And I'm willing to bet it is exactly that, with an updated algorithm.
I would love to see a comparison.
the instrument player mentality is so true, it took me years to "retrain" my ears from when I was a player and basically I had a monitoring mix for years in my ears and not a release mix
For the record, machine learning is a field of AI. What people are calling AI nowadays are just new applications of machine learning. Kinda weird to explain it by text in a comment box but given today's technology the two terms are interchangeable since there's no "modern" AI without ML and ML is AI by definition.
Yeah, good for you to understand this)
They can say that gullfoss is AI too, but it's not)
@@EpilepsyGod appreciate it, mate. It's what I do for a living actually. :) I'm pretty sure that most of the AI plugins are not AI but marketing. It's questionable that AI plugins pop up everywhere when the technology needed for this exists for years. Sonible looks legit though. They've been pioneering this before it was a fad.
Is sidechaining complicated in pro tools or something? Its a quick drop down menu in Ableton, never had a problem even when I first started
I have a feeling that a smart:unmask version (with full visuals and more controls) may be in the works, but they can't get the inter-plugin communication to work with stability yet, as seen in their smart:eq4 release which STILL doesn't work properly in some DAWs for some people, including myself.
Pretty much smartcomp with inter process communication then
You are completely correct about the routing. It is way more intuitive to do it the opposite way.
Obviously subjective, I had never thought someone would find it more logical this way - I find it confusing if it was the other way around.
@@KYTHERAOfficial same here I couldn't get what the problem was with the routing. Like the plugin should be on the track being ducked and not the one ducking
@@yeezythabest I agree with you
I totally get why sonible did it that way. The other way around would mean that you boost frequencies to compete with the background track which could mess up gain staging. Manual unmasking would also most likely be done with a EQ cut on the track that occupies too much spectral space.
I get it though that this might feel unusual when you are used to work with smart:eq that uses „inter-plugin communication“😊.
You know I wanna see that new unit on too of your Gainlab chanel strip is 😈
Ya The need to rethink this.. It like having to climb behind a rack and patch something in. Some times it just a bit more work then Im up to doing. If they have a way to make it easy! Then why not.
It's fine if you take it as a little baby plugin like the "pure" series is in general. But I have similar plugins that are very advanced and work very very well...
I saw that it is part of the pure bundle and one-knob plugins are not my favorite but maybe some people will be able to fit it in their projects. I dunno 🤷
might be useful if used in conjunction with blue cat audio mb-7 as there's not enough control for me in the unit itself. That said.....I doubt it'll do a better job than trackspacer.
Soothe replaced trackspacer for me. Side chain it and duck the instrumental under vox... A few dB does wonders
another reason I love you channel is the music you use in your demonstrations, ( ironically in this particular video the music you used was terrible imo, but usually its super awesome ) . you should like make a playlist of all the songs you used in your videos, because all of those sandstorms are amazing.
Funnily, if you're mixing you OWN vocals, you tend to make them 2-4 DB too quiet. It's universal!
Wow this is true. I do this myself
I find that the FASTreveal is the best unmasking plug-in. Arent’t they the same people behind it?
I don't see much difference between "inter plugin communication" (IPC) and sidechain routing.
For the first you need two instances of the plugin on both channels (one on the channel you want to hear in front and one you want to have in the back), for the latter you need only one instance (less CPU maybe?) on the channel you want to have in the background and select the channel you want to have in the front as sidechain, or am I overseeing something?
In addition, for the first (IPC) mathod, when you do processing after the instance of the first plugin you have to move the plugin to the last position of the effect chain, to maintain the correct signal passed to the second instance. Seems cumbersome. But yeah, there are much other plugins who do the same thing with less CPU usage.
I would say its an alternative to trackspacer, or?
Interesting about the mix getting a tendency to the musicians instrument.
I wonder what you would say about my music, since I play all the instruments and sing also... 🤔
_"They could have done better"_
I'm guess that sometime in the near future they will release 'smart:unmask' (rather than the simpler 'pure:unmask').
Doesn't smart:EQ already do this?
@@sburton84yes
@@sburton84 not in my DAW. The group function doesn’t work at all
@@tez4792 What DAW are you using? If you're using a DAW that can isolate plugin processes such as Bitwig (which is what I use) then you might need to make sure it's configured to create an instance per plugin type, rather than for each individual plugin, as the cross-instance communication won't work if the instances are separated into different processes. With this configured it works fine for me in Bitwig.
@@sburton84 _"Doesn't smart:EQ already do this?"_
Yeah, it does, so there would be some overlap, but smart:EQ focuses on EQ rather than spectral shaping (but then I suppose you could argue that spectral shaping is just EQ) . . . Hmmmm, who knows, maybe you're right.
Sidechain routing perhaps keeps the cost down? And fulfills a marketing need for users to step up to Smart EQ. The pure bundle are just baby steps into the jungle.
I mean if you own soothe all of these unmasking plugins become redundant…
exactly
Soothe Costs exactly 10 times more 🤣
@@jesperborgstrm3385 Fair point
Soothe is expensive… but since you can do much more with it and it is in many producer/engineer’s arsenal already, this plugin only makes sense if you don’t already own soothe/trackspacer/even ozone or neutron
@White Sea Studio It would have been helpful on a video like this one if you had explained what result you were hoping to achieve. In other words, you mentioned there was mud, but what did you want to be happening with the ducking. I really couldn't hear any change with the drums, but you never said what we should be listening for.
Oh, and this is doing pretty similar things to what TrackSpacer and iZotope Neutron is doing. Haven't tried this to see if it sounds better, but not sure there's too much in the code that you could do differently.
Gullfoss has sidechain functionality too, but I’ve never tried it
You gave the price for the whole Pure bundle - not the Pure:Unmask plug-in on it's own - which is about 15-20 Euros depending on the other plug-ins you own.
The first fact is clearly written in the video at 12:56.
I think you are trying to fix problems that dont exist. Pure:unmix only works with ONE signal. Do you know how complicated it is to explain the plug-in which instance it should react to? Also Trackspacer, soothe2, etc. are working with the same system. Sidechain is one of the easiest things to understand. In deutsch würden wir sagen.: Du hast aus einer Mücke einen Elefanten gemacht.
It doesn’t make any sense? What do you mean? Lol you usually do sidechain a track into unmasking plugins, unless it’s like neutron which will do that on it’s own (using 2 instances), like, you can do it with soothe 2 just like that too. You send the signal of the track that you want ‘in front’ and let it spectral correct a different track.
Although on a second thought, it sounds like you did it backwards somehow? It sounds like the guitar is now being unmasked from the drums instead of the other way around? Unless that’s what you intended… to actually mask the drums lol.
I’m wondering if they’ll add the feature into there’s next comp I mean they’re on smart eq 4 they need a smart comp 3 by now
Trackspacer already does this and works in exactly the same way?
The difference I believe is that Trackspacer doesn’t know what type of sound or instrument the sidechain signal is and just uses the signal’s frequency response. Whereas Unmask has some nuance to be able to react differently depending on what profile you select for the sidechain signal.
Exactly the lengths this difference goes to or the effectiveness of it, I am unsure at this point.
How would you do it the opposite way around?... it logically would need to be on the channel it is creating space for... otherwise it would be adding things to the channel you want forward to be more forward and it aims to make space... so it need to be on the channel space is to be made... I'm puzzled how you think it's better the other way around.. it will need the side channel to also hear the track it is making space to otherwise it's just a general setting for drums or bass or vocal or whatever... that setting is only so it can better act to what it is reading... if there is no side chain, it can not use any form of machine learning
If it is good or not is another thing... but could not see it working the other way around
It is not bad as a quick fixer, but not really a professional tool... maybe just to get a idea if you are a bit lost... then mimic it and work from there
Trackspacer has m/s control, I don’t like ai deciding for me on what frequencies to cut more, I tried this and couldn’t get same results as trackspacer…
1:51 just how trackspacer works. The sidechain ducks out the masked frequencies out of the track. Its just unfortunate that they called it unmask because it actually masks the track (ducks it out). You can also sidechain from a group track or side chain a group of instruments (all guitars] and let one vocal through…. Hiw else would you do it, have 2 instances of the plugin and call it receiver and sender? And then deal with different send groups and basically reinvent the wheel…
So if you had a kick and bass clashing, the kick would be “masked” by the bass. So, you use the plugin on the bass to “unmask” the kick. I think that’s why it’s named that way.
Unmask is good for integrating a vocal in the mix.
I can't see the issues with this ? You put the plugin on what you want to mask. Makes perfect senes to me.
It makes perfect sense. Trackspacer does the same thing. It works!
hi
I used soothe 2 for unmasking purpose in side chain mode.
does anybody have the possibility to compare pure:unmask to soothe 2?
***
To do the smart EQ balancing you need to put plugin on every channel you want to balance, and the plugins are talking to each other (one ring to rule them all). this you put it only in the channel you want to effect. like dynamic eq with side chain on channel you want to make room for, but AI handling the frequencies being compressed/ducked. I get the issue with side channel understanding, it's an engineers tool, not every guitar player gets it.😜
wait... was that Sandstorm by Darude?
Does anybody know what the latest update to smart eq 4 is?
They have released an update?
I never got an email about an update, though I've heard others have.
smart:eq4's bugs make the plugin complete crap + mostly unusable
@@JohanBuurma I’ve not had an email thus far. I am eagerly awaiting the update because I can’t use the groups function on my DAW at all. I saw on a forum that Sonible’s support team told a couple of people an update was coming very soon. The version on the website is still 1.0.1. Hopefully soon!
at this point I hope Sonible pays you a hell of a sponsorship because every plugin I bought from them was because of your reviews! :)
Can you do this w soothe 2
wouldn't you want to duck out the low end of guitar and let the bass handle low end? seems backwards at 4 min. i wonder if you can do both ways, limit bass ducking to mids and guitar ducking allowing bass to handle lows.
That sounds like you’d create a feedback loop of 2 channels infinitely sidechaining each other haha
it works exactly like any other plugins this type. You put it on a track that needs to be ducked.
This is exactly what soothe does (can do) no?
When playing live it is important for the player to hear themselves just a bit louder than every other player on stage. It always caries over to recording. Even when you play all of the instruments yourself it just gets worse, then you can't determine which instruments should be louder. 😜
Shouldn't be unintuitive for audio engineer to use (spectral) side chain compression... Sidechain profile doesn't make it that much more complicated.
Focusrite fast reveal is so much better...
People who understand unmasking should be able to understand sidechaining. Sidechaning is a breeze in Cubase. Just my 2 cents.
Oh, you have a new intro?! I like the smooth sound!
So basically, the exact same thing Pro Q3 can do but with an AI twist? is that it...?
Basically a thing that soothe2 in sidechain mode can do on a wider band with more control
I followed your link and you got the price way wrong somehow? Maybe only if you already own other Sonible plugin bundles as I do, but I just got it (and a buy one get one free €50 saturator) for just €28. Also, I demand sympathy. I have an ear infection in my right ear. I can barely hear a thing, and nothing over 250hz, on my heads right stereo in. Worst nightmare.
He was talking about the bundle price (as written in the video at 12:56) - not for unmask alone. E.g.: I could get unmask for 19 € if I wanted.
@@Fastvoice He didn't say bundle...but what about symphony for my ear?
There is some very serious "machine learning" tech on the way, ground breaking and extremely exciting⏳
its a bit like trackspacer
It's a ducker, it compresses freq band when side chain gets signal. nothing new. just AI applied and done by freq.
Faser is easier for this I feel
the best one for doing this is mspectraldynamics. still the best after all those years and super deep. Uses around the same amount of cpu and is linear phase as opposed to this. Sonible's AI isn't great, it's better to have manual control over it
Side chaining in logic is a breeze.
This is pretty much trackspacer
Big fan of Sonible as well but TrackSpacer is an order of magnitude better than Unmask.
WAITING FOR SMART:UNMASK :D
Anyone talking about Fuser by Mastering The Mix? I really like it, and have the Sonible, Fuser and Trackspacer. Just wondering what people think of Fuser in the conversation
I like Fuser’s loudness match meter and phase rotation feature, and will definitely continue using it for that, but I wasn’t so pleased with its unmasking ability. From my brief testing pure:unmask seems to be more effective so far. I’d be tempted to run it through both, particular on a bass.